1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.553 2008/09/05 16:59:47 fanf2 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TK/01 Added patch by Johannes Berg that expands the main option
10 "spamd_servers" if it starts with a dollar sign.
12 TK/02 Write list of recipients to X-Envelope-Sender header when building
13 the mbox-format spool file for content scanning (suggested by Jakob
16 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
17 (http://www.dcc-servers.net/) support via dccifd. Activated by
18 setting EXPERIMENTAL_DCC=yes in Local/Makefile. Check out
19 experimental_spec.txt for more documentation.
21 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
22 by Mark Daniel Reidel <mr@df.eu>.
24 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
25 When building exim an external PCRE library is now needed -
26 PCRE is a system library on the majority of modern systems.
27 See entry on PCRE_LIBS in EDITME file.
29 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
30 conversation. Added nologin parameter to request.
31 Patch contributed by Kirill Miazine
33 TF/01 Do not log submission mode rewrites if they do not change the address.
35 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
37 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
38 log files in place. Contributed by Roberto Lima
40 NM/04 Bugzilla 667: close socket used by dovecot authenticator
42 TF/03 Bugzilla 615: When checking the local_parts router precondition
43 after a local_part_suffix or local_part_prefix option, Exim now
44 does not use the address's named list lookup cache, since this
45 contains cached lookups for the whole local part.
47 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
48 Robert Millan. Documentation is in experimental-spec.txt
50 TF/04 Bugzilla 668: Fix parallel build (make -j).
52 NM/05 Bugzilla 437: Prevent Maildix aux files being created with mode 000
54 NM/05 Bugzilla 598: Improvedment to Dovecot authenticator handling.
55 Patch provided by Jan Srzednicki
57 TF/05 Leading white space used to be stripped from $spam_report which
58 wrecked the formatting. Now it is preserved.
60 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
61 that they are available at delivery time.
63 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
65 TF/08 TLS error reporting now respects the incoming_interface and
66 incoming_port log selectors.
68 TF/09 Produce a more useful error message if an SMTP transport's hosts
69 setting expands to an empty string.
75 TK/01 Add preliminary DKIM support. Currently requires a forked version of
76 ALT-N's libdkim that I have put here:
77 http://duncanthrax.net/exim-experimental/
79 Note to Michael Haardt: I had to rename some vars in sieve.c. They
80 were called 'true' and it seems that C99 defines that as a reserved
81 keyword to be used with 'bool' variable types. That means you could
82 not include C99-style headers which use bools without triggering
83 build errors in sieve.c.
85 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
86 as mailq or other aliases. Changed the --help handling significantly
87 to do whats expected. exim_usage() emits usage/help information.
89 SC/01 Added the -bylocaldomain option to eximstats.
91 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
93 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
95 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
101 PH/01 Another patch from the Sieve maintainer.
103 PH/02 When an IPv6 address is converted to a string for single-key lookup
104 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
105 dots are used instead of colons so that keys in lsearch files need not
106 contain colons. This was done some time before quoting was made available
107 in lsearch files. However, iplsearch files do require colons in IPv6 keys
108 (notated using the quote facility) so as to distinguish them from IPv4
109 keys. This meant that lookups for IP addresses in host lists did not work
110 for iplsearch lookups.
112 This has been fixed by arranging for IPv6 addresses to be expressed with
113 colons if the lookup type is iplsearch. This is not incompatible, because
114 previously such lookups could never work.
116 The situation is now rather anomolous, since one *can* have colons in
117 ordinary lsearch keys. However, making the change in all cases is
118 incompatible and would probably break a number of configurations.
120 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
123 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
124 conversion specification without a maximum field width, thereby enabling
125 a rogue spamd server to cause a buffer overflow. While nobody in their
126 right mind would setup Exim to query an untrusted spamd server, an
127 attacker that gains access to a server running spamd could potentially
128 exploit this vulnerability to run arbitrary code as the Exim user.
130 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
131 $primary_hostname instead of what libspf2 thinks the hosts name is.
133 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
134 a directory entry by the name of the lookup key. Previously, if a
135 symlink pointed to a non-existing file or a file in a directory that
136 Exim lacked permissions to read, a lookup for a key matching that
137 symlink would fail. Now it is enough that a matching directory entry
138 exists, symlink or not. (Bugzilla 503.)
140 PH/03 The body_linecount and body_zerocount variables are now exported in the
143 PH/04 Added the $dnslist_matched variable.
145 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
146 This means they are set thereafter only if the connection becomes
149 PH/06 Added the client_condition to authenticators so that some can be skipped
150 by clients under certain conditions.
152 PH/07 The error message for a badly-placed control=no_multiline_responses left
153 "_responses" off the end of the name.
155 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
157 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
158 (without spaces) instead of just copying the configuration text.
160 PH/10 Added the /noupdate option to the ratelimit ACL condition.
162 PH/11 Added $max_received_linelength.
164 PH/12 Added +ignore_defer and +include_defer to host lists.
166 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
167 way in which PCRE > 7.0 is built.
169 PH/14 Implemented queue_only_load_latch.
171 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
172 MAIL command. The effect was to mangle the value on 64-bit systems.
174 PH/16 Another patch from the Sieve maintainer.
176 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
178 PH/18 If a system quota error occurred while trying to create the file for
179 a maildir delivery, the message "Mailbox is full" was not appended to the
180 bounce if the delivery eventually timed out. Change 4.67/27 below applied
181 only to a quota excession during the actual writing of the file.
183 PH/19 It seems that peer DN values may contain newlines (and other non-printing
184 characters?) which causes problems in log lines. The DN values are now
185 passed through string_printing() before being added to log lines.
187 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
188 and InterBase are left for another time.)
190 PH/21 Added message_body_newlines option.
192 PH/22 Guard against possible overflow in moan_check_errorcopy().
194 PH/23 POSIX allows open() to be a macro; guard against that.
196 PH/24 If the recipient of an error message contained an @ in the local part
197 (suitably quoted, of course), incorrect values were put in $domain and
198 $local_part during the evaluation of errors_copy.
204 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
205 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
208 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
209 issue a MAIL command.
211 PH/02 In an ACL statement such as
213 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
215 if a client was not listed at all, or was listed with a value other than
216 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
217 the condition was not true (as it should be), so access was not denied.
218 The bug was that the ! inversion was incorrectly passed on to the second
219 item. This has been fixed.
221 PH/03 Added additional dnslists conditions == and =& which are different from
222 = and & when the dns lookup returns more than one IP address.
224 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
225 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
227 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
228 FSYNC, which compiles an option called disable_fsync that allows for
229 bypassing fsync(). The documentation is heavily laced with warnings.
231 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
233 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
234 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
235 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
236 including adding "make clean"; (3) Added -fPIC when compiling the test
237 dynamically loaded module, to get rid of a warning.
239 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
240 message fails, move_frozen_messages = true and ignore_bounce_errors_after
241 = 0s. The bug is otherwise harmless.
243 PH/07 There was a bug in the dovecot authenticator such that the value of
244 $auth1 could be overwritten, and so not correctly preserved, after a
245 successful authentication. This usually meant that the value preserved by
246 the server_setid option was incorrect.
248 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
250 PH/09 Installed PCRE release 7.0.
252 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
253 run for batched SMTP input. It is now run at the start of every message
254 in the batch. While fixing this I discovered that the process information
255 (output by running exiwhat) was not always getting set for -bs and -bS
256 input. This is fixed, and it now also says "batched" for BSMTP.
258 PH/11 Added control=no_pipelining.
260 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
261 patch, slightly modified), and move the expansion of helo_data till after
262 the connection is made in the smtp transport (so it can use these
265 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
267 PH/14 Added log_selector = +pid.
269 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
271 PH/16 Add ${if forany and ${if forall.
273 PH/17 Added dsn_from option to vary the From: line in DSNs.
275 PH/18 Flush SMTP output before performing a callout, unless control =
276 no_callout_flush is set.
278 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
279 was true (the default) a successful delivery failed to delete the retry
280 item, thus causing premature timeout of the address. The bug is now
283 PH/20 Added hosts_avoid_pipelining to the smtp transport.
285 PH/21 Long custom messages for fakedefer and fakereject are now split up
286 into multiline reponses in the same way that messages for "deny" and
287 other ACL rejections are.
289 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
290 with slight modification.
292 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
293 draft, changing the syntax and factoring some duplicate code".
295 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
296 for deliveries of the second and subsequent messages over the same SMTP
299 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
300 ${reduce, with only minor "tidies".
302 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
304 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
305 expansion side effects.
307 PH/27 When a message times out after an over-quota error from an Exim-imposed
308 quota, the bounce message says "mailbox is full". This message was not
309 being given when it was a system quota that was exceeded. It now should
312 MH/03 Made $recipients available in local_scan(). local_scan() already has
313 better access to the recipient list through recipients_list[], but
314 $recipients can be useful in postmaster-provided expansion strings.
316 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
317 in the case of a MAIL command with additional options following the
318 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
319 were accidentally chopped off.
321 PH/29 SMTP synchronization checks are implemented when a command is read -
322 there is a check that no more input is waiting when there shouldn't be
323 any. However, for some commands, a delay in an ACL can mean that it is
324 some time before the response is written. In this time, more input might
325 arrive, invalidly. So now there are extra checks after an ACL has run for
326 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
327 pipelining has not been advertised.
329 PH/30 MH's patch to allow iscntrl() characters to be list separators.
331 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
332 returned in the SMTP response when smtp_return_error_details was false.
335 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
336 instead of the C I/O that was originally supplied, because problems were
339 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
340 Exim which did not show up earlier: it was assuming that a call to
341 SSL_CTX_set_info_callback() might give an error value. In fact, there is
342 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
343 was a macro that became an assignment, so it seemed to work. This has
344 changed to a proper function call with a void return, hence the compile
345 error. Exim's code has been fixed.
347 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
350 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
352 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
354 PH/37 If a message is not accepted after it has had an id assigned (e.g.
355 because it turns out to be too big or there is a timeout) there is no
356 "Completed" line in the log. When some messages of this type were
357 selected by exigrep, they were listed as "not completed". Others were
358 picked up by some special patterns. I have improved the selection
359 criteria to be more general.
361 PH/38 The host_find_failed option in the manualroute router can now be set
362 to "ignore", to completely ignore a host whose IP address cannot be
363 found. If all hosts are ignored, the behaviour is controlled by the new
364 host_all_ignored option.
366 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
367 homing or because of multiple MX records with /mx) generated more than
368 one IP address, and the following item turned out to be the local host,
369 all the secondary addresses of the first item were incorrectly removed
370 from the list, along with the local host and any following hosts (which
371 is what is supposed to happen).
373 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
374 whoever called Exim into the -H file. In the case of the daemon it was
375 behaving confusingly. When first started, it used values for whoever
376 started the daemon, but after a SIGHUP it used the Exim user (because it
377 calls itself on a restart). I have changed the code so that it now always
380 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
381 message are rejected with the same error (e.g. no authentication or bad
382 sender address), and a DATA command is nevertheless sent (as can happen
383 with PIPELINING or a stupid MUA), the error message that was given to the
384 RCPT commands is included in the rejection of the DATA command. This is
385 intended to be helpful for MUAs that show only the final error to their
388 PH/42 Another patch from the Sieve maintainer.
390 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
391 Eximstats - Fixed some broken HTML links and added missing column headers
393 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
394 columns for Rejects, Temp Rejects, Ham, and Spam rows.
396 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
398 PH/43 Yet another patch from the Sieve maintainer.
400 PH/44 I found a way to check for a TCP/IP connection going away before sending
401 the response to the final '.' that terminates a message, but only in the
402 case where the client has not sent further data following the '.'
403 (unfortunately, this is allowed). However, in many cases there won't be
404 any further data because there won't be any more messages to send. A call
405 to select() can be used: if it shows that the input is "ready", there is
406 either input waiting, or the socket has been closed. An attempt to read
407 the next input character can distinguish the two cases. Previously, Exim
408 would have sent an OK response which the client would never have see.
409 This could lead to message repetition. This fix should cure that, at
410 least in a lot of common cases.
412 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
413 advertised in response to EHLO.
419 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
420 fixed by 4.65/MH/01 (is this a record?) are fixed:
422 (i) An empty string was always treated as zero by the numeric comparison
423 operators. This behaviour has been restored.
425 (ii) It is documented that the numeric comparison operators always treat
426 their arguments as decimal numbers. This was broken in that numbers
427 starting with 0 were being interpreted as octal.
429 While fixing these problems I realized that there was another issue that
430 hadn't been noticed. Values of message_size_limit (both the global option
431 and the transport option) were treated as octal if they started with 0.
432 The documentation was vague. These values are now always treated as
433 decimal, and I will make that clear in the documentation.
439 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
440 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
443 MH/01 Don't check that the operands of numeric comparison operators are
444 integers when their expansion is in "skipping" mode (fixes bug
445 introduced by 4.64-PH/07).
447 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
448 child addresses, Exim now panics and dies. Previously, because the count
449 is held in a short int, deliveries were likely to be lost. As such a
450 large number of recipients for a single message is ridiculous
451 (performance will be very, very poor), I have chosen to impose a limit
452 rather than extend the field.
458 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
459 leftover -K file (the existence of which was triggered by #402).
460 While we were at it, introduced process PID as part of the -K
461 filename. This should rule out race conditions when creating
464 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
465 processing considerably. Previous code took too long for large mails,
466 triggering a timeout which in turn triggers #401.
468 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
469 in the DK code in transports.c. sendfile() is not really portable,
470 hence the _LINUX specificness.
472 TF/01 In the add_headers option to the mail command in an Exim filter,
473 there was a bug that Exim would claim a syntax error in any
474 header after the first one which had an odd number of characters
477 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
478 callout verification, Exim cached a "reject" for the entire domain. This
479 is correct for most verifications, but it is not correct for a recipient
480 verification with use_sender or use_postmaster set, because in that case
481 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
482 case of MAIL FROM:<> rejection from other early rejections (e.g.
483 rejection of HELO). When verifying a recipient using a non-null MAIL
484 address, the cache is ignored if it shows MAIL FROM:<> rejection.
485 Whatever the result of the callout, the value of the domain cache is
486 left unchanged (for any other kind of callout, getting as far as trying
487 RCPT means that the domain itself is ok).
489 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
492 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
493 manifest itself as EPIPE rather than ECONNECT. When tidying away a
494 session, the daemon ignores ECONNECT errors and logs others; it now
495 ignores EPIPE as well.
497 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
498 (quoted-printable decoding).
500 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
501 later the small subsequent patch to fix an introduced bug.
503 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
505 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
507 PH/08 An error is now given if message_size_limit is specified negative.
509 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
510 to be given (somewhat) arbitrary names.
512 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
515 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
516 miscellaneous code fixes
518 PH/10 Added the log_reject_target ACL modifier to specify where to log
521 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
522 hostname. This is wrong, because it relates to the incoming message (and
523 probably the interface on which it is arriving) and not to the outgoing
524 callout (which could be using a different interface). This has been
525 changed to use the value of the helo_data option from the smtp transport
526 instead - this is what is used when a message is actually being sent. If
527 there is no remote transport (possible with a router that sets up host
528 addresses), $smtp_active_hostname is used.
530 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
531 tweaks were necessary in order to get it to work (see also 21 below):
532 (a) The code assumed that strncpy() returns a negative number on buffer
533 overflow, which isn't the case. Replaced with Exim's string_format()
535 (b) There were several signed/unsigned issues. I just did the minimum
536 hacking in of casts. There is scope for a larger refactoring.
537 (c) The code used strcasecmp() which is not a standard C function.
538 Replaced with Exim's strcmpic() function.
539 (d) The code set only $1; it now sets $auth1 as well.
540 (e) A simple test gave the error "authentication client didn't specify
541 service in request". It would seem that Dovecot has changed its
542 interface. Fortunately there's a specification; I followed it and
543 changed what the client sends and it appears to be working now.
545 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
548 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
549 address A is aliased to B and C, where B exists and C does not. Without
550 -v the output is "A verified" because verification stops after a
551 successful redirection if more than one address is generated. However,
552 with -v the child addresses are also verified. Exim was outputting "A
553 failed to verify" and then showing the successful verification for C,
554 with its parentage. It now outputs "B failed to verify", showing B's
555 parentage before showing the successful verification of C.
557 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
558 look up a TXT record in a specific list after matching in a combined
561 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
562 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
563 they consult the DNS. I had assumed they would set it the way they
564 wanted; and indeed my experiments on Linux seem to show that in some
565 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
566 To be on the safe side, however, I have now made the interface to
567 host_find_byname() similar to host_find_bydns(), with an argument
568 containing the DNS resolver options. The host_find_byname() function now
569 sets these options at its start, just as host_find_bydns() does. The smtp
570 transport options dns_qualify_single and dns_search_parents are passed to
571 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
572 of host_find_byname() use the default settings of RES_DEFNAMES
573 (qualify_single) but not RES_DNSRCH (search_parents).
575 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
576 spool_read_header() do less string testing, by means of a preliminary
577 switch on the second character of optional "-foo" lines. (This is
578 overdue, caused by the large number of possibilities that now exist.
579 Originally there were few.) While I was there, I also converted the
580 str(n)cmp tests so they don't re-test the leading "-" and the first
581 character, in the hope this might squeeze out yet more improvement.
583 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
584 flag allowing group syntax was set by the header_syntax check but not
585 turned off, possible causing trouble later; (2) The flag was not being
586 set at all for the header_verify test, causing "group"-style headers to
587 be rejected. I have now set it in this case, and also caused header_
588 verify to ignore an empty address taken from a group. While doing this, I
589 came across some other cases where the code for allowing group syntax
590 while scanning a header line wasn't quite right (mostly, not resetting
591 the flag correctly in the right place). These bugs could have caused
592 trouble for malformed header lines. I hope it is now all correct.
594 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
595 with the "reply" argument non-NULL. The code, however (which originally
596 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
597 but it didn't always do it. This confused somebody who was copying the
598 code for some other use. I have removed all the tests.
600 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
601 feature that was used to support insecure browsers during the U.S. crypto
602 embargo. It requires special client support, and Exim is probably the
603 only MTA that supported it -- and would never use it because real RSA is
604 always available. This code has been removed, because it had the bad
605 effect of slowing Exim down by computing (never used) parameters for the
606 RSA_EXPORT functionality.
608 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
609 authenticator to fail if there's a tab character in the incoming data
610 (there should never be unless someone is messing about, as it's supposed
611 to be base64-encoded). Also added, on Timo's advice, the "secured" option
612 if the connection is using TLS or if the remote IP is the same as the
613 local IP, and the "valid-client-cert option" if a client certificate has
616 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
617 authenticators. This can be used for authorization after authentication
618 succeeds. (In the case of plaintext, it servers for both authentication
621 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
622 if any retry times were supplied.
624 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
625 connection, where there is no client IP address to check. In this
626 situation, the verify now always succeeds.
628 PH/25 Applied John Jetmore's -Mset patch.
630 PH/26 Added -bem to be like -Mset, but loading a message from a file.
632 PH/27 In a string expansion for a processed (not raw) header when multiple
633 headers of the same name were present, leading whitespace was being
634 removed from all of them, but trailing whitespace was being removed only
635 from the last one. Now trailing whitespace is removed from each header
636 before concatenation. Completely empty headers in a concatenation (as
639 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
640 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
642 PH/29 [Removed. This was a change that I later backed out, and forgot to
643 correct the ChangeLog entry (that I had efficiently created) before
644 committing the later change.]
646 PH/30 Exim was sometimes attempting to deliver messages that had suffered
647 address errors (4xx response to RCPT) over the same connection as other
648 messages routed to the same hosts. Such deliveries are always "forced",
649 so retry times are not inspected. This resulted in far too many retries
650 for the affected addresses. The effect occurred only when there were more
651 hosts than the hosts_max_try setting in the smtp transport when it had
652 the 4xx errors. Those hosts that it had tried were not added to the list
653 of hosts for which the message was waiting, so if all were tried, there
654 was no problem. Two fixes have been applied:
656 (i) If there are any address or message errors in an SMTP delivery, none
657 of the hosts (tried or untried) are now added to the list of hosts
658 for which the message is waiting, so the message should not be a
659 candidate for sending over the same connection that was used for a
660 successful delivery of some other message. This seems entirely
661 reasonable: after all the message is NOT "waiting for some host".
662 This is so "obvious" that I'm not sure why it wasn't done
663 previously. Hope I haven't missed anything, but it can't do any
664 harm, as the worst effect is to miss an optimization.
666 (ii) If, despite (i), such a delivery is accidentally attempted, the
667 routing retry time is respected, so at least it doesn't keep
668 hammering the server.
670 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
671 in ${readsocket because some servers need this prod.
673 PH/32 Added some extra debug output when updating a wait-xxx database.
675 PH/33 The hint "could be header name not terminated by colon", which has been
676 given for certain expansion errors for a long time, was not being given
677 for the ${if def:h_colon_omitted{... case.
679 PH/34 The spec says: "With one important exception, whenever a domain list is
680 being scanned, $domain contains the subject domain." There was at least
681 one case where this was not true.
683 PH/35 The error "getsockname() failed: connection reset by peer" was being
684 written to the panic log as well as the main log, but it isn't really
685 panic-worthy as it just means the connection died rather early on. I have
686 removed the panic log writing for the ECONNRESET error when getsockname()
689 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
690 runs only) independently of the message's sender address. This meant
691 that, if the 4xx error was in fact related to the sender, a different
692 message to the same recipient with a different sender could confuse
693 things. In particualar, this can happen when sending to a greylisting
694 server, but other circumstances could also provoke similar problems.
695 I have changed the default so that the retry time for these errors is now
696 based a combination of the sender and recipient addresses. This change
697 can be overridden by setting address_retry_include_sender=false in the
700 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
701 remote server are returned as part of bounce messages. This was not
702 happening for LMTP over a pipe (the lmtp transport), but now it is the
703 same for both kinds of LMTP.
705 PH/38 Despite being documented as not happening, Exim was rewriting addresses
706 in header lines that were in fact CNAMEs. This is no longer the case.
708 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
709 and queue runs started by the daemon processed all messages. This has
710 been fixed so that -R and -S can now usefully be given with -q<time>.
712 PH/40 Import PCRE release 6.7 (fixes some bugs).
714 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
716 PH/42 Give an error if -q is specified more than once.
718 PH/43 Renamed the variables $interface_address and $interface_port as
719 $received_ip_address and $received_port, to make it clear that these
720 values apply to message reception, and not to the outgoing interface when
721 a message is delivered. (The old names remain recognized, of course.)
723 PH/44 There was no timeout on the connect() call when using a Unix domain
724 socket in the ${readsocket expansion. There now is.
726 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
727 be meaningful with "accept".
729 SC/01 Eximstats V1.43
730 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
732 SC/02 Eximstats V1.44
733 Use a glob alias rather than an array ref in the generated
734 parser. This improves both readability and performance.
736 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
737 Collect SpamAssassin and rejection statistics.
738 Don't display local sender or destination tables unless
739 there is data to show.
740 Added average volumes into the top table text output.
742 SC/04 Eximstats V1.46
743 Collect data on the number of addresses (recipients)
744 as well as the number of messages.
746 SC/05 Eximstats V1.47
747 Added 'Message too big' to the list of mail rejection
748 reasons (thanks to Marco Gaiarin).
750 SC/06 Eximstats V1.48
751 Mainlog lines which have GMT offsets and are too short to
752 have a flag are now skipped.
754 SC/07 Eximstats V1.49 (Alain Williams)
755 Added the -emptyok flag.
757 SC/08 Eximstats V1.50
758 Fixes for obtaining the IP address from reject messages.
760 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
761 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
762 whitesspace changes from 4.64-PH/27
764 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
767 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
768 are found, allow negative numbers in numeric criteria)
770 JJ/06 exipick.20061117.2, added new $message_body_missing variable
772 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
773 to match changes made in 4.64-PH/43
775 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
777 PH/47 Put in an explicit test for a DNS lookup of an address record where the
778 "domain" is actually an IP address, and force a failure. This locks out
779 those revolvers/nameservers that support "A-for-A" lookups, in
780 contravention of the specifications.
782 PH/48 When a host name was looked up from an IP address, and the subsequent
783 forward lookup of the name timed out, the host name was left in
784 $sender_host_name, contrary to the specification.
786 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
787 restricted to single-key lookups, Exim was not diagnosing an error if
788 * or *@ was used with a query-style lookup.
790 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
792 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
793 long ago, but noone interested enough thought of it. Let's just say that
794 the "1.1" means that there are some new functions that weren't there at
795 some point in the past.
797 PH/51 Error processing for expansion failure of helo_data from an smtp
798 transport during callout processing was broken.
800 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
801 tested/used via the -bh/-bhc/-bs options.
803 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
804 bug, fixed in subsequent PCRE releases).
806 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
807 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
809 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
815 SC/01 Use a glob alias rather than an array ref in eximstats generated
816 parser. This improves both readability and performance.
818 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
819 Don't display local sender or destination tables in eximstats unless
820 there is data to show.
821 Added average volumes into the eximstats top table text output.
823 SC/03 Collect data on the number of addresses (recipients) as well
824 as the number of messages in eximstats.
826 TF/01 Correct an error in the documentation for the redirect router. Exim
827 does (usually) call initgroups() when daemonizing.
829 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
830 with consistent privilege compared to when running as a daemon.
832 TF/03 Note in the spec that $authenticated_id is not set for local
833 submissions from trusted users.
835 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
836 Thanks to Dean Brooks <dean@iglou.com> for the patch.
838 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
839 by adding some example configuration directives to the default
840 configuration file. A little bit of work is required to uncomment the
841 directives and define how usernames and passwords are checked, but
842 there is now a framework to start from.
844 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
845 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
846 without this. I don't know how relevant this is to other LDAP libraries.
848 PH/02 Add the verb name to the "unknown ACL verb" error.
850 PH/03 Magnus Holmgren's patch for filter_prepend_home.
852 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
854 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
855 directory not expanded when it should be if an expanded home directory
856 was set for the address (which is overridden by the transport).
858 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
861 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
862 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
863 because it is too late at that time, and has no effect.
865 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
866 security issue with \' (bugzilla #107). I could not use the
867 PQescapeStringConn() function, because it needs a PGconn value as one of
870 PH/08 When testing addresses using -bt, indicate those final addresses that
871 are duplicates that would not cause an additional delivery. At least one
872 person was confused, thinking that -bt output corresponded to deliveries.
873 (Suppressing duplicates isn't a good idea as you lose the information
874 about possibly different redirections that led to the duplicates.)
876 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
877 systems where poll() doesn't work, in particular OS X.
879 PH/10 Added more information to debugging output for retry time not reached.
881 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
882 operations in malware.c.
884 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
887 PH/13 If write_rejectlog was set false when logging was sent to syslog with
888 syslog_duplication set false, log lines that would normally be written
889 both the the main log and to the reject log were not written to syslog at
892 PH/14 In the default configuration, change the use of "message" in ACL warn
893 statements to "add_header".
895 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
896 not followed by a command (e.g. "seen endif").
898 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
899 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
902 PH/17 Added extra conditions to the default value of delay_warning_condition
906 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
907 { match{$h_precedence:}{(?i)bulk|list|junk} } \
908 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
911 The Auto-Submitted: and various List- headers are standardised, whereas I
912 don't think Precedence: ever was.
914 PH/18 Refactored debugging code in route_finduser() to show more information,
915 in particular, the error code if getpwnam() issues one.
917 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
918 This is apparently needed in addition to the PH/07 change above to avoid
919 any possible encoding problems.
921 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
922 but not after initializing Perl.
924 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
925 output them only if debugging. By default they are written stderr,
926 apparently, which is not desirable.
928 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
931 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
934 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
936 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
937 authenticated or an ident call has been made. Suppress the default
938 values for $authenticated_id and $authenticated_sender (but permit -oMai
939 and -oMas) when testing with -bh.
941 PH/24 Re-jigged the order of the tests in the default configuration so that the
942 tests for valid domains and recipients precede the DNS black list and CSA
943 tests, on the grounds that those ones are more expensive.
945 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
946 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
947 This bug exists in every version of Exim that I still have, right back to
950 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
951 However, an attempt to turn on case-sensitivity in a regex key by
952 including (?-i) didn't work because the subject string was already
953 lowercased, and the effects were non-intuitive. It turns out that a
954 one-line patch can be used to allow (?-i) to work as expected.
960 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
961 other effects) broke the use of negated acl sub-conditions.
963 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
966 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
967 "Deny" causes Exim to reject the incoming connection with a 554 error.
968 Unfortunately, if there is a major crisis, such as a disk failure,
969 tcp-wrappers gives "deny", whereas what one would like would be some
970 kind of temporary error. A kludge has been added to help with this.
971 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
972 554 error is used if errno is still zero or contains ENOENT (which occurs
973 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
976 PH/03 Add -lutil to the default FreeBSD LIBS setting.
978 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
979 errors. Otherwise a message that provokes a temporary error (when other
980 messages do not) can cause a whole host to time out.
982 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
983 addresses were routed directly to files or pipes from a redirect router.
984 File deliveries just didn't batch; pipe deliveries might have suffered
987 PH/06 A failure to get a lock for a hints database would erroneously always say
988 "Failed to get write lock", even when it was really a read lock.
990 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
991 of 0600. This has been changed to use the value of the lockfile_mode
992 option (which defaults to 0600).
994 PH/08 Applied small patch from the Sieve maintainer.
996 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
997 folder from quota calculations, a direct delivery into this folder messed
998 up the contents of the maildirsize file. This was because the regex was
999 used only to exclude .Trash (or whatever) when the size of the mailbox
1000 was calculated. There was no check that a delivery was happening into an
1001 excluded directory. This bug has been fixed by ignoring all quota
1002 processing for deliveries into excluded directories.
1004 PH/10 Added the maildirfolder_create_regex option to appendfile.
1010 PH/01 The code for finding all the local interface addresses on a FreeBSD
1011 system running IPv6 was broken. This may well have applied to all BSD
1012 systems, as well as to others that have similar system calls. The broken
1013 code found IPv4 interfaces correctly, but gave incorrect values for the
1014 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1015 that it would not match correctly against @[] and not recognize the IPv6
1018 PH/02 The ipliteral router was not recognizing addresses of the form user@
1019 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1021 PH/03 Added disable_ipv6.
1023 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1024 decoded form, because it is most often used to construct To: headers
1025 lines in autoreplies, and the decoded form may well be syntactically
1026 invalid. However, $reply_address has leading white space removed, and all
1027 newlines turned into spaces so that the autoreply transport does not
1030 PH/05 If group was specified without a user on a router, and no group or user
1031 was specified on a transport, the group from the router was ignored.
1033 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1034 for visible compile-time settings that can be used to change these
1035 numbers, for those that want even more. Backwards compatibility with old
1036 spool files has been maintained. However, going back to a previous Exim
1037 release will lost any variables that are in spool files.
1039 PH/07 Two small changes when running in the test harness: increase delay when
1040 passing a TCP/IP connection to a new process, in case the original
1041 process has to generate a bounce, and remove special handling of
1042 127.0.0.2 (sic), which is no longer necessary.
1044 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1045 be the same on different OS.
1047 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1050 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1051 whether --show-vars was specified or not
1053 JJ/02 exipick: Added support for new ACL variable spool format introduced
1056 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1057 syntactically invalid From: or Reply-to: line, and a filter used this to
1058 generate an autoreply, and therefore failed to obtain an address for the
1059 autoreply, Exim could try to deliver to a non-existent relative file
1060 name, causing unrelated and misleading errors. What now happens is that
1061 it logs this as a hard delivery error, but does not attempt to create a
1064 PH/11 The exinext utility has a -C option for testing purposes, but although
1065 the given file was scanned by exinext itself; it wasn't being passed on
1066 when Exim was called.
1068 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1069 an end-of-file indication when reading a command response.
1071 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1072 compiled. In many other places in Exim, IPv6 addresses are always
1073 recognized, so I have changed this. It also means that IPv4 domain
1074 literals of the form [IPV4:n.n.n.n] are now always recognized.
1076 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1077 used if the router is not running as root, for example, when verifying at
1078 ACL time, or when using -bh. The debugging output from this situation was
1079 non-existent - all you got was a failure to exec. I have made two
1082 (a) Failures to set uid/gid, the current directory, or a process leader
1083 in a subprocess such as that created by queryprogram now generate
1084 suitable debugging ouput when -d is set.
1086 (b) The queryprogram router detects when it is not running as root,
1087 outputs suitable debugging information if -d is set, and then runs
1088 the subprocess without attempting to change uid/gid.
1090 PH/15 Minor change to Makefile for building test_host (undocumented testing
1093 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1094 additional section of a DNS packet that returns MX or SRV records.
1095 Instead, it always explicitly searches for A/AAAA records. This avoids
1096 major problems that occur when a DNS server includes only records of one
1097 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1098 fixed another bug: if SRV records were looked up and the corresponding
1099 address records were *not* found in the additional section, the port
1100 values from the SRV records were lost.
1102 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1103 using the correct key (the original address) when searching the retry
1104 rules in order to find which one to use for generating the retry hint.
1106 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1107 adding the default one. Similarly, if it contains a Reply-To: header, the
1108 errors_reply_to option, if set, is not used.
1110 PH/19 When calculating a retry time, Exim used to measure the "time since
1111 failure" by looking at the "first failed" field in the retry record. Now
1112 it does not use this if it is later than than the arrival time of the
1113 message. Instead it uses the arrival time. This makes for better
1114 behaviour in cases where some deliveries succeed, thus re-setting the
1115 "first failed" field. An example is a quota failure for a huge message
1116 when small messages continue to be delivered. Without this change, the
1117 "time since failure" will always be short, possible causing more frequent
1118 delivery attempts for the huge message than are intended.
1119 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1121 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1122 $1, $2, $3) because the numerical variables can be reset during some
1123 expansion items (e.g. "match"), thereby losing the authentication data.
1125 PH/21 Make -bV show the size of off_t variables so that the test suite can
1126 decide whether to run tests for quotas > 2G.
1128 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1129 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1130 filecount value is greater than 2G or if a quota value is greater than 2G
1131 on a system where the size of off_t is not greater than 4, a panic error
1134 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1135 never match. The debug and -bh output now contains an explicit error
1136 message indicating a malformed IPv4 address or mask.
1138 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1139 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1140 PH/23 above applies.
1142 PH/25 Do not write to syslog when running in the test harness. The only
1143 occasion when this arises is a failure to open the main or panic logs
1144 (for which there is an explicit test).
1146 PH/26 Added the /no_tell option to "control=freeze".
1148 PH/27 If a host name lookup failed very early in a connection, for example, if
1149 the IP address matched host_lookup and the reverse lookup yielded a name
1150 that did not have a forward lookup, an error message of the form "no IP
1151 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1152 could be logged. Now it outputs the IP address instead of "NULL".
1154 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1155 allows the sender and the authenticated sender to be set when
1156 submitting a message from within Exim. Since child_open_exim() is
1157 documented for local_scan(), the new function should be too.
1159 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1160 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1161 results in an empty string is now treated as unset.
1163 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1165 PH/31 Added sender_verify_fail logging option.
1167 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1168 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1169 tidied the source and removed it altogether.
1171 PH/33 When a queue run was abandoned because the load average was too high, a
1172 log line was always written; now it is written only if the queue_run log
1173 selector is set. In addition, the log line for abandonment now contains
1174 information about the queue run such as the pid. This is always present
1175 in "start" and "stop" lines but was omitted from the "abandon" line.
1177 PH/34 Omit spaces between a header name and the colon in the error message that
1178 is given when verify = headers_syntax fails (if there are lots of them,
1179 the message gets confusing).
1181 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1182 names, as there are now some PTR records that contain slashes. This check
1183 is only to protect against broken name servers that fall over on strange
1184 characters, so the fact that it applies to all lookups doesn't matter.
1186 PH/36 Now that the new test suite is complete, we can remove some of the
1187 special code in Exim that was needed for the old test suite. For example,
1188 sorting DNS records because real resolvers return them in an arbitrary
1189 order. The new test suite's fake resolver always returns records in the
1192 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1193 bounces) except when queue_only is set, to avoid logging races between
1194 the different processes.
1196 PH/38 Panic-die if .include specifies a non-absolute path.
1198 PH/39 A tweak to the "H" retry rule from its user.
1200 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1201 a label. They prevented compilation on older perls.
1203 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1204 a warning to be raised on newish perls.
1206 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1207 on queue. Changes to match documented behaviour of showing count of
1208 messages matching specified criteria.
1210 PH/40 Changed the default ident timeout from 30s to 5s.
1212 PH/41 Added support for the use of login_cap features, on those BSD systems
1213 that have them, for controlling the resources used by pipe deliveries.
1215 PH/42 The content-scanning code uses fopen() to create files in which to put
1216 message data. Previously it was not paying any attention to the mode of
1217 the files. Exim runs with umask(0) because the rest of the code creates
1218 files with open(), and sets the required mode explicitly. Thus, these
1219 files were ending up world-writeable. This was not a big issue, because,
1220 being within the spool directory, they were not world-accessible. I have
1221 created a function called modefopen, which takes an additional mode
1222 argument. It sets umask(777), creates the file, chmods it to the required
1223 mode, then resets the umask. All the relevant calls to fopen() in the
1224 content scanning code have been changed to use this function.
1226 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1227 to 24 hours. This avoids potential overflow problems when processing G
1228 and H retry rules. I suspect nobody ever tinkers with this value.
1230 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1232 PH/45 When the plaintext authenticator is running as a client, the server's
1233 challenges are checked to ensure they are valid base64 strings. By
1234 default, the authentication attempt is cancelled if an invalid string is
1235 received. Setting client_ignore_invalid_base64 true ignores these errors.
1236 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1237 they are received. Thus, the responses can be made to depend on the
1238 challenges. If an invalid string is ignored, an empty string is placed in
1241 PH/46 Messages that are created by the autoreply transport now contains a
1242 References: header, in accordance with RFCs 2822 and 3834.
1244 PH/47 Added authenticated_sender_force to the smtp transport.
1246 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1248 PH/49 Installed latest patch from the Sieve maintainer.
1250 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1251 was also set, the appendfile transport was unnecessarily scanning a
1252 directory of message files (e.g. for maildir delivery) to find the count
1253 of files (along with the size), even though it did not need this
1254 information. It now does the scan only if it needs to find either the
1255 size of the count of files.
1257 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1259 PH/52 Two bugs concerned with error handling when the smtp transport is
1262 (i) Exim was not creating retry information for temporary errors given
1263 for individual recipients after the DATA command when the smtp transport
1264 was used in LMTP mode. This meant that they could be retried too
1265 frequently, and not timed out correctly.
1267 (ii) Exim was setting the flag that allows error details to be returned
1268 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1269 recipients that were returned after the DATA command.
1271 PH/53 This is related to PH/52, but is more general: for any failing address,
1272 when detailed error information was permitted to be returned to the
1273 sender, but the error was temporary, then after the final timeout, only
1274 "retry timeout exceeded" was returned. Now it returns the full error as
1275 well as "retry timeout exceeded".
1277 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1278 do this, and (what is worse) MTAs that accept it.
1280 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1281 will now be deprecated.
1283 PH/56 New os.c-cygwin from the Cygwin maintainer.
1285 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1286 formats (previously only available in exim formats via -bpr, -bpru,
1287 and -bpra. Now also available in native and exiqgrep formats)
1289 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1290 with very large, slow to parse queues
1292 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1294 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1296 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1297 responses to authentication challenges, though it was showing the
1298 challenges; (ii) I've removed the CR characters from the debug output for
1301 PH/58 Allow for the insertion of a newline as well as a space when a string
1302 is turned into more than one encoded-word during RFC 2047 encoding. The
1303 Sieve code now uses this.
1305 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1306 data_4xx, lost_connection, tls_required.
1308 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1309 message was being sent as an SMTP response.
1311 PH/61 Add -l and -k options to exicyclog.
1313 PH/62 When verifying, if an address was redirected to one new address, so that
1314 verification continued, and the new address failed or deferred after
1315 having set something in $address_data, the value of $address_data was not
1316 passed back to the ACL. This was different to the case when no
1317 redirection occurred. The value is now passed back in both cases.
1319 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1320 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1321 use login_cap.h, so on its own it isn't the distinguishing feature. The
1322 new name refers directly to the setclassresources() function.
1324 PH/65 Added configuration files for NetBSD3.
1326 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1328 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1329 is preferred over IPv4.
1331 PH/68 The bounce_return_message and bounce_return_body options were not being
1332 honoured for bounces generated during the reception of non-SMTP messages.
1333 In particular, this applied to messages rejected by the ACL. This bug has
1334 been fixed. However, if bounce_return_message is true and bounce_return_
1335 body is false, the headers that are returned for a non-SMTP message
1336 include only those that have been read before the error was detected.
1337 (In the case of an ACL rejection, they have all been read.)
1339 PH/69 The HTML version of the specification is now built in a directory called
1340 spec_html instead of spec.html, because the latter looks like a path with
1341 a MIME-type, and this confuses some software.
1343 PH/70 Catch two compiler warnings in sieve.c.
1345 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1346 function verify_get_ident() calls ip_connect() to connect a socket, but
1347 if the "connect()" function timed out, ip_connect() used to close the
1348 socket. However, verify_get_ident() also closes the socket later, and in
1349 between Exim writes to the log, which may get opened at this point. When
1350 the socket was closed in ip_connect(), the log could get the same file
1351 descriptor number as the socket. This naturally causes chaos. The fix is
1352 not to close the socket in ip_connect(); the socket should be closed by
1353 the function that creates it. There was only one place in the code where
1354 this was missing, in the iplookup router, which I don't think anybody now
1355 uses, but I've fixed it anyway.
1357 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1358 well as to direct DNS lookups. Otherwise the handling of names in host
1359 lists is inconsistent and therefore confusing.
1365 PH/01 Two changes to the default runtime configuration:
1367 (1) Move the checks for relay_from_hosts and authenticated clients from
1368 after to before the (commented out) DNS black list checks.
1370 (2) Add control=submission to the relay_from_hosts and authenticated
1371 clients checks, on the grounds that messages accepted by these
1372 statements are most likely to be submissions.
1374 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1376 (1) Generate an error if the third argument for the ${prvs expansion is
1379 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1382 (3) Reset the variables that are obtained from the first argument of
1383 ${prvscheck and used in the second argument before leaving the code,
1384 because their memory is reclaimed, so using them afterwards may do
1387 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1388 one (it's much easier than Tom thought :-).
1390 (5) Because of (4), we can now allow for the use of $prvscheck_result
1391 inside the third argument.
1393 PH/03 For some reason, the default setting of PATH when running a command from
1394 a pipe transport was just "/usr/bin". I have changed it to
1397 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1398 anything to be listed in the output from -bV.
1400 PH/05 When a filter generated an autoreply, the entire To: header line was
1401 quoted in the delivery log line, like this:
1403 => >A.N.Other <ano@some.domain> <original@ddress> ...
1405 This has been changed so that it extracts the operative address. There
1406 may be more than one such address. If so, they are comma-separated, like
1409 => >ano@some.domain,ona@other.domain <original@ddress> ...
1411 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1412 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1413 not being looked up in the rDNS to get a host name, Exim was showing the
1414 IP address twice in Received: lines, even though the IP addresses were
1415 identical. For example:
1417 Received: from [1.2.3.4] (helo=[1.2.3.4])
1419 However, if the real host name was known, it was omitting the HELO data
1420 if it matched the actual IP address. This has been tidied up so that it
1421 doesn't show the same IP address twice.
1423 PH/07 When both +timestamp and +memory debugging was on, the value given by
1424 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1425 called by the debug printing, thereby overwriting the timestamp buffer.
1426 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1428 PH/08 When the original message was included in an autoreply transport, it
1429 always said "this is a copy of the message, including all the headers",
1430 even if body_only or headers_only was set. It now gives an appropriate
1433 PH/09 Applied a patch from the Sieve maintainer which:
1435 o fixes some comments
1436 o adds the (disabled) notify extension core
1437 o adds some debug output for the result of if/elsif tests
1438 o points to the current vacation draft in the documentation
1439 and documents the missing references header update
1443 o fixes a bug in processing the envelope test (when testing
1444 multiple envelope elements, the last element determinted the
1447 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1448 Electronic Mail") by including:
1450 Auto-submitted: auto-generated
1452 in the messages that it generates (bounce messages and others, such as
1453 warnings). In the case of bounce messages for non-SMTP mesages, there was
1454 also a typo: it was using "Auto_submitted" (underscore instead of
1455 hyphen). Since every message generated by Exim is necessarily in response
1456 to another message, thes have all been changed to:
1458 Auto-Submitted: auto-replied
1460 in accordance with these statements in the RFC:
1462 The auto-replied keyword:
1464 - SHOULD be used on messages sent in direct response to another
1465 message by an automatic process,
1467 - MUST NOT be used on manually-generated messages,
1469 - MAY be used on Delivery Status Notifications (DSNs) and Message
1470 Disposition Notifications (MDNs),
1472 - MUST NOT be used on messages generated by automatic or periodic
1473 processes, except for messages which are automatic responses to
1476 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1477 to the default Received: header definition.
1479 PH/12 Added log selector acl_warn_skipped (default on).
1481 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1482 variables because (a) they are in the wrong storage pool and (b) even if
1483 they were copied, it wouldn't work properly because of the caching.
1485 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1486 checking when decoding. Apparently there are clients that generate
1487 overlong encoded strings. Why am I not surprised?
1489 PH/15 If the first argument of "${if match_address" was not empty, but did not
1490 contain an "@" character, Exim crashed. Now it writes a panic log message
1491 and treats the condition as false.
1493 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1495 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1496 extension "envelope-auth". The code is finished and in agreement with
1497 other implementations, but there is no documentation so far and in fact,
1498 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1499 not changing the active code.
1501 Print executed "if" and "elsif" statements when debugging is used. This
1502 helps a great deal to understand what a filter does.
1504 Document more things not specified clearly in RFC3028. I had all this
1505 sorted out, when out of a sudden new issues came to my mind. Oops."
1507 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1510 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1511 canonical form (as documented). However, after a host name lookup from
1512 the IP address, check_host() was doing a simple string comparison with
1513 addresses acquired from the DNS when checking that the found name did
1514 have the original IP as one of its addresses. Since any found IPv6
1515 addresses are likely to be in abbreviated form, the comparison could
1516 fail. Luckily, there already exists a function for doing the comparison
1517 by converting both addresses to binary, so now that is used instead of
1518 the text comparison.
1520 PH/20 There was another similar case to PH/19, when a complete host name was
1521 given in a host list; looking up its IP address could give an abbreviated
1522 form, whereas the current host's name might or might not be abbreviated.
1523 The same fix has been applied.
1529 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1530 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1533 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1534 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1536 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1538 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1539 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1540 Also, for "Auto-Submitted", treat anything other than "no" as
1541 non-personal, in accordance with RFC 3834. (Previously it treated
1542 anything starting "auto-" as non-personal.)
1544 TF/01 The control=submission/name=... option had a problem with syntax
1545 errors if the name included a slash character. The /name= option
1546 now slurps the rest of the string, so it can include any characters
1547 but it must come last in the list of options (after /sender_retain
1550 PH/05 Some modifications to the interface to the fake nameserver for the new
1558 TK/01 Added the "success_on_redirect" address verification option. See
1559 NewStuff for rationale and an example.
1561 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1563 PH/02 Patch to exigrep to allow it to work on syslog lines.
1565 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1566 fread() to skip over the body file's header line, because in Cygwin the
1567 header line is locked and is inaccessible.
1569 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1570 co-exist for some time) to make it clear that it is the Exim ID that is
1571 referenced, not the Message-ID: header line.
1573 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1574 string_format() function, because snprintf() does not exist on all
1577 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1578 ${stat: expansion item.
1580 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1581 protocol synchronization error", to keep the pedants happy.
1583 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1584 well as for IRIX systems, when gcc is being used. See the host.c source
1587 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1589 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1592 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1593 transport and to the smtp transport in LMTP mode.
1595 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1597 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1598 run a filter in a subprocess. This could lead to confusion in subsequent
1599 lookups in the parent process. There should also be a search_tidyup() at
1600 the end of the subprocess.
1602 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1603 only if the host matched helo_try_verify_hosts, which caused the
1604 verification to occur when the EHLO/HELO command was issued. The ACL just
1605 tested the remembered result. Now, if a previous verification attempt has
1606 not happened, "verify = helo" does it there and then.
1608 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1610 TK/03 Fix log output including CR from clamd.
1612 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1613 exist provoked a memory error which could cause a segfault.
1615 PH/15 Installed PCRE 6.2
1617 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1619 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1620 of the problem. Specifically, suggested +O2 rather than +O1 for the
1623 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1625 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1626 record, the port was not correctly logged when the outgoing_port log
1627 selector was set (it logged the transort's default port).
1629 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1630 fallback_hosts, and "hosts" in the smtp transport.
1632 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1633 host errors such as "Connection refused".
1635 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1636 authentication with radiusclient 0.4.9:
1638 - Error returned from rc_read_config was caught wrongly
1639 - Username/password not passed on to radius server due to wrong length.
1641 The presumption is that some radiusclient API changes for 4.51/PH/17
1642 were not taken care of correctly. The code is still untested by me (my
1643 Linux distribution still has 0.3.2 of radiusclient), but it was
1644 contributed by a Radius user.
1646 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1647 expanding the "port" option of the smtp transport.
1649 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1650 while reading a MIME header. Thanks to Tom Hughes for a patch.
1652 PH/24 Include config.h inside local_scan.h so that configuration settings are
1655 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1656 that in an ACL for RCPT (for example), you can examine exactly what was
1659 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1660 commands, but it was not correctly comparing the address with the actual
1661 client host address. Thus, it would show the EHLO address in Received:
1662 header lines when this was not necessary.
1664 PH/27 Added the % operator to ${eval:}.
1666 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1667 it should be ignoring failures (because with -C, for example, it has lost
1668 privilege). It wasn't ignoring creation failures other than "already
1671 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1674 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1675 because an input line was too long, either on its own, or by virtue of
1676 too many continuations, the temporary file was not being removed, and the
1677 return code was incorrect.
1679 PH/31 Missing "BOOL" in function definition in filtertest.c.
1681 PH/32 Applied Sieve patches from the maintainer.
1683 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1685 PH/33 Added "verify = not_blind".
1687 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1688 Local/Makefile (with some defaults set). These are used in built scripts
1689 such as exicyclog, but they have never been used in the exim_install
1690 script (though there are many overriding facilities there). I have
1691 arranged that the exim_install script now takes note of these two
1694 PH/35 Installed configuration files for Dragonfly.
1696 PH/36 When a locally submitted message by a trusted user did not contain a
1697 From: header, and the sender address was obtained from -f or from an SMTP
1698 MAIL command, and the trusted user did not use -F to supply a sender
1699 name, $originator_name was incorrectly used when constructing a From:
1700 header. Furthermore, $originator_name was used for submission mode
1701 messages from external hosts without From: headers in a similar way,
1702 which is clearly wrong.
1704 PH/37 Added control=suppress_local_fixups.
1706 PH/38 When log_selector = +received_sender was set, and the addition of the
1707 sender made the log line's construction buffer exactly full, or one byte
1708 less than full, an overflow happened when the terminating "\n" was
1711 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1712 when the result of a list match is failure because a DNS lookup failed.
1714 PH/40 RM_COMMAND is now used in the building process.
1716 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1717 the "build-* directories that it finds.
1719 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1720 address in a domain literal was a prefix of an interface address.
1722 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1723 when verifying a sender address, unless rewrite_headers is false.
1725 PH/44 Wrote a long comment about why errors_to addresses are verified as
1726 recipients, not senders.
1728 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1729 the ratelimit ACL was added.
1731 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1733 PH/46 Added extra information about PostgreSQL errors to the error string.
1735 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1736 avoiding the need to install special zones in a real server. This is
1737 backwards compatible; if it can't find the fake resolver, it drops back.
1738 Thus, both old and new test suites can be run.
1740 TF/02 Added util/ratelimit.pl
1742 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1743 clock is set back in time.
1745 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1746 Candler <B.Candler@pobox.com>.
1748 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1749 OK for addresses that are the result of redirections.
1751 PH/48 A number of further additions for the benefit of the new test suite,
1752 including a fake gethostbyname() that interfaces to the fake DNS resolver
1755 TF/06 The fix for widen_domains has also been applied to qualify_single and
1756 search_parents which are the other dnslookup options that can cause
1759 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1762 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1764 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1765 service" if it encountered an MX record with an empty target hostname.
1766 The message is now "an MX or SRV record indicated no SMTP service".
1768 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1769 if the DNS of the sending site is misconfigured. This is quite a
1770 common situation. This change restores the behaviour of treating a
1771 helo verification defer as a failure.
1773 PH/51 If self=fail was set on a router, the bounce message did not include the
1774 actual error message.
1780 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1782 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1783 command itself ended in error, the underlying message about the transport
1784 filter timeout was being overwritten with the pipe command error. Now the
1785 underlying error message should be appended to the second error message.
1787 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1789 PH/02 Reduce the amount of output that "make" produces by default. Full output
1790 can still be requested.
1792 PH/03 The warning log line about a condition test deferring for a "warn" verb
1793 was being output only once per connection, rather than after each
1794 occurrence (because it was using the same function as for successful
1795 "warn" verbs). This seems wrong, so I have changed it.
1797 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1798 it should not have, which might have caused a crash in the right
1799 circumstances, but probably never did.
1801 PH/04 Installed a modified version of Tony Finch's patch to make submission
1802 mode fix the return path as well as the Sender: header line, and to
1803 add a /name= option so that you can make the user's friendly name appear
1806 TF/03 Added the control = fakedefer ACL modifier.
1808 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1809 Mark Lowes for thorough testing.
1811 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1813 TK/03 Merged latest SRS patch from Miles Wilton.
1815 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1816 with the definition in sysexits.h (which is #included earlier).
1817 Fortunately, Exim does not actually use EX_OK. The code used to try to
1818 preserve the sysexits.h value, by assumimg that macro definitions were
1819 scanned for macro replacements. I have been disabused of this notion,
1820 so now the code just undefines EX_OK before #including unistd.h.
1822 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1823 in the smtp transport. When a block could not be written in a single
1824 write() function, the timeout was being re-applied to each part-write.
1825 This seems wrong - if the receiver was accepting one byte at a time it
1826 would take for ever. The timeout is now adjusted when this happens. It
1827 doesn't have to be particularly precise.
1829 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1830 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1832 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1833 without a domain if the check to <postmaster@domain> fails.
1835 SC/01 Eximstats: added -xls and the ability to specify output files
1836 (patch written by Frank Heydlauf).
1838 SC/02 Eximstats: use FileHandles for outputing results.
1840 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1842 SC/04 Eximstats: fixed display of large numbers with -nvr option
1844 SC/05 Eximstats: fixed merging of reports with empty tables.
1846 SC/06 Eximstats: added the -include_original_destination flag
1848 SC/07 Eximstats: removed tabs and trailing whitespace.
1850 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1852 TK/06 MBOX spool code: Add real "From " MBOX separator line
1853 so the .eml file is really in mbox format (even though
1854 most programs do not really care). Patch from Alex Miller.
1856 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1857 The latter is generated from $received_to and is only set if the
1858 message has one envelope recipient. SA can use these headers,
1859 obviously out-of-the-box. Patch from Alex Miller.
1861 PH/08 The ${def test on a variable was returning false if the variable's
1862 value was "0", contrary to what the specification has always said!
1863 The result should be true unless the variable is empty.
1865 PH/09 The syntax error of a character other than { following "${if
1866 def:variable_name" (after optional whitespace) was not being diagnosed.
1867 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1868 accidental colon was present, for example, could give incorrect results.
1870 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1871 result is used (not including appendfile, where other changes are about
1874 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1875 This involved changing a lot of size variables from int to off_t. It
1876 should work with maildirs and everything.
1878 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1879 spamd dying while we are connected to it.
1881 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1884 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1885 function for BDB 4.3.
1887 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1889 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1890 the sources for PCRE have been re-arranged and more files are now
1893 PH/15 The code I had for printing potentially long long variables in PH/11
1894 above was not the best (it lost precision). The length of off_t variables
1895 is now inspected at build time, and an appropriate printing format (%ld
1896 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1897 to be "long long int" or "long int". This is needed for the internal
1898 formatting function string_vformat().
1900 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1901 the configuration file to be ":syslog", then the script "guesses" where
1902 the logs files are, rather than using the compiled in default. In our
1903 case the guess is not the same as the compiled default, so the script
1904 suddenly stopped working when I started to use syslog. The patch checks
1905 to see if log_file_path is "". If so, it attempts to read it from exim
1906 with no configuration file to get the compiled in version, before it
1907 falls back to the previous guessing code."
1909 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1910 implementing BATV in an Exim configuration. See NewStuff for the gory
1913 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1914 Makefile that are specific to HP-UX.
1916 PH/18 If the "use_postmaster" option was set for a recipient callout together
1917 with the "random" option, the postmaster address was used as the MAIL
1918 FROM address for the random test, but not for the subsequent recipient
1919 test. It is now used for both.
1921 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1922 patch removes a few documentation additions to RFC 3028, because the
1923 latest draft now contains them. It adds the new en;ascii-case comparator
1924 and a new error check for 8bit text in MIME parts. Comparator and
1925 require names are now matched exactly. I enabled the subaddress
1926 extension, but it is not well tested yet (read: it works for me)."
1928 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1929 rework some of the code of TK/09 above to avoid the hardwired use of
1930 "%lld" and "long long". Replaced the call to snprintf() with a call to
1933 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1934 records point to non-existent hosts", "retry timeout exceeded", and
1935 "retry time not reached for any host after a long failure period".
1937 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1938 experimental DomainKeys support:
1940 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1941 (2) On an error such as an illegally used "control", the wrong name for
1942 the control was given.
1944 These problems did NOT occur unless DomainKeys support was compiled.
1946 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1948 PH/24 Added ${if match_ip condition.
1950 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1951 databases so that it will be absolutely obvious if a crash occurs in the
1952 DB library. This is a regular occurrence (often caused by mis-matched
1955 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1956 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1957 picked up on a user's system that detects such things. There doesn't seem
1958 to be a gcc warning option for this - only an attribute that has to be
1959 put on the function's prototype. It seems that in Fedora Core 4 they have
1960 set this on a number of new functions. No doubt there will be more in due
1963 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1964 specify a transport. However, if an address that was verified by such a
1965 router was the subject of a callout, Exim crashed because it tried to
1966 read the rcpt_include_affixes from the non-existent transport. Now it
1967 just assumes that the setting of that option is false. This bug was
1968 introduced by 4.51/PH/31.
1970 PH/28 Changed -d+all to exclude +memory, because that information is very
1971 rarely of interest, but it makes the output a lot bigger. People tend to
1972 do -d+all out of habit.
1974 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1975 code in os-type was giving problems when libc.so lives in lib64, like on
1978 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1979 aren't the modern standard, and it seems that some systems' include files
1980 don't always have them. Exim was already checking for some of the newer
1981 ones like T_AAAA, and defining it itself. I've added checks for all the
1982 record types that Exim uses.
1984 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1985 not automatically generating a new one, as it is supposed to. This
1986 prevented TLS from working. If the file did exist, but contained invalid
1987 data, a new version was generated, as expected. It was only the case of a
1988 non-existent file that was broken.
1990 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1991 with a change in libdomainkeys > 0.64.
1993 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1994 from DNS. If the selector record carries the flag, it now has
1995 precedence over the domain-wide flag.
1997 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1999 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2000 the use of a port name that isn't defined in /etc/services) occurred, the
2001 message was deferred as in a normal delivery, and thus remained on the
2002 spool, instead of being failed because of the mua_wrapper setting. This
2003 is now fixed, and I tidied up some of the mua_wrapper messages at the
2006 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2007 the messages in a hash of arrays rather than using individual hashes.
2008 This is a bit cleaner and results in dramatic memory savings, albeit
2009 at a slight CPU cost.
2011 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2012 as requested by Marc Sherman.
2014 SC/10 Eximstats: added histograms for user specified patterns as requested
2017 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2019 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2020 fopen() in the content-scanning modules that did not already have it.
2026 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2027 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2029 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2031 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2033 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2034 to test Sieve filters that use "vacation".
2036 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2037 that changes the way the GnuTLS parameters are stored in the cache file.
2038 The new format can be generated externally. For backward compatibility,
2039 if the data in the cache doesn't make sense, Exim assumes it has read an
2040 old-format file, and it generates new data and writes a new file. This
2041 means that you can't go back to an older release without removing the
2044 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2045 work if there are any delivery delays because "one_time" forces the
2046 parent to be marked "delivered", so its unseen clone is never tried
2047 again. For this reason, Exim now forbids the simultaneous setting of
2050 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2051 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2052 entry if you want to know the details. The fix, however, neglected to
2053 consider the case when local delivery batching is involved. The test for
2054 "previously delivered" was not happening when checking to see if an
2055 address could be batched with a previous (undelivered) one; under
2056 certain circumstances this could lead to multiple deliveries to the same
2059 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2060 in its include files, and this causes problems building Exim.
2062 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2063 header_syntax) but Exim was just ignoring anything given after a slash.
2064 In particular, this caused confusion with an attempt to use "verify =
2065 reverse_host_lookup/defer_ok". An error is now given when options are
2066 supplied for verify items that do not have them. (Maybe reverse_host_
2067 lookup should have a defer_ok option, but that's a different point.)
2069 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2070 defined by RFC 821) to 2048, because there were problems with some AUTH
2071 commands, and RFC 1869 says the size should be increased for extended
2072 SMTP commands that take arguments.
2074 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2077 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2078 "unknown" error; now it says that the functionality isn't in the binary.
2080 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2081 an address' error message when a string expansion fails (syntax or
2082 whatever). Otherwise the password may appear in the log. Following change
2083 PH/42 below, there is no longer a chance of it appearing in a bounce
2086 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2088 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2089 of its addresses was ever tried. (Bugzilla bug #2.)
2091 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2092 the result incorrectly in the debug output. (It correctly added a newline
2093 to what was transported.)
2095 TF/01 Added $received_time.
2097 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2098 commented out examples of how to interface to a virus scanner and to
2099 SpamAssassin. Also added commented examples of av_scanner and
2100 spamd_address settings.
2102 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2103 and controls are allowed in which ACLs. There were a couple of minor
2104 errors. Some of the entries in the conditions table (which is a table of
2105 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2106 negation of where the condition IS allowed.
2108 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2110 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2111 header file does not have a version number, so I've had to invent a new
2112 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2113 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2114 radiusclient), but it was contributed by a Radius user.
2116 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2117 files or directories, for OpenSSL.
2119 PH/19 When an Exim process that is running as root has to create an Exim log
2120 file, it does so in a subprocess that runs as exim:exim so as to get the
2121 ownership right at creation (otherwise, other Exim processes might see
2122 the file with the wrong ownership). There was no test for failure of this
2123 fork() call, which would lead to the process getting stuck as it waited
2124 for a non-existent subprocess. Forks do occasionally fail when resources
2125 run out. I reviewed all the other calls to fork(); they all seem to check
2128 PH/20 When checking for unexpected SMTP input at connect time (before writing
2129 the banner), Exim was not dealing correctly with a non-positive return
2130 from the read() function. If the client had disconnected by this time,
2131 the result was a log entry for a synchronization error with an empty
2132 string after "input=" when read() returned zero. If read() returned -1
2133 (an event I could not check), uninitialized data bytes were printed.
2134 There were reports of junk text (parts of files, etc) appearing after
2137 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2139 PH/22 Added support for macro redefinition, and (re)definition in between
2140 driver and ACL definitions.
2142 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2143 forgetting to use the resulting value; it was using the unexpanded value.
2145 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2146 hadn't been configured. The fix is from Juergen Kreileder, who
2147 understands it better than I do:
2149 "Here's what I see happening with three configured cyrus_sasl
2150 authenticators configured (plain, login, cram-md5):
2152 On startup auth_cyrus_sasl_init() gets called for each of these.
2153 This means three calls to sasl_listmech() without a specified mech_list.
2154 => SASL tests which mechs of all available mechs actually work
2155 => three warnings about OTP not working
2156 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2158 With the patch, sasl_listmech() also gets called three times. But now
2159 SASL's mech_list option is set to the server_mech specified in the the
2160 authenticator. Or in other words, the answer from sasl_listmech()
2161 gets limited to just the mech you're testing for (which is different
2163 => the return list contains just 'plain' or 'login', 'cram-md5' or
2164 nothing depending on the value of ob->server_mech.
2166 I've just tested the patch: Authentication still works fine,
2167 unavailable mechs specified in the exim configuration are still
2168 caught, and the auth.log warnings about OTP are gone."
2170 PH/25 When debugging is enabled, the contents of the command line are added
2171 to the debugging output, even when log_selector=+arguments is not
2174 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2175 answer is "GNU", and only if the return is "GNU/something" is the answer
2178 PH/27 $acl_verify_message is now set immediately after the failure of a
2179 verification in an ACL, and so is available in subsequent modifiers. In
2180 particular, the message can be preserved by coding like this:
2182 warn !verify = sender
2183 set acl_m0 = $acl_verify_message
2185 Previously, $acl_verify_message was set only while expanding "message"
2186 and "log_message" when a very denied access.
2188 PH/28 Modified OS/os.c-Linux with
2190 -#ifndef OS_LOAD_AVERAGE
2191 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2193 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2194 nomenclature these days.)
2196 PH/29 Installed patch from the Sieve maintainer that adds the options
2197 sieve_useraddress and sieve_subaddress to the redirect router.
2199 PH/30 In these circumstances:
2200 . Two addresses routed to the same list of hosts;
2201 . First host does not offer TLS;
2202 . First host accepts first address;
2203 . First host gives temporary error to second address;
2204 . Second host offers TLS and a TLS session is established;
2205 . Second host accepts second address.
2206 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2207 and peerdn, if requested) that were in fact used only for the second
2210 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2211 attention to any local part prefix or suffix that was matched by the
2212 router that accepted the address. It now behaves in the same way as it
2213 does for delivery: the affixes are removed from the local part unless
2214 rcpt_include_affixes is set on the transport.
2216 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2217 timeout during the DATA phase of an incoming message.
2219 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2220 applied a patch sanctioned by the Sieve maintainer.
2222 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2223 the uid or gid is negative. A case of a negative gid caused this to be
2224 noticed. The fix allows for either to be negative.
2226 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2227 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2230 PH/36 The change PH/12 above was broken. Fixed it.
2232 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2233 the grounds that routing the same address twice would always produce the
2234 same answer. This might have been true once, but it is certainly no
2235 longer true now. Routing a child address may depend on the previous
2236 routing that produced that child. Some complicated redirection strategies
2237 went wrong when messages had multiple recipients, and made Exim's
2238 behaviour dependent on the order in which the addresses were given.
2240 I have moved the duplicate checking until after the routing is complete.
2241 Exim scans the addresses that are assigned to local and remote
2242 transports, and removes any duplicates. This means that more work will be
2243 done, as duplicates will always all be routed, but duplicates are
2244 presumably rare, so I don't expect this is of any significance.
2246 For deliveries to pipes, files, and autoreplies, the duplicate checking
2247 still happens during the routing process, since they are not going to be
2250 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2251 It corrects a timeout issue with spamd. This is Ian's comment: "The
2252 background is that sometimes spamd either never reads data from a
2253 connection it has accepted, or it never writes response data. The exiscan
2254 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2255 blindly assumes that writes won't block so it may never time out."
2257 PH/39 Allow G after quota size as well as K and M.
2259 PH/40 The value set for $authenticated_id in an authenticator may not contain
2260 binary zeroes or newlines because the value is written to log lines and
2261 to spool files. There was no check on this. Now the value is run through
2262 the string_printing() function so that such characters are converted to
2263 printable escape sequences.
2265 PH/41 $message_linecount is a new variable that contains the total number of
2266 lines in the message. Compare $body_linecount, which is the count for the
2269 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2270 bounce and delay warning messages, except in certain special cases, which
2273 (a) An SMTP error message from a remote host;
2274 (b) A message specified in a :fail: redirection;
2275 (c) A message specified in a "fail" command in a system filter;
2276 (d) A message specified in a FAIL return from the queryprogram router;
2277 (e) A message specified by the cannot_route_message router option.
2279 In these cases only, Exim does include the error details in bounce and
2280 warning messages. There are also a few cases where bland messages such
2281 as "unrouteable address" or "local delivery error" are given.
2283 PH/43 $value is now also set for the "else" part of a ${run expansion.
2285 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2286 being worked on, but at least Exim now implements the latest version to
2289 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2290 process to complete was treated as a delivery failure, a timeout while
2291 writing the message to the pipe was logged, but erroneously treated as a
2292 successful delivery. Such timeouts include transport filter timeouts. For
2293 consistency with the overall process timeout, these timeouts are now
2294 treated as errors, giving rise to delivery failures by default. However,
2295 there is now a new Boolean option for the pipe transport called
2296 timeout_defer, which, if set TRUE, converts the failures into defers for
2297 both kinds of timeout. A transport filter timeout is now identified in
2300 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2301 systems where "make" and "gmake" are different, calling "gmake" at top
2302 level broke things. I've arranged for the value of $(MAKE) to be passed
2303 from the Makefile to this script so that it can call the same version of
2307 A note about Exim versions 4.44 and 4.50
2308 ----------------------------------------
2310 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2311 changes of various kinds. As a consequence, a big documentation update was
2312 needed. This delayed the release for rather longer than seemed good, especially
2313 in the light of a couple of (minor) security issues. Therefore, the changes
2314 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2315 release. So 4.44 and 4.50 are in effect two different branches that both start
2318 I have left the 4.50 change log unchanged; it contains all the changes since
2319 4.43. The change log for 4.44 is below; many of its items are identical to
2320 those for 4.50. This seems to be the most sensible way to preserve the
2321 historical information.
2327 1. Minor wording change to the doc/README.SIEVE file.
2329 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2330 computation of the current number of files was incorrect.
2332 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2333 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2336 4. Give more explanation in the error message when the command for a transport
2337 filter fails to execute.
2339 5. There are several places where Exim runs a non-Exim command in a
2340 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2341 was being done only for the command run by the queryprogram router. It is
2342 now done for all such subprocesses. The other cases are: ${run, transport
2343 filters, and the commands run by the lmtp and pipe transports.
2345 6. Added CONFIGURE_GROUP build-time option.
2347 7. Some older OS have a limit of 256 on the maximum number of file
2348 descriptors. Exim was using setrlimit() to set 1000 as a large value
2349 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2350 systems. I've change it so that if it can't get 1000, it tries for 256.
2352 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2353 was an oversight, and furthermore, ever since the addition of extra
2354 controls (e.g. 4.43/32), the checks on when to allow different forms of
2355 "control" were broken. There should now be diagnostics for all cases when a
2356 control that does not make sense is encountered.
2358 9. Added the /retain_sender option to "control=submission".
2360 10. $recipients is now available in the predata ACL (oversight).
2362 11. Tidy the search cache before the fork to do a delivery from a message
2363 received from the command line. Otherwise the child will trigger a lookup
2364 failure and thereby defer the delivery if it tries to use (for example) a
2365 cached ldap connection that the parent has called unbind on.
2367 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2368 of $address_data from the recipient verification was clobbered by the
2369 sender verification.
2371 13. The value of address_data from a sender verification is now available in
2372 $sender_address_data in subsequent conditions in the ACL statement.
2374 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2376 15. Added a new option "connect=<time>" to callout options, to set a different
2379 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2380 was its contents. (It was OK if the option was not defined at all.)
2382 17. A "Completed" log line is now written for messages that are removed from
2383 the spool by the -Mrm option.
2385 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2386 information about exactly what failed.
2388 19. Added -dd to debug only the daemon process.
2390 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2391 handles timeouts, both on the server side and network timeouts. Renamed the
2392 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2394 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2395 It is now set to "smtps".
2397 22. $host_address is now set to the target address during the checking of
2398 ignore_target_hosts.
2400 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2401 being passed; this would have caused $sender_host_name to have been used if
2402 matching the list had actually called for a host name (not very likely,
2403 since this list is usually IP addresses). A host name is now passed as
2406 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2407 code that specifies a non-restarting handler (typically sigaction() in
2408 modern systems) in an attempt to fix a rare and obscure crash bug.
2410 25. Narrowed the window for a race in the daemon that could cause it to ignore
2411 SIGCHLD signals. This is not a major problem, because they are used only to
2412 wake it up if nothing else does.
2414 26. A malformed maildirsize file could cause Exim to calculate negative values
2415 for the mailbox size or file count. Odd effects could occur as a result.
2416 The maildirsize information is now recalculated if the size or filecount
2419 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2420 support for a long time. Removed HAVE_SYS_VFS_H.
2422 28. Installed the latest version of exipick from John Jetmore.
2424 29. In an address list, if the pattern was not a regular expression, an empty
2425 subject address (from a bounce message) matched only if the pattern was an
2426 empty string. Non-empty patterns were not even tested. This was the wrong
2427 because it is perfectly reasonable to use an empty address as part of a
2428 database query. An empty address is now tested by patterns that are
2429 lookups. However, all the other forms of pattern expect the subject to
2430 contain a local part and a domain, and therefore, for them, an empty
2431 address still always fails if the pattern is not itself empty.
2433 30. Exim went into a mad DNS loop when attempting to do a callout where the
2434 host was specified on an smtp transport, and looking it up yielded more
2435 than one IP address.
2437 31. Re-factored the code for checking spool and log partition space into a
2438 function that finds that data and another that does the check. The former
2439 is then used to implement four new variables: $spool_space, $log_space,
2440 $spool_inodes, and $log_inodes.
2442 32. The RFC2047 encoding function was originally intended for short strings
2443 such as real names; it was not keeping to the 75-character limit for
2444 encoded words that the RFC imposes. It now respects the limit, and
2445 generates multiple encoded words if necessary. To be on the safe side, I
2446 have increased the buffer size for the ${rfc2047: expansion operator from
2449 33. It is now permitted to omit both strings after an "if" condition; if the
2450 condition is true, the result is "true". As before, when the second string
2451 is omitted, a false condition yields an empty string. This makes it less
2452 cumbersome to write custom ACL and router conditions.
2454 34. Failure to deliver a bounce message always caused it to be frozen, even if
2455 there was an errors_to setting on the router. The errors_to setting is now
2458 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2459 canonical form (fully expanded) before being placed in
2460 $sender_host_address.
2462 36. The table in the code that translates DNS record types into text (T_A to
2463 "A" for instance) was missing entries for NS and CNAME. It is just possible
2464 that this could have caused confusion if both these types were looked up
2465 for the same domain, because the text type is used as part of Exim's
2466 per-process caching. But the chance of anyone hitting this buglet seems
2469 37. The dnsdb lookup has been extended in a number of ways.
2471 (1) There is a new type, "zns", which walks up the domain tree until it
2472 finds some nameserver records. It should be used with care.
2474 (2) There is a new type, "mxh", which is like "mx" except that it returns
2475 just the host names, not the priorities.
2477 (3) It is now possible to give a list of domains (or IP addresses) to be
2478 looked up. The behaviour when one of the lookups defers can be
2479 controlled by a keyword.
2481 (4) It is now possible to specify the separator character for use when
2482 multiple records are returned.
2484 38. The dnslists ACL condition has been extended: it is now possible to supply
2485 a list of IP addresses and/or domains to be looked up in a particular DNS
2488 39. Added log_selector=+queue_time_overall.
2490 40. When running the queue in the test harness, wait just a tad after forking a
2491 delivery process, to get repeatability of debugging output.
2493 41. Include certificate and key file names in error message when GnuTLS fails
2494 to set them up, because the GnuTLS error message doesn't include the name
2495 of the failing file when there is a problem reading it.
2497 42. Allow both -bf and -bF in the same test run.
2499 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2501 44. The "Exiscan patch" is now merged into the mainline Exim source.
2503 45. Sometimes the final signoff response after QUIT could fail to get
2504 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2505 < 0 before doing a fflush(). This bug looks as though it goes back to the
2506 introduction of TLS in release 3.20, but "sometimes" must have been rare
2507 because the tests only now provoked it.
2509 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2510 (this can affect the format of dates).
2512 47. exim_tidydb, when checking for the continued existence of a message for
2513 which it has found a message-specific retry record, was not finding
2514 messages that were in split spool directories. Consequently, it was
2515 deleting retry records that should have stayed in existence.
2517 48. Steve fixed some bugs in eximstats.
2519 49. The SPA authentication driver was not abandoning authentication and moving
2520 on to the next authenticator when an expansion was forced to fail,
2521 contradicting the general specification for all authenticators. Instead it
2522 was generating a temporary error. It now behaves as specified.
2524 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2525 (the order specifies the preference for clients). The order is now AES256,
2526 AES128, 3DES, ARCFOUR128.
2528 51. Small patch to Sieve code - explicitly set From: when generating an
2531 52. Exim crashed if a remote delivery caused a very long error message to be
2532 recorded - for instance if somebody sent an entire SpamAssassin report back
2533 as a large number of 550 error lines. This bug was coincidentally fixed by
2534 increasing the size of one of Exim's internal buffers (big_buffer) that
2535 happened as part of the Exiscan merge. However, to be on the safe side, I
2536 have made the code more robust (and fixed the comments that describe what
2539 53. Now that there can be additional text after "Completed" in log lines (if
2540 the queue_time_overall log selector is set), a one-byte patch to exigrep
2541 was needed to allow it to recognize "Completed" as not the last thing in
2544 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2545 patch that reportedly fixes this has been added. I am not expert enough to
2546 create a test for it. This is what the patch creator wrote:
2548 "I found a little strange behaviour of ldap code when working with
2549 Windows 2003 AD Domain, where users was placed in more than one
2550 Organization Units. When I tried to give exim partial DN, the exit code
2551 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2552 But simultaneously result of request was absolutely normal ldap result,
2553 so I produce this patch..."
2555 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2556 REFERENCE, so I have modified the code to exclude the patch when that macro
2559 55. Some experimental protocols are using DNS PTR records for new purposes. The
2560 keys for these records are domain names, not reversed IP addresses. The
2561 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2562 leaves it alone. Component reversal etc. now happens only for IP addresses.
2565 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2567 57. Double the size of the debug message buffer (to 2048) so that more of very
2568 long debug lines gets shown.
2570 58. The exicyclog utility now does better if the number of log files to keep
2571 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2573 59. Two changes related to the smtp_active_hostname option:
2575 (1) $smtp_active_hostname is now available as a variable.
2576 (2) The default for smtp_banner uses $smtp_active_hostname instead
2577 of $primary_hostname.
2579 60. The host_aton() function is supposed to be passed a string that is known
2580 to be a valid IP address. However, in the case of IPv6 addresses, it was
2581 not checking this. This is a hostage to fortune. Exim now panics and dies
2582 if the condition is not met. A case was found where this could be provoked
2583 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2584 components; fortuitously, this particular loophole had already been fixed
2585 by change 4.50/55 above.
2587 If there are any other similar loopholes, the new check in host_aton()
2588 itself should stop them being exploited. The report I received stated that
2589 data on the command line could provoke the exploit when Exim was running as
2590 exim, but did not say which command line option was involved. All I could
2591 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2592 running as the user.
2595 61. There was a buffer overflow vulnerability in the SPA authentication code
2596 (which came originally from the Samba project). I have added a test to the
2597 spa_base64_to_bits() function which I hope fixes it.
2600 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2601 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2603 63. The daemon start-up calls getloadavg() while still root for those OS that
2604 need the first call to be done as root, but it missed one case: when
2605 deliver_queue_load_max is set with deliver_drop_privilege. This is
2606 necessary for the benefit of the queue runner, because there is no re-exec
2607 when deliver_drop_privilege is set.
2609 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2610 This has been fixed.
2612 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2613 was in use, was not putting the data itself into the right store pool;
2614 consequently, it could be overwritten for a subsequent message in the same
2615 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2618 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2620 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2621 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2622 the function were treating the return as a boolean value, which happened to
2623 work because 0=false and not-0=true, but is not correct code.
2625 68. The host_aton() function was not handling scoped IPv6 addresses (those
2626 with, for example, "%eth0" on the end) correctly.
2628 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2629 negated items (that is, ~something) in unsigned ints. Some compilers
2630 apparently mutter when there is no cast.
2632 70. If an address verification called from an ACL failed, and did not produce a
2633 user-specific message (i.e. there was only a "system" message), nothing was
2634 put in $acl_verify_message. In this situation, it now puts the system
2637 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2638 session; change 4.31/43 added the unwanted input to the log line - except
2639 that it did not do this in the start of session case. It now does.
2641 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2642 This is wrong and can cause the other end to generate a synchronization
2643 error if it is another Exim or anything else that does the synchronization
2644 check. A QUIT command is no longer sent after a timeout.
2646 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2647 during host lookups.
2649 74. The defer_ok option of callout verification was not working if it was used
2650 when verifying addresses in header lines, that is, for this case:
2652 verify = header_sender/callout=defer_ok
2654 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2655 those file descriptors could be used for SMTP connections. If anything
2656 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2657 could be sent to the SMTP client, causing chaos. The daemon now opens
2658 stdin, stdout, and stderr to /dev/null when it puts itself into the
2661 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2662 log by default. The user can override this with suitable Perl magic.
2664 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2665 the log message when discard triggers, was not working for the DATA ACL or
2666 for the non-SMTP ACL.
2668 78. Error message wording change in sieve.c.
2670 79. If smtp_accept_max_per_host was set, the number of connections could be
2671 restricted to fewer than expected, because the daemon was trying to set up
2672 a new connection before checking whether the processes handling previous
2673 connections had finished. The check for completed processes is now done
2674 earlier. On busy systems, this bug wouldn't be noticed because something
2675 else would have woken the daemon, and it would have reaped the completed
2678 80. If a message was submitted locally by a user whose login name contained one
2679 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2680 It caused a spool format error. I have fixed the spool reading code. A
2681 related problem was that the "from" clause in the Received: line became
2682 illegal because of the space(s). It is now covered by ${quote_local_part.
2684 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2687 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2689 83. Patch from Sieve maintainer to fix unterminated string problem in
2690 "vacation" handling.
2692 84. Some minor changes to the Linux configuration files to help with other
2693 OS variants using glibc.
2695 85. One more patch for Sieve to update vacation handling to latest spec.
2698 ----------------------------------------------------
2699 See the note above about the 4.44 and 4.50 releases.
2700 ----------------------------------------------------
2706 1. Change 4.43/35 introduced a bug that caused file counts to be
2707 incorrectly computed when quota_filecount was set in an appendfile
2710 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2711 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2714 3. Give more explanation in the error message when the command for a transport
2715 filter fails to execute.
2717 4. There are several places where Exim runs a non-Exim command in a
2718 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2719 was being done only for the command run by the queryprogram router. It is
2720 now done for all such subprocesses. The other cases are: ${run, transport
2721 filters, and the commands run by the lmtp and pipe transports.
2723 5. Some older OS have a limit of 256 on the maximum number of file
2724 descriptors. Exim was using setrlimit() to set 1000 as a large value
2725 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2726 systems. I've change it so that if it can't get 1000, it tries for 256.
2728 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2729 was an oversight, and furthermore, ever since the addition of extra
2730 controls (e.g. 4.43/32), the checks on when to allow different forms of
2731 "control" were broken. There should now be diagnostics for all cases when a
2732 control that does not make sense is encountered.
2734 7. $recipients is now available in the predata ACL (oversight).
2736 8. Tidy the search cache before the fork to do a delivery from a message
2737 received from the command line. Otherwise the child will trigger a lookup
2738 failure and thereby defer the delivery if it tries to use (for example) a
2739 cached ldap connection that the parent has called unbind on.
2741 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2742 of $address_data from the recipient verification was clobbered by the
2743 sender verification.
2745 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2746 was its contents. (It was OK if the option was not defined at all.)
2748 11. A "Completed" log line is now written for messages that are removed from
2749 the spool by the -Mrm option.
2751 12. $host_address is now set to the target address during the checking of
2752 ignore_target_hosts.
2754 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2755 being passed; this would have caused $sender_host_name to have been used if
2756 matching the list had actually called for a host name (not very likely,
2757 since this list is usually IP addresses). A host name is now passed as
2760 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2761 code that specifies a non-restarting handler (typically sigaction() in
2762 modern systems) in an attempt to fix a rare and obscure crash bug.
2764 15. Narrowed the window for a race in the daemon that could cause it to ignore
2765 SIGCHLD signals. This is not a major problem, because they are used only to
2766 wake it up if nothing else does.
2768 16. A malformed maildirsize file could cause Exim to calculate negative values
2769 for the mailbox size or file count. Odd effects could occur as a result.
2770 The maildirsize information is now recalculated if the size or filecount
2773 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2774 support for a long time. Removed HAVE_SYS_VFS_H.
2776 18. Updated exipick to current release from John Jetmore.
2778 19. Allow an empty sender to be matched against a lookup in an address list.
2779 Previously the only cases considered were a regular expression, or an
2782 20. Exim went into a mad DNS lookup loop when doing a callout where the
2783 host was specified on the transport, if the DNS lookup yielded more than
2786 21. The RFC2047 encoding function was originally intended for short strings
2787 such as real names; it was not keeping to the 75-character limit for
2788 encoded words that the RFC imposes. It now respects the limit, and
2789 generates multiple encoded words if necessary. To be on the safe side, I
2790 have increased the buffer size for the ${rfc2047: expansion operator from
2793 22. Failure to deliver a bounce message always caused it to be frozen, even if
2794 there was an errors_to setting on the router. The errors_to setting is now
2797 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2798 canonical form (fully expanded) before being placed in
2799 $sender_host_address.
2801 24. Updated eximstats to version 1.33
2803 25. Include certificate and key file names in error message when GnuTLS fails
2804 to set them up, because the GnuTLS error message doesn't include the name
2805 of the failing file when there is a problem reading it.
2807 26. Expand error message when OpenSSL has problems setting up cert/key files.
2810 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2811 (this can affect the format of dates).
2813 28. exim_tidydb, when checking for the continued existence of a message for
2814 which it has found a message-specific retry record, was not finding
2815 messages that were in split spool directories. Consequently, it was
2816 deleting retry records that should have stayed in existence.
2818 29. eximstats updated to version 1.35
2819 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2820 1.35 - bugfix such that pie charts by volume are generated correctly
2822 30. The SPA authentication driver was not abandoning authentication and moving
2823 on to the next authenticator when an expansion was forced to fail,
2824 contradicting the general specification for all authenticators. Instead it
2825 was generating a temporary error. It now behaves as specified.
2827 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2828 (the order specifies the preference for clients). The order is now AES256,
2829 AES128, 3DES, ARCFOUR128.
2831 31. Small patch to Sieve code - explicitly set From: when generating an
2834 32. Exim crashed if a remote delivery caused a very long error message to be
2835 recorded - for instance if somebody sent an entire SpamAssassin report back
2836 as a large number of 550 error lines. This bug was coincidentally fixed by
2837 increasing the size of one of Exim's internal buffers (big_buffer) that
2838 happened as part of the Exiscan merge. However, to be on the safe side, I
2839 have made the code more robust (and fixed the comments that describe what
2842 33. Some experimental protocols are using DNS PTR records for new purposes. The
2843 keys for these records are domain names, not reversed IP addresses. The
2844 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2845 leaves it alone. Component reversal etc. now happens only for IP addresses.
2848 34. The host_aton() function is supposed to be passed a string that is known
2849 to be a valid IP address. However, in the case of IPv6 addresses, it was
2850 not checking this. This is a hostage to fortune. Exim now panics and dies
2851 if the condition is not met. A case was found where this could be provoked
2852 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2853 components; fortuitously, this particular loophole had already been fixed
2854 by change 4.50/55 or 4.44/33 above.
2856 If there are any other similar loopholes, the new check in host_aton()
2857 itself should stop them being exploited. The report I received stated that
2858 data on the command line could provoke the exploit when Exim was running as
2859 exim, but did not say which command line option was involved. All I could
2860 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2861 running as the user.
2864 35. There was a buffer overflow vulnerability in the SPA authentication code
2865 (which came originally from the Samba project). I have added a test to the
2866 spa_base64_to_bits() function which I hope fixes it.
2869 36. The daemon start-up calls getloadavg() while still root for those OS that
2870 need the first call to be done as root, but it missed one case: when
2871 deliver_queue_load_max is set with deliver_drop_privilege. This is
2872 necessary for the benefit of the queue runner, because there is no re-exec
2873 when deliver_drop_privilege is set.
2875 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2876 was in use, was not putting the data itself into the right store pool;
2877 consequently, it could be overwritten for a subsequent message in the same
2878 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2881 38. Sometimes the final signoff response after QUIT could fail to get
2882 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2883 < 0 before doing a fflush(). This bug looks as though it goes back to the
2884 introduction of TLS in release 3.20, but "sometimes" must have been rare
2885 because the tests only now provoked it.
2891 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2892 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2893 Now it returns an integer. A number of calls were still expecting a T/F
2894 return. Fortuitously, in all cases, the tests worked in OK situations,
2895 which is the norm. However, things would have gone wrong on any write
2896 failures on the smtp file descriptor. This function is used when sending
2897 messages over SMTP and also when doing verify callouts.
2899 2. When Exim is called to do synchronous delivery of a locally submitted
2900 message (the -odf or -odi options), it no longer closes stderr before doing
2903 3. Implemented the mua_wrapper option.
2905 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2907 5. Implemented the functions header_remove(), header_testname(),
2908 header_add_at_position(), and receive_remove_recipient(), and exported them
2911 6. If an ACL "warn" statement specified the addition of headers, Exim already
2912 inserted X-ACL-Warn: at the start if there was no header name. However, it
2913 was not making this test for the second and subsequent header lines if
2914 there were newlines in the string. This meant that an invalid header could
2915 be inserted if Exim was badly configured.
2917 7. Allow an ACL "warn" statement to add header lines at the start or after all
2918 the Received: headers, as well as at the end.
2920 8. Added the rcpt_4xx retry error code.
2922 9. Added postmaster_mailfrom=xxx to callout verification option.
2924 10. Added mailfrom=xxxx to the callout verification option, for verify=
2927 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2928 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2930 12. Inserted some casts to stop certain compilers warning when using pointer
2931 differences as field lengths or precisions in printf-type calls (mostly
2932 affecting debugging statements).
2934 13. Added optional readline() support for -be (dynamically loaded).
2936 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2937 same clock tick as a message's arrival, so that its received time was the
2938 same as the "first fail" time on the retry record, and that message
2939 remained on the queue past the ultimate address timeout, every queue runner
2940 would try a delivery (because it was past the ultimate address timeout) but
2941 after another failure, the ultimate address timeout, which should have then
2942 bounced the address, did not kick in. This was a "< instead of <=" error;
2943 in most cases the first failure would have been in the next clock tick
2944 after the received time, and all would be well.
2946 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2947 being recognized when the domain list was tested by the match_domain
2948 condition in an expansion string.
2950 16. Added the ${str2b64: operator.
2952 17. Exim was always calling setrlimit() to set a large limit for the number of
2953 processes, without checking whether the existing limit was already
2954 adequate. (It did check for the limit on file descriptors.) Furthermore,
2955 errors from getrlimit() and setrlimit() were being ignored. Now they are
2956 logged to the main and panic logs, but Exim does carry on, to try to do its
2957 job under whatever limits there are.
2959 18. Imported PCRE 5.0.
2961 19. Trivial typo in log message " temporarily refused connection" (the leading
2964 20. If the log selector return_path_on_delivery was set and an address was
2965 redirected to /dev/null, the delivery process crashed because it assumed
2966 that a return path would always be set for a "successful" delivery. In this
2967 case, the whole delivery is bypassed as an optimization, and therefore no
2970 21. Internal re-arrangement: the function for sending a challenge and reading
2971 a response while authentication was assuming a zero-terminated challenge
2972 string. It's now changed to take a pointer and a length, to allow for
2973 binary data in such strings.
2975 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2977 23. Exim was not respecting finduser_retries when seeking the login of the
2978 uid under which it was called; it was always trying 10 times. (The default
2979 setting of finduser_retries is zero.) Also, it was sleeping after the final
2980 failure, which is pointless.
2982 24. Implemented tls_on_connect_ports.
2984 25. Implemented acl_smtp_predata.
2986 26. If the domain in control=submission is set empty, Exim assumes that the
2987 authenticated id is a complete email address when it generates From: or
2988 Sender: header lines.
2990 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2991 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2992 chown and chgrp in /bin and hostname in /usr/bin.
2994 28. Exim was keeping the "process log" file open after each use, just as it
2995 does for the main log. This opens the possibility of it remaining open for
2996 long periods when the USR1 signal hits a daemon. Occasional processlog
2997 errors were reported, that could have been caused by this. Anyway, it seems
2998 much more sensible not to leave this file open at all, so that is what now
3001 29. The long-running daemon process does not normally write to the log once it
3002 has entered its main loop, and it closes the log before doing so. This is
3003 so that log files can straightforwardly be renamed and moved. However,
3004 there are a couple of unusual error situations where the daemon does write
3005 log entries, and I had neglected to close the log afterwards.
3007 30. The text of an SMTP error response that was received during a remote
3008 delivery was being truncated at 512 bytes. This is too short for some of
3009 the long messages that one sometimes sees. I've increased the limit to
3012 31. It is now possible to make retry rules that apply only when a message has a
3013 specific sender, in particular, an empty sender.
3015 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3016 it possible to be selective about when SMTP synchronization is enforced.
3018 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3020 32. Implemented hosts_connection_nolog.
3022 33. Added an ACL for QUIT.
3024 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3027 35. Added mailbox_size and mailbox_filecount to appendfile.
3029 36. Added control = no_multiline_responses to ACLs.
3031 37. There was a bug in the logic of the code that waits for the clock to tick
3032 in the case where the clock went backwards by a substantial amount such
3033 that the microsecond fraction of "now" was more than the microsecond
3034 fraction of "then" (but the whole seconds number was less).
3036 38. Added support for the libradius Radius client library this is found on
3037 FreeBSD (previously only the radiusclient library was supported).
3043 1. When certain lookups returned multiple values in the form name=value, the
3044 quoting of the values was not always being done properly. Specifically:
3045 (a) If the value started with a double quote, but contained no whitespace,
3047 (b) If the value contained whitespace other than a space character (i.e.
3048 tabs or newlines or carriage returns) it was not quoted.
3049 This fix has been applied to the mysql and pgsql lookups by writing a
3050 separate quoting function and calling it from the lookup code. The fix
3051 should probably also be applied to nisplus, ibase and oracle lookups, but
3052 since I cannot test any of those, I have not disturbed their existing code.
3054 2. A hit in the callout cache for a specific address caused a log line with no
3055 reason for rejecting RCPT. Now it says "Previous (cached) callout
3056 verification failure".
3058 3. There was an off-by-one bug in the queryprogram router. An over-long
3059 return line was truncated at 256 instead of 255 characters, thereby
3060 overflowing its buffer with the terminating zero. As well as fixing this, I
3061 have increased the buffer size to 1024 (and made a note to document this).
3063 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3064 when Exim is waiting for an SMTP response from a remote server, Exim
3065 restarts its select() call on the socket, thereby resetting its timeout.
3066 This is not a problem when such interrupts are rare. Somebody set up a cron
3067 job to run exiwhat every 2 minutes, which is less than the normal select()
3068 timeout (5 or 10 minutes). This meant that the select() timeout never
3069 kicked in because it was always reset. I have fixed this by comparing the
3070 time when an interrupt arrives with the time at the start of the first call
3071 to select(). If more time than the timeout has elapsed, the interrupt is
3072 treated as a timeout.
3074 5. Some internal re-factoring in preparation for the addition of Sieve
3075 extensions (by MH). In particular, the "personal" test is moved to a
3076 separate function, and given an option for scanning Cc: and Bcc: (which is
3077 not set for Exim filters).
3079 6. When Exim created an email address using the login of the caller as the
3080 local part (e.g. when creating a From: or Sender: header line), it was not
3081 quoting the local part when it contained special characters such as @.
3083 7. Installed new OpenBSD configuration files.
3085 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3086 try to make them clearer.
3088 9. Callout options, other than the timeout value, were being ignored when
3089 verifying sender addresses in header lines. For example, when using
3091 verify = header_sender/callout=no_cache
3093 the cache was (incorrectly) being used.
3095 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3096 only the Cygwin environment.
3098 11. When return_path_on_delivery was set as a log selector, if different remote
3099 addresses in the same message used different return paths and parallel
3100 remote delivery occurred, the wrong values would sometimes be logged.
3101 (Whenever a remote delivery process finished, the return path value from
3102 the most recently started remote delivery process was logged.)
3104 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3105 lines when AUTH and/or TLS are in use. This is the "received protocol"
3106 field. Exim used to use "asmtp" for authenticated SMTP, without any
3107 indication (in the protocol name) for TLS use. Now it follows the RFC and
3108 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3109 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3110 names appear in log lines as well as in Received: header lines.
3112 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3113 extensions, and comparison tests, and to fix some bugs.
3115 14. Changes to the "personal" filter test:
3117 (1) The test was buggy in that it was just doing the equivalent of
3118 "contains" tests on header lines. For example, if a user's address was
3119 anne@some.where, the "personal" test would incorrectly be true for
3121 To: susanne@some.where
3123 This test is now done by extracting each address from the header in turn,
3124 and checking the entire address. Other tests that are part of "personal"
3125 are now done using regular expressions (for example, to check local parts
3126 of addresses in From: header lines).
3128 (2) The list of non-personal local parts in From: addresses has been
3129 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3130 taken from the Sieve specification recommendations.
3132 (3) If the message contains any header line starting with "List-" it is
3133 treated as non-personal.
3135 (4) The test for "circular" in the Subject: header line has been removed
3136 because it now seems ill-conceived.
3138 15. Minor typos in src/EDITME comments corrected.
3140 16. Installed latest exipick from John Jetmore.
3142 17. If headers_add on a router specified a text string that was too long for
3143 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3144 of string_sprintf() is now avoided.
3146 18. $message_body_size was not set (it was always zero) when running the DATA
3147 ACL and the local_scan() function.
3149 19. For the "mail" command in an Exim filter, no default was being set for
3150 the once_repeat time, causing a random time value to be used if "once" was
3151 specified. (If the value happened to be <= 0, no repeat happened.) The
3152 default is now 0s, meaning "never repeat". The "vacation" command was OK
3153 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3154 (I found it when inspecting the code).
3156 20. There is now an overall timeout for performing a callout verification. It
3157 defaults to 4 times the callout timeout, which applies to individual SMTP
3158 commands during the callout. The overall timeout applies when there is more
3159 than one host that can be tried. The timeout is checked before trying the
3160 next host. This prevents very long delays if there are a large number of
3161 hosts and all are timing out (e.g. when the network connections are timing
3162 out). The value of the overall timeout can be changed by specifying an
3163 additional sub-option for "callout", called "maxwait". For example:
3165 verify = sender/callout=5s,maxwait=20s
3167 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3168 to the end before writing, but this should make it even safer).
3170 22. Exim was forgetting that it had advertised PIPELINING for the second and
3171 subsequent messages on an SMTP connection. It was also not resetting its
3172 memory on STARTTLS and an internal HELO.
3174 23. When Exim logs an SMTP synchronization error within a session, it now
3175 records whether PIPELINING has been advertised or not.
3177 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3178 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3179 rather than long int.
3181 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3183 26. Added the never_mail option to autoreply.
3189 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3190 crash if the getsockname() call failed; this can happen if a connection is
3191 closed very soon after it is established. The problem was simply in the
3192 order in which certain operations were done, causing Exim to try to write
3193 to the SMTP stream before it had set up the file descriptor. The bug has
3194 been fixed by making things happen in the correct order.
3200 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3201 before the connection was closed, thus losing the rejection response.
3203 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3204 some early Solaris releases, but causes trouble in current releases where
3205 socklen_t is defined.
3207 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3210 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3213 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3214 that releasing the top of it at the end releases what was used for sub-
3215 expansions (unless the block got too big). However, discard this block if
3216 the first thing is a variable or header, so that we can use its block when
3217 it is dynamic (useful for very large $message_headers, for example).
3219 6. Lookups now cache *every* query, not just the most recent. A new, separate
3220 store pool is used for this. It can be recovered when all lookup caches are
3221 flushed. Lookups now release memory at the end of their result strings.
3222 This has involved some general refactoring of the lookup sources.
3224 7. Some code has been added to the store_xxx() functions to reduce the amount
3225 of flapping under certain conditions.
3227 8. log_incoming_interface used to affect only the <= reception log lines. Now
3228 it causes the local interface and port to be added to several more SMTP log
3229 lines, for example "SMTP connection from", and rejection lines.
3231 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3233 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3235 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3236 could be overwritten at the end of the current message (or the start of a
3237 new message if it was set in a HELO ACL). The value is now preserved for
3238 the duration of the SMTP connection.
3240 12. If a transport had a headers_rewrite setting, and a matching header line
3241 contained an unqualified address, that address was qualified, even if it
3242 did not match any rewriting rules. The underlying bug was that the values
3243 of the flags that permit the existence of unqualified sender and recipient
3244 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3245 non-local messages, and by -bnq for local messages) were not being
3246 preserved with the message after it was received.
3248 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3249 "next input=" as part of the text comprising the host identity instead of
3250 the correct text. The code was using the same buffer for two different
3251 strings. However, depending on which order the printing function evaluated
3252 its arguments, the bug did not always show up. Under Linux, for example, my
3253 test suite worked just fine.
3255 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3256 doesn't work with some older versions of Perl. It has been changed to "my",
3257 which in any case is probably the better facility to use.
3259 15. A really picky compiler found some instances of statements for creating
3260 error messages that either had too many or two few arguments for the format
3263 16. The size of the buffer for calls to the DNS resolver has been increased
3264 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3265 for addresses that have a lot of PTR records. This alleviates a problem; it
3266 does not fully solve it.
3268 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3269 buffer now truncates the list and logs the incident, which is the same
3270 action as happens when Exim is looking up a host name and its aliases.
3271 Previously in this situation something unpredictable would happen;
3272 sometimes it was "internal error: store_reset failed".
3274 18. If a server dropped the connection unexpectedly when an Exim client was
3275 using GnuTLS and trying to read a response, the client delivery process
3276 crashed while trying to generate an error log message.
3278 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3279 string, for example:
3281 warn message = H1: something\nH2: something
3283 the text was added as a single header line from Exim's point of view
3284 though it ended up OK in the delivered message. However, searching for the
3285 second and subsequent header lines using $h_h2: did not work. This has been
3286 fixed. Similarly, if a system filter added multiple headers in this way,
3287 the routers could not see them.
3289 20. Expanded the error message when iplsearch is called with an invalid key to
3290 suggest using net-iplsearch in a host list.
3292 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3293 ACLs are no longer actually imposed (and a message to that effect is
3296 22. If a "gecos" field in a passwd entry contained escaped characters, in
3297 particular, if it contained a \" sequence, Exim got it wrong when building
3298 a From: or a Sender: header from that name. A second bug also caused
3299 incorrect handling when an unquoted " was present following a character
3300 that needed quoting.
3302 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3303 was not being matched caselessly.
3305 24. Arranged for all hyphens in the exim.8 source to be escaped with
3308 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3309 properly. Recipient callout cache records were still being keyed to include
3310 the sender, even when use_sender was set false. This led to far more
3311 callouts that were necessary. The sender is no longer included in the key
3312 when use_sender is false.
3314 26. Added "control = submission" modifier to ACLs.
3316 27. Added the ${base62d: operator to decode base 62 numbers.
3318 28. dnsdb lookups can now access SRV records.
3320 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3321 the configuration file.
3323 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3324 (-v) mode. This makes the output for a verbose queue run more intelligible.
3326 31. Added a use_postmaster feature to recipient callouts.
3328 32. Added the $body_zerocount variable, containing the number of binary zero
3329 bytes in the message body.
3331 33. The time of last modification of the "new" subdirectory is now used as the
3332 "mailbox time last read" when there is a quota error for a maildir
3335 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3337 35. Added +ignore_unknown as a special item in host lists.
3339 36. Code for decoding IPv6 addresses in host lists is now included, even if
3340 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3341 address was recognized as an IP address, but was then not correctly decoded
3342 into binary, causing unexpected and incorrect effects when compared with
3349 1. Very minor rewording of debugging text in manualroute to say "list of
3350 hosts" instead of "hostlist".
3352 2. If verify=header_syntax was set, and a header line with an unqualified
3353 address (no domain) and a large number of spaces between the end of the
3354 name and the colon was received, the reception process suffered a buffer
3355 overflow, and (when I tested it) crashed. This was caused by some obsolete
3356 code that should have been removed. The fix is to remove it!
3358 3. When running in the test harness, delay a bit after writing a bounce
3359 message to get a bit more predictability in the log output.
3361 4. Added a call to search_tidyup() just before forking a reception process. In
3362 theory, someone could use a lookup in the expansion of smtp_accept_max_
3363 per_host which, without the tidyup, could leave open a database connection.
3365 5. Added the variables $recipient_data and $sender_data which get set from a
3366 lookup success in an ACL "recipients" or "senders" condition, or a router
3367 "senders" option, similar to $domain_data and $local_part_data.
3369 6. Moved the writing of debug_print from before to after the "senders" test
3372 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3373 problems for message scanning, either using a data ACL, or using
3374 local_scan() because the Received: header was not generated till after they
3375 were called (in order to set the time as the time of reception completion).
3376 I have revised the way this works. The header is now generated after the
3377 body is received, but before the ACL or local_scan() are called. After they
3378 are run, the timestamp in the header is updated.
3384 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3385 before starting a queue runner without re-exec. This happened only when
3386 deliver_drop_privilege was set or when the Exim user was set to root. The
3387 effect of the bug was that timeouts during subsequent deliveries caused
3388 crashes instead of being properly handled. The handler is now left at its
3389 default (and expected) setting.
3391 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3392 message, again when deliver_drop_privilege is set or Exim is run as root.
3393 The bug described in (1) was not present in this case, but the tidying up
3394 of the other signals was missing. I have made the two cases consistent.
3396 3. The ignore_target_hosts setting on a manualroute router was being ignored
3397 for hosts that were looked up using the /MX notation.
3399 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3402 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3403 operated on the sender address. After changing the $sender_address to <>
3404 for the sender address verify, Exim was re-instated it as the original
3405 (before rewriting) address, but remembering that it had rewritten it, so it
3406 wasn't rewriting it again. This bug also had the effect of breaking the
3407 sender address verification caching when the sender address was rewritten.
3409 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3410 This has been changed so that if the ip literal address matches
3411 ignore_target_hosts, the router declines.
3413 7. Added expansion conditions match_domain, match_address, and match_local_
3414 part (NOT match_host).
3416 8. The placeholder for the Received: header didn't have a length field set.
3418 9. Added code to Exim itself and to exim_lock to test for a specific race
3419 condition that could lead to file corruption when using MBX delivery. The
3420 issue is with the lockfile that is created in /tmp. If this file is removed
3421 after a process has opened it but before that process has acquired a lock,
3422 there is the potential for a second process to recreate the file and also
3423 acquire a lock. This could lead to two Exim processes writing to the file
3424 at the same time. The added code performs the same test as UW imapd; it
3425 checks after acquiring the lock that its file descriptor still refers to
3426 the same named file.
3428 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3429 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3432 11. Added the smtp_active_hostname option. If used, this will typically be made
3433 to depend on the incoming interface address. Because $interface_address is
3434 not set up until the daemon has forked a reception process, error responses
3435 that can happen earlier (such as "too many connections") no longer contain
3438 12. If an expansion in a condition on a "warn" statement fails because a lookup
3439 defers, the "warn" statement is abandoned, and the next ACL statement is
3440 processed. Previously this caused the whole ACL to be aborted.
3442 13. Added the iplsearch lookup type.
3444 14. Added ident_timeout as a log selector.
3446 15. Added tls_certificate_verified as a log selector.
3448 16. Added a global option tls_require_ciphers (compare the smtp transport
3449 option of the same name). This controls incoming TLS connections.
3451 17. I finally figured out how to make tls_require_ciphers do a similar thing
3452 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3453 before starting the TLS session.
3455 18. Tabs are now shown as \t in -bP output.
3457 19. If the log selector return_path_on_delivery was set, Exim crashed when
3458 bouncing a message because it had too many Received: header lines.
3460 20. If two routers both had headers_remove settings, and the first one included
3461 a superfluous trailing colon, the final name in the first list and the
3462 first name in the second list were incorrectly joined into one item (with a
3463 colon in the middle).
3469 1. Added -C and -D options to the exinext utility, mainly to make it easier
3470 to include in the automated testing, but these could be helpful when
3471 multiple configurations are in use.
3473 2. The exinext utility was not formatting the output nicely when there was
3474 an alternate port involved in the retry record key, nor when there was a
3475 message id as well (for retries that were specific to a specific message
3476 and a specific host). It was also confused by IPv6 addresses, because of
3477 the additional colons they contain. I have fixed the IPv4 problem, and
3478 patched it up to do a reasonable job for IPv6.
3480 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3481 delivery, the log line now contains "pipelined" if PIPELINING was used.
3483 4. An SMTP transport process used to panic and die if the bind() call to set
3484 an explicit outgoing interface failed. This has been changed; it is now
3485 treated in the same way as a connect() failure.
3487 5. A reference to $sender_host_name in the part of a conditional expansion
3488 that was being skipped was still causing a DNS lookup. This no longer
3491 6. The def: expansion condition was not recognizing references to header lines
3492 that used bh_ and bheader_.
3494 7. Added the _cache feature to named lists.
3496 8. The code for checking quota_filecount in the appendfile transport was
3497 allowing one more file than it should have been.
3499 9. For compatibility with Sendmail, the command line option
3507 and sets the incoming protocol and host name (for trusted callers). The
3508 host name and its colon can be omitted when only the protocol is to be set.
3509 Note the Exim already has two private options, -pd and -ps, that refer to
3510 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3511 "s", but I don't think that's a major issue.
3513 10. A number of refactoring changes to the code, none of which should affect
3516 (a) The number of logging options was getting close to filling up the
3517 32-bit word that was used as a bit map. I have split them into two classes:
3518 those that are passed in the argument to log_write(), and those that are
3519 only ever tested independently outside of that function. These are now in
3520 separate 32-bit words, so there is plenty of room for expansion again.
3521 There is no change in the user interface or the logging behaviour.
3523 (b) When building, for example, log lines, the code previously used a
3524 macro that called string_cat() twice, in order to add two strings. This is
3525 not really sufficiently general. Furthermore, there was one instance where
3526 it was actually wrong because one of the argument was used twice, and in
3527 one call a function was used. (As it happened, calling the function twice
3528 did not affect the overall behaviour.) The macro has been replaced by a
3529 function that can join an arbitrary number of extra strings onto a growing
3532 (c) The code for expansion conditions now uses a table and a binary chop
3533 instead of a serial search (which was left over from when there were very
3534 few conditions). Also, it now recognizes conditions like "pam" even when
3535 the relevant support is not compiled in: a suitably worded error message is
3536 given if an attempt is made to use such a condition.
3538 11. Added ${time_interval:xxxxx}.
3540 12. A bug was causing one of the ddress fields not to be passed back correctly
3541 from remote delivery subprocesses. The field in question was not being
3542 subsequently used, so this caused to problems in practice.
3544 13. Added new log selectors queue_time and deliver_time.
3546 14. Might have fixed a bug in maildirsizefile handling that threw up
3547 "unexpected character" debug warnings, and recalculated the data
3548 unnecessarily. In any case, I expanded the warning message to give more
3551 15. Added the message "Restricted characters in address" to the statements in
3552 the default ACL that block characters like @ and % in local parts.
3554 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3555 Three changes have been made:
3557 (a) There was a serious bug; a negative response to MAIL caused the whole
3558 recipient domain to be cached as invalid, thereby blocking all messages
3559 to all local parts at the same domain, from all senders. This bug has
3560 been fixed. The domain is no longer cached after a negative response to
3561 MAIL if the sender used is not empty.
3563 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3566 (c) A new callout option, "use_sender" has been added for people who want
3567 the modified behaviour.
3573 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3576 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3577 indeed breaks things for older releases.
3579 3. Added additional logging to the case where there is a problem reading data
3580 from a filter that is running in a subprocess using a pipe, in order to
3581 try to track down a specific problem.
3583 4. Testing facility fudge: when running in the test harness and attempting
3584 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3585 getting "No route to host". Convert this to a timeout.
3587 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3590 6. Some OS don't have socklen_t but use size_t instead. This affects the
3591 fifth argument of getsockopt() amongst other things. This is now
3592 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3593 can be set for individual OS. I have set it for SunOS5, OSF1, and
3594 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3595 some earlier ones do not.
3597 7. Change 4.30/15 was not doing the test caselessly.
3599 8. The standard form for an IPv6 address literal was being rejected by address
3600 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3601 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3602 this, as well as the form without the "IPv6" on the front (but only when
3603 address literals are enabled, of course).
3605 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3607 10. Exim crashed if a message with an empty sender address specified by -f
3608 encountered a router with an errors_to setting. This could be provoked only
3609 by a command such as
3613 where an empty string was supplied; "<>" did not hit this bug.
3615 11. Installed PCRE release 4.5.
3617 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3618 remained set. It is now erased.
3620 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3621 times from message ids (which are base 36 rather than the normal 62).
3623 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3624 were being counted as actual protocol errors, and logged if the log
3625 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3626 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3627 and DATA following a set of rejected RCPTs do not count as protocol errors.
3628 In other words, Exim assumes they were pipelined, though this may not
3629 actually be the case. Of course, in all cases the client gets an
3630 appropriate error code.
3632 15. If a lookup fails in an ACL condition, a message about the failure may
3633 be available; it is used if testing the ACL cannot continue, because most
3634 such messages specify what the cause of the deferral is. However, some
3635 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3636 that caused an old message to be retained and used if a later statement
3637 caused a defer, replacing the real cause of the deferral.
3639 16. If an IP address had so many PTR records that the DNS lookup buffer
3640 was not large enough to hold them, Exim could crash while trying to process
3641 the truncated data. It now detects and logs this case.
3643 17. Further to 4.21/58, another change has been made: if (and only if) the
3644 first line of a message (the first header line) ends with CRLF, a bare LF
3645 in a subsequent header line has a space inserted after it, so as not to
3646 terminate the header.
3648 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3649 unnecessarily, used atoi() instead of strtol(), and didn't check the
3650 termination when getting file sizes from file names by regex.
3652 19. Completely re-implemented the support for maildirsize files, in the light
3653 of a number of problems with the previous contributed implementation
3654 (4.30/29). In particular:
3656 . If the quota is zero, the maildirsize file is maintained, but no quota is
3659 . If the maildir directory does not exist, it is created before any attempt
3660 to write a maildirsize file.
3662 . The quota value in the file is just a cache; if the quota is changed in
3663 the transport, the new value overrides.
3665 . A regular expression is available for excluding directories from the
3668 20. The autoreply transport checks the characters in options that define the
3669 message's headers; it allows continued headers, but it was checking with
3670 isspace() after an embedded newline instead of explicitly looking for a
3673 21. If all the "regular" hosts to which an address was routed had passed their
3674 expiry times, and had not reached their retry times, the address was
3675 bounced, even if fallback hosts were defined. Now Exim should go on to try
3678 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3679 equivalent code in the SMTP transport. Some hosts send humungous responses
3680 to HELO/EHLO, more than 1024 it seems.
3682 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3683 gives compiler warnings in some environments. I've now done it "properly",
3686 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3687 (because of problems with the built-in one) was declared to return uschar *
3688 instead of char *, causing compiler failure.
3690 25. Fixed a file descriptor leak when processing alias/forward files.
3692 26. Fixed a minor format string issue in dbfn.c.
3694 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3696 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3697 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3700 29. When a sender address is verified, it is cached, to save repeating the test
3701 when there is more than one recipient in a message. However, when the
3702 verification involves a callout, it is possible for different callout
3703 options to be set for different recipients. It is too complicated to keep
3704 track of this in the cache, so now Exim always runs a verification when a
3705 callout is required, relying on the callout cache for the optimization.
3706 The overhead is duplication of the address routing, but this should not be
3709 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3710 to be verified with callout=postmaster, and the main callout worked but the
3711 postmaster check failed, the verification correctly failed. However, if a
3712 subsequent RCPT command asked for sender verification *without* the
3713 postmaster check, incorrect caching caused this verification also to fail,
3716 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3717 it was not caching the DNS options (qualify_single, search_parents) that
3718 were used when the lookup failed. A subsequent lookup with different
3719 options therefore always gave the same answer, though there were cases
3720 where it should not have. (Example: a "domains = !$mx_any" option on a
3721 dnslookup router: the "domains" option is always processed without any
3722 widening, but the router might have qualify_single set.) Now Exim uses the
3723 cached value only when the same options are set.
3725 32. Added John Jetmore's "exipick" utility to the distribution.
3727 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3728 than a timeout (e.g. a certificate is required, and is not provided), an
3729 Exim server now closes the connection immediately. Previously it waited for
3730 the client to close - but if the client is SSL, it seems that they each
3731 wait for each other, leading to a delay before one of them times out.
3733 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3734 maintained 0.8.x compatibility because I don't think many are using it, and
3735 it is clearly obsolete.
3737 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3740 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3741 client certificate was expired. A simple patch fixes this, though I don't
3742 understand the full logic of why the verify callback is called multiple
3745 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3746 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3747 which causes problems with some clients (such as the Certicom SSL Plus
3748 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3749 disables the coutermeasure allowing Eudora to connect."
3751 38. Exim was not checking that a write() to a log file succeeded. This could
3752 lead to Bad Things if a log got too big, in particular if it hit a file
3753 size limit. Exim now panics and dies if it cannot write to a log file, just
3754 as it does if it cannot open a log file.
3756 39. Modified OS/Makefile-Linux so that it now contains
3758 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3760 The two -D definitions ensure that Exim is compiled with large file
3761 support, which makes it possible to handle log files that are bigger than
3764 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3765 instance) a domain was checked against a named list that involved a lookup,
3766 causing $domain_data to be set, then another domain was checked against the
3767 same list, then the first domain was re-checked, the value of $domain_data
3768 after the final check could be wrong. In particular, if the second check
3769 failed, it could be set empty. This bug probably also applied to
3772 41. The strip_trailing_dot option was not being applied to the address given
3773 with the -f command-line option.
3775 42. The code for reading a message's header from the spool was incrementing
3776 $received_count, but never initializing it. This meant that the value was
3777 incorrect (doubled) while delivering a message in the same process in which
3778 it was received. In the most common configuration of Exim, this never
3779 happens - a fresh exec is done - but it can happen when
3780 deliver_drop_privilege is set.
3782 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3783 it now includes up to 150 characters of the unexpected data in the log
3786 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3787 and building data strings. The size of both of these buffers was 10 000
3788 bytes - far larger than anybody would *ever* want, thought I. Needless to
3789 say, somebody hit the limit. I have increased the maximum line length to
3790 20 000 and the maximum data length of concatenated lines to 100 000. I have
3791 also fixed two bugs, because there was no checking on these buffers. Tsk,
3792 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3793 buffer is too small.
3795 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3796 lsearch lookups. Now it does.
3798 46. When parsing a route_list item in a manualroute router, a fixed-length
3799 buffer was used for the list of hosts. I made this 1024 bytes long,
3800 thinking that nobody would ever have a list of hosts that long. Wrong.
3801 Somebody had a whole pile of complicated expansion conditions, and the
3802 string was silently truncated, leading to an expansion error. It turns out
3803 that it is easier to change to an unlimited length (owing to other changes
3804 that have happened since this code was originally written) than to build
3805 structure for giving a limitation error. The length of the item that
3806 expands into the list of hosts is now unlimited.
3808 47. The lsearch lookup could not handle data where the length of text line was
3809 more than 4095 characters. Such lines were truncated, leading to shortened
3810 data being returned. It should now handle lines of any length.
3812 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3813 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3816 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3817 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3818 no longer happens in comment lines. A list of replacements is now placed
3819 at the head of all of the source files, except those whose only change is
3820 to replace PERL_COMMAND in the very first #! line.
3822 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3823 messages on the queue, with a bottom-up merge sort, using code contributed
3824 by Michael Haardt. This should make operations like -bp somewhat faster on
3825 large queues. It won't affect queue runners, except when queue_run_in_order
3828 51. Installed eximstats 1.31 in the distribution.
3830 52. Added support for SRV lookups to the dnslookup router.
3832 53. If an ACL referred to $message_body or $message_body_end, the value was not
3833 reset for any messages that followed in the same SMTP session.
3835 54. The store-handling optimization for building very long strings was not
3836 differentiating between the different store pools. I don't think this
3837 actually made any difference in practice, but I've tidied it.
3839 55. While running the routers to verify a sender address, $sender_address
3840 was still set to the sender address. This is wrong, because when routing to
3841 send a bounce to the sender, it would be empty. Therefore, I have changed
3842 it so that, while verifying a sender address, $sender_address is set to <>.
3843 (There is no change to what happens when verifying a recipient address.)
3845 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3846 target A or AAAA records (if not already returned) without resetting the
3847 qualify_single or search_parents options of the DNS resolver. These are
3848 inappropriate in this case because the targets of MX and SRV records must
3849 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3850 target that, when qualified, matched something in the local domain. These
3851 two options are now turned off when doing these lookups.
3853 57. It seems that at least some releases of Reiserfs (which does not have the
3854 concept of a fixed number of inodes) returns zero and not -1 for the
3855 number of available inodes. This interacted badly with check_spool_inodes,
3856 which assumed that -1 was the "no such thing" setting. What I have done is
3857 to check that the total number of inodes is greater than zero before doing
3858 the test of how many are available.
3860 58. When a "warn" ACL statement has a log_message modifier, the message is
3861 remembered, and not repeated. This is to avoid a lot of repetition when a
3862 message has many recipients that cause the same warning to be written.
3863 Howewer, Exim was preserving the list of already written lines for an
3864 entire SMTP session, which doesn't seem right. The memory is now reset if a
3865 new message is started.
3867 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3868 debugging output unless log_rewrite was also set.
3870 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3871 of (int)(handle) when we know that handle contains (void *)(-1).
3873 61. The Exim daemon panic-logs an error return when it closes the incoming
3874 connection. However "connection reset by peer" seems to be common, and
3875 isn't really an error worthy of noting specially, so that particular error
3878 62. When Exim is trying to find all the local interfaces, it used to panic and
3879 die if the ioctl to get the interface flags failed. However, it seems that
3880 on at least one OS (Solaris 9) it is possible to have an interface that is
3881 included in the list of interfaces, but for which you get a failure error
3882 for this call. This happens when the interface is not "plumbed" into a
3883 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3884 failure of the "get flags" call assumes that the interface is down.
3886 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3887 makes life easier for people who are doing arithmetic on fields extracted
3888 from dates, where you often get leading zeros that should not be
3889 interpreted as octal.
3891 64. Added qualify_domain to the redirect router, to override the global
3894 65. If a pathologically long header line contained very many addresses (the
3895 report of this problem mentioned 10 000) and each of them was rewritten,
3896 Exim could use up a very large amount of memory. (It kept on making new
3897 copies of the header line as it rewrote, and never released the old ones.)
3898 At the expense of a bit more processing, the header rewriting function has
3899 been changed so that it no longer eats memory in this way.
3901 66. The generation of the Received: header has been moved from the time that a
3902 message starts to be received, to the time that it finishes. The timestamp
3903 in the Received: header should now be very close to that of the <= log
3904 line. There are two side-effects of this change:
3906 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3907 logged header lines no longer include the local Received: line, because
3908 it has not yet been created. The same applies to a copy of the message
3909 that is returned to a non-SMTP sender when a message is rejected.
3911 (b) When a filter file is tested using -bf, no additional Received: header
3912 is added to the test message. After some thought, I decided that this
3915 This change does not affect the value of $received_for. It is still set
3916 after address rewriting, but before local_scan() is called.
3918 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3920 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3921 gave an unhelpful panic error message, and a defer error. I have managed to
3922 change this behaviour so that it now rejects any supplied certificate,
3923 which seems right, as the list of acceptable certificates is empty.
3925 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3926 gave an unhelpful defer error. I have not managed to make this reject any
3927 supplied certificates, but the error message it gives is "no certificate
3928 supplied", which is not helpful.
3930 70. exigrep's output now also includes lines that are not associated with any
3931 message, but which match the given pattern. Implemented by a patch from
3932 Martin Sluka, which also tidied up the Perl a bit.
3934 71. Recipient callout verification, like sender verification, was using <> in
3935 the MAIL FROM command. This isn't really the right thing, since the actual
3936 sender may affect whether the remote host accepts the recipient or not. I
3937 have changed it to use the actual sender in the callout; this means that
3938 the cache record is now keyed on a recipient/sender pair, not just the
3939 recipient address. There doesn't seem to be a real danger of callout loops,
3940 since a callout by the remote host to check the sender would use <>.
3941 [SEE ABOVE: changed after hitting problems.]
3943 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3944 temporary errors. However, in the case of such a code being given after
3945 the end of a data transmission (i.e. after ".") Exim was failing to write
3946 a retry record for the message. (Yes, there was some broken host that was
3947 actually sending 8xx at this point.)
3949 73. An unknown lookup type in a host list could cause Exim to panic-die when
3950 the list was checked. (An example that provoked this was putting <; in the
3951 middle of a list instead of at the start.) If this happened during a DATA
3952 ACL check, a -D file could be left lying around. This kind of configuration
3953 error no longer causes Exim to die; instead it causes a defer errror. The
3954 incident is still logged to the main and panic logs.
3956 74. Buglet left over from Exim 3 conversion. The message "too many messages
3957 in one connection" was written to the rejectlog but not the mainlog, except
3958 when address rewriting (yes!) was being logged.
3960 75. Added write_rejectlog option.
3962 76. When a system filter was run not as root (that is, when system_filter_user
3963 was set), the values of the $n variables were not being returned to the
3964 main process; thus, they were not subsequently available in the $sn
3967 77. Added +return_path_on_delivery log selector.
3969 78. A connection timeout was being treated differently from recipients deferred
3970 when testing hosts_max_try with a message that was older than the host's
3971 retry timeout. (The host should not be counted, thus allowing all hosts to
3972 be tried at least once before bouncing.) This may have been the cause of an
3973 occasionally reported bug whereby a message would remain on the queue
3974 longer than the retry timeout, but would be bounced if a delivery was
3975 forced. I say "may" because I never totally pinned down the problem;
3976 setting up timeout/retry tests is difficult. See also the next item.
3978 79. The ultimate address timeout was not being applied to errors that involved
3979 a combination of host plus message (for example, a timeout on a MAIL
3980 command). When an address resolved to a number of possible hosts, and they
3981 were not all tried for each delivery (e.g. because of hosts_max_try), a
3982 message could remain on the queue longer than the retry timeout.
3984 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3987 81. Fixed an obscure SMTP outgoing bug which required at least the following
3988 conditions: (a) there was another message waiting for the same server;
3989 (b) the server returned 5xx to all RCPT commands in the first message so
3990 that the message was not completed; (c) the server dropped the connection
3991 or gave a negative response to the RSET that Exim sends to abort the
3992 transaction. The observed case was a dropped connection after DATA that had
3993 been sent in pipelining mode. That is, the server had advertised PIPELINING
3994 but was not implementing it correctly. The effect of the bug was incorrect
3995 behaviour, such as trying another host, and this could lead to a crash.
4001 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4002 and daemon.c were passed as pointers to ints; they should have been
4003 pointers to socklen_t variables (which are typically unsigned ints).
4005 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4008 3. Fixed a really odd bug that affected only the testing scheme; patching a
4009 certain fixed string in the binary changed the value of another string that
4010 happened to be identical to the end of the original first string.
4012 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4013 name", it returns that address as the IP address. On some operating
4014 systems (e.g. Solaris), it also passes back the IP address string as the
4015 "host name". However, on others (e.g. Linux), it passes back an empty
4016 string. Exim wasn't checking for this, and was changing the host name to an
4017 empty string, assuming it had been canonicized.
4019 5. Although rare, it is permitted to have more than one PTR record for a given
4020 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4021 all the names associated with an address, because they do in Solaris.
4022 However, it seems that they do not in Linux for data that comes from the
4023 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4024 I found this out when I moved to a new Linux workstation and tried to run
4025 the Exim test suite.
4027 To get round this problem I have changed the code so that it now does its
4028 own call to the DNS to look up PTR records when searching for a host name.
4029 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4030 addresses that are only in /etc/hosts are still found.
4032 This behaviour is, however, controlled by an option called host_lookup_
4033 order, which defaults to "bydns:byaddr". If people want to use the other
4034 order, or indeed, just use one or the other means of lookup, they can
4035 specify it in this variable.
4037 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4038 some operating systems, this comes back from gethostbyaddr() as an empty
4039 string, and this is what Exim used to test for. However, it seems that in
4040 other systems, "." is yielded. Exim now tests for this case too.
4042 7. The values of check_spool_space and check_log_space are now held internally
4043 as a number of kilobytes instead of an absolute number of bytes. If a
4044 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4045 kilobyte. This means that much larger values can be stored.
4047 8. Exim monitor: an attempt to get the action menu when not actually pointing
4048 at a message produces an empty menu entitled "No message selected". This
4049 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4050 no entries in it ("Shell widget menu has zero width and/or height"). So I
4051 have added a single, blank menu entry in this case.
4053 9. Added ${quote_local_part.
4055 10. MIME decoding is now applied to the contents of Subject: header lines when
4058 11. Now that a reference to $sender_host_address automatically causes a reverse
4059 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4060 host lookup before query-style lookups in lists that might use this
4061 variable. This has therefore been abolished, and the "net-" prefix is no
4062 longer necessary for query-style lookups.
4064 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4065 have been a typo for LFLAGS, so it has been changed.
4067 13. The install script calls Exim with "-C /dev/null" in order to find the
4068 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4069 to be output. Howeve, since Exim outputs its version number before the
4070 error, it didn't break the script. It just looked ugly. I fixed this by
4071 always allowing "-C /dev/null" if the caller is root.
4073 14. Ignore overlarge ACL variable number when reading spool file - insurance
4074 against a later release with more variables having written the file.
4076 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4077 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4078 this, as well as the form without the "IPv6" on the front.
4080 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4081 OS/Makefile-Darwin file.
4083 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4084 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4086 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4087 variable. However, some OS use ints for inodes. I've added cast to long int
4088 to get rid of the compiler warning.
4090 19. I had forgotten to lock out "/../" in configuration file names when
4091 ALT_CONFIG_PREFIX was set.
4093 20. Routers used for verification do not need to specify transports. However,
4094 if such a router generated a host list, and callout was configured, Exim
4095 crashed, because it could not find a port number from the (non-existent)
4096 transport. It now assumes port 25 in this circumstance.
4098 21. Added the -t option to exigrep.
4100 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4101 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4102 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4103 them is set, LOOKUP_LSEARCH is forced.
4105 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4107 24. Added sender and host information to the "rejected by local_scan()" log
4108 line; previously there was no indication of these.
4110 25. Added .include_if_exists.
4112 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4113 Linux. It turns out that not all file systems support this. Apparently some
4114 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4115 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4116 not supported on the file descriptor, is now ignored when Exim is trying to
4117 sync a directory. This applies only to Linux.
4119 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4121 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4122 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4125 29. Added support for maildirsize files from supplied patch (modified a bit).
4127 30. The use of :fail: followed by an empty string could lead Exim to respond to
4128 sender verification failures with (e.g.):
4130 550 Verification failed for <xxx>
4131 550 Sender verify failed
4133 where the first response line was missing the '-' that indicates it is not
4134 the final line of the response.
4136 31. The loop for finding the name of the user that called Exim had a hardwired
4137 limit of 10; it now uses the value of finduser_retries, which is used for
4138 all other user lookups.
4140 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4143 33. Exim was neglecting to zero errno before one call of strtol() when
4144 expanding a string and expecting an integer value. On some systems this
4145 resulted in spurious "integer overflow" errors. Also, it was casting the
4146 result into an int without checking.
4148 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4149 did not work. The code looks as if it has *never* worked, though it appears
4150 to have been documented since at least releast 1.62. I have made it work.
4152 35. The "timeout_DNS" error in retry rules, also documented since at least
4153 1.62, also never worked. As it isn't clear exactly what this means, and
4154 clearly it isn't a major issue, I have abolished the feature by treating it
4155 as "timeout", and writing a warning to the main and panic logs.
4157 36. The display of retry rules for -brt wasn't always showing the error code
4160 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4161 timeout_connect_A, timeout_connect_MX.
4163 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4164 to the empty sender.
4166 39. The daemon was not analysing the content of -oX till after it had closed
4167 stderr and disconnected from the controlling terminal. This meant that any
4168 syntax errors were only noted on the panic log, and the return code from
4169 the command was 0. By re-arranging the code a little, I've made the
4170 decoding happen first, so such errors now appear on stderr, and the return
4171 code is 1. However, the actual setting up of the sockets still happens in
4172 the disconnected process, so errors there are still only recorded on the
4175 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4176 connections (as happens on some IP stacks) was logged at start up time as
4177 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4178 it from "IPv6 and IPv4", which means that two separate sockets are being
4181 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4182 says whether AF_INET or AF_INET6 was passed as an argument.
4184 42. Exiwhat output was messed up when time zones were included in log
4187 43. Exiwhat now gives more information about the daemon's listening ports,
4188 and whether -tls-on-connect was used.
4190 44. The "port" option of the smtp transport is now expanded.
4192 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4193 silently ignored. Now an error message is written to the main and panic
4196 46. There's a new ACL modifier called "logwrite" which writes to a log file
4197 as soon as it is encountered.
4199 47. Added $local_user_uid and $local_user_gid at routing time.
4201 48. Exim crashed when trying to verify a sender address that was being
4204 49. Exim was recognizing only a space character after ".include". It now also
4205 recognizes a tab character.
4207 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4208 extracting the relevant information from the specification. The man page no
4209 longer contains scrambled data for the -d option, and I've added a section
4210 at the front about calling Exim under different names.
4212 51. Added "extra_headers" argument to the "mail" command in filter files.
4214 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4217 53. Installed eximstats 1.29.
4219 54. Added transport_filter_timeout as a generic transport option.
4221 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4222 Cc: header lines. This was required by RFC 822, but it not required by RFC
4225 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4226 incoming messages that did not have them. Now it does so only if the
4227 message originates locally, that is, if there is no associated remote host
4228 address. When Resent- header lines are present, this applies to the Resent-
4229 lines rather than the non-Resent- lines.
4231 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4232 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4234 58. Messages for configuration errors now include the name of the main
4235 configuration file - useful now that there may be more than one file in a
4236 list (.included file names were always shown).
4238 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4239 for those rare installations that do not start the daemon as root or run it
4240 setuid root. I've cut out the call to initgroups() if the daemon is not
4243 60. The Exim user and group can now be bound into the binary as text strings
4244 that are looked up at the start of Exim's processing.
4246 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4248 62. Added $mailstore_basename variable.
4250 63. Installed patch to sieve.c from Michael Haardt.
4252 64. When Exim failed to open the panic log after failing to open the main log,
4253 the original message it was trying to log was written to stderr and debug
4254 output, but if they were not available (the usual case in production), it
4255 was lost. Now it is written to syslog before the two lines that record the
4256 failures to open the logs.
4258 65. Users' Exim filters run in subprocesses under the user's uid. It is
4259 possible for a "deliver" command or an alias in a "personal" command to
4260 provoke an address rewrite. If logging of address rewriting is configured,
4261 this fails because the process is not running as root or exim. There may be
4262 a better way of dealing with this, but for the moment (because 4.30 needs
4263 to be released), I have disabled address rewrite logging when running a
4264 filter in a non-root, non-exim process.
4270 1. The buildconfig auxiliary program wasn't quoting the value set for
4271 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4272 not defined. This bug was masked in 4.22 by the effect that was fixed in
4275 2. Some messages that were rejected after a message id was allocated were
4276 shown as "incomplete" by exigrep. It no longer does this for messages that
4277 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4279 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4280 have allow_domain_literals set, the ID did not get logged in the <= line.
4281 Domain literals are now always recognized in Message-ID: header lines.
4283 4. The first argument for a ${extract expansion item is the key name or field
4284 number. Leading and trailing spaces in this item were not being ignored,
4285 causing some misleading effects.
4287 5. When deliver_drop_privilege was set, single queue runner processes started
4288 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4289 same command in the process it spins off) were not dropping privilege.
4291 6. When the daemon running as "exim" started a queue runner, it always
4292 re-executed Exim in the spun-off process. This is a waste of effort when
4293 deliver_drop_privilege is set. The new process now just calls the
4294 queue-runner function directly.
4300 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4303 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4304 ignored. Though the use of -oP was forcing the writing of a pid file, it
4305 was always written to the default place.
4307 3. If the message "no IP address found for host xxxx" is generated during
4308 incoming verification, it is now followed by identification of the incoming
4309 connection (so you can more easily find what provoked it).
4311 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4313 5. Added some features to "harden" Exim a bit more against certain attacks:
4315 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4316 be put in Local/Makefile. This is like the never_users runtime option,
4317 but it cannot be overridden. The default setting is "root".
4319 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4320 prefix string with which any file named in a -C command line option
4323 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4324 is retained for -C and -D only if the caller of Exim is root. Without
4325 it, the exim user may also use -C and -D and retain privilege.
4327 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4328 command line option is disabled.
4330 6. Macro names set by the -D option must start with an upper case letter, just
4331 like macro names defined in the configuration file.
4333 7. Added "dereference=" facility to LDAP.
4335 8. Two instances of the typo "uknown" in the source files are fixed.
4337 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4338 the Configure-Makefile script screwed up while processing it.
4340 10. Incorporated PCRE 4.4.
4342 11. The SMTP synchronization check was not operating right at the start of an
4343 SMTP session. For example, it could not catch a HELO sent before the client
4344 waited for the greeting. There is now a check for outstanding input at the
4345 point when the greeting is written. Because of the duplex, asynchronous
4346 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4347 way, but not yet received, when the check is performed.
4349 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4350 on TCP/IP sockets, because this apparently causes some broken clients to
4353 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4354 unchanged) from the Cygwin maintainer.
4356 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4357 was supported instead of testing for mbx. Effectively a typo.
4359 15. The spa authenticator server code was not checking that the input it
4360 received was valid base64.
4362 16. The debug output line for the "set" modifier in ACLs was not showing the
4363 name of the variable that was being set.
4365 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4367 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4368 lookup on its own. Something else was needed to trigger the lookup. For
4369 example, a match in host_lookup or the need for a host name in a host list.
4370 Now, if $sender_host_name is referenced and the host name has not yet been
4371 looked up, a lookup is performed. If the lookup fails, the variable remains
4372 empty, and $host_lookup_failed is set to "1".
4374 19. Added "eqi" as a case-independent comparison operator.
4376 20. The saslauthd authentication condition could segfault if neither service
4377 nor realm was specified.
4379 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4380 error message that was logged was misleading, and incoming SMTP
4381 connections were dropped. The message is now more accurate, and temporary
4382 errors are given to SMTP connections.
4384 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4385 (except RSET) with a 421 error, until QUIT is received. However, it was
4386 failing to send a response to QUIT.
4388 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4389 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4390 ACL is now run after the helo_try_verify_hosts code.
4392 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4393 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4394 case-independent, so other case variants are also recognized). Apparently
4395 some systems use these upper case variants.
4397 25. If more than two messages were waiting for the same host, and a transport
4398 filter was specified for the transport, Exim sent two messages over the
4399 same TCP/IP connection, and then failed with "socket operation on non-
4400 socket" when it tried to send the third.
4402 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4404 27. The extern definition of crypt16() in expand.c was not being excluded when
4405 the OS had its own crypt16() function.
4407 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4408 as a preferred synonym for return_size_limit, both as an option and as an
4411 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4413 30. Changed the default configuration ACL to relax the local part checking rule
4414 for addresses that are not in any local domains. For these addresses,
4415 slashes and pipe symbols are allowed within local parts, but the sequence
4416 /../ is explicitly forbidden.
4418 31. SPA server authentication was not clearing the challenge buffer before
4421 32. log_message in a "warn" ACL statement was writing to the reject log as
4422 well as to the main log, which contradicts the documentation and doesn't
4423 seem right (because no rejection is happening). So I have stopped it.
4425 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4426 However, I am unable to do any testing of this.
4428 34. Fixed an infelicity in the appendfile transport. When checking directories
4429 for a mailbox, to see if any needed to be created, it was accidentally
4430 using path names with one or more superfluous leading slashes; tracing
4431 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4433 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4434 contents are added to the log line that is written for every discarded
4435 recipient. (Previously a log_message setting was ignored.)
4437 36. The ${quote: operator now quotes the string if it is empty.
4439 37. The install script runs exim in order to find its version number. If for
4440 some reason other than non-existence or emptiness, which it checks, it
4441 could not run './exim', it was installing it with an empty version number,
4442 i.e. as "exim-". This error state is now caught, and the installation is
4445 38. An argument was missing from the function that creates an error message
4446 when Exim fails to connect to the socket for saslauthd authentication.
4447 This could cause Exim to crash, or give a corrupted message.
4449 39. Added isip, isip4, and isip6 to ${if conditions.
4451 40. The ACL variables $acl_xx are now saved with the message, and can be
4452 accessed later in routers, transports, and filters.
4454 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4455 strings in the file are not string-expanded.
4457 42. If a MAIL command specified a SIZE value that was too large to fit into an
4458 int variable, the check against message_size_limit failed. Such values are
4459 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4460 day this will have to be increased, but I don't think I want to be around
4461 when emails are that large.
4468 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4469 iconv() is not standard in FreeBSD.
4471 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4472 IPv6 enabled. The observed symptom was a segmentation fault on return from
4473 the function os_common_find_running_interfaces() in src/os.c.
4475 3. In the check_special_case() function in daemon.c I had used "errno" as an
4476 argument name, which causes warnings on some systems. This was basically a
4477 typo, since it was named "eno" in the comments!
4479 4. The code that waits for the clock to tick (at a resolution of some fraction
4480 of a second) so as to ensure message-id uniqueness was always waiting for
4481 at least one whole tick, when it could have waited for less. [This is
4482 almost certainly not relevant at current processor speeds, where it is
4483 unlikely to ever wait at all. But we try to future-proof.]
4485 5. The function that sleeps for a time interval that includes fractions of a
4486 second contained a race. It did not block SIGALRM between setting the
4487 timer, and suspending (a couple of lines later). If the interval was short
4488 and the sigsuspend() was delayed until after it had expired, the suspension
4489 never ended. On busy systems this could lead to processes getting stuck for
4492 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4493 process, before it forks any delivery processes. The open lookup caching
4494 mechanism meant that the open file or database connection was passed into
4495 the delivery process. The problem was that delivery processes always tidy
4496 up cached lookup data. This could cause a problem for the next delivery
4497 process started by the queue runner, because the external queue runner
4498 process does not know about the closure. So the next delivery process
4499 still has data in the lookup cache. In the case of a file lookup, there was
4500 no problem because closing a file descriptor in a subprocess doesn't affect
4501 the parent. However, if the lookup was caching a connection to a database,
4502 the connection was closed, and the second delivery process was likely to
4503 see errors such as "PGSQL: query failed: server closed the connection
4504 unexpectedly". The problem has been fixed by closing all cached lookups
4505 in a queue runner before running a delivery process.
4507 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4508 seem to have the "const" qualifier which it has on other OS. I've
4511 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4512 *of the same type* (client or server) with the same public name that an
4513 error should be diagnosed.
4515 9. When Exim looked up a host name for an IP address, but failed to find the
4516 original IP address when looking up the host name (a safety check), it
4517 output the message "<ip address> does not match any IP for NULL", which was
4518 confusing, to say the least. The bug was that the host name should have
4519 appeared instead of "NULL".
4521 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4522 user that is built into the binary, and the -C or -D options is used, root
4523 privilege is dropped before the configuration file is read. In addition,
4524 logging is switched to stderr instead of the normal log files. If the
4525 configuration then re-defines the Exim user, the unprivileged environment
4526 is probably not what is expected, so Exim logs a panic warning message (but
4529 However, if deliver_drop_privilege is set, the unprivileged state may well
4530 be exactly what is intended, so the warning has been cut out in that case,
4531 and Exim is allowed to try to write to its normal log files.
4537 1. smtp_return_error_details was not giving details for temporary sender
4538 or receiver verification errors.
4540 2. Diagnose a configuration error if two authenticators have the same public
4543 3. Exim used not to create the message log file for a message until the first
4544 delivery attempt. This could be confusing when incoming messages were held
4545 for policy or load reasons. The message log file is now created at the time
4546 the message is received, and an initial "Received" line is written to it.
4548 4. The automatically generated man page for command line options had a minor
4549 bug that caused no ill effects; however, a more serious problem was that
4550 the procedure for building the man page automatically didn't always
4551 operate. Consequently, release 4.20 contains an out-of-date version. This
4552 shouldn't happen again.
4554 5. When building Exim with embedded Perl support, the script that builds the
4555 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4556 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4558 6. The freeze_tell option was not being used for messages that were frozen on
4559 arrival, either by an ACL or by local_scan().
4561 7. Added the smtp_incomplete_transaction log selector.
4563 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4564 was accepting AUTH without a new EHLO.
4566 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4567 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4570 10. Logging of TCP/IP connections (when configured) now happens in the main
4571 daemon process instead of the child process, so that the TCP/IP connection
4572 count is more accurate (but it can never be perfect).
4574 11. The use of "drop" in a nested ACL was not being handled correctly in the
4575 outer ACL. Now, if condition failure induced by the nested "drop" causes
4576 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4577 or "require"), the connection is dropped.
4579 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4580 that yield "discard" can now be used with an "accept" or a "discard" verb,
4581 but an error is generated for any others (because I can't see a useful way
4582 to define what should happen).
4584 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4585 are now processed in the same way as lines in the Exim configuration file.
4586 In particular, continuation lines are supported.
4588 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4590 15. Added -ti meaning -t -i.
4592 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4593 domains, and warn by logging if others are found.
4595 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4596 returned from GIFCONF when Exim is trying to find the list of interfaces on
4597 a host. The code in os.c has been modified to copy each ifreq to an aligned
4598 structure in all cases.
4600 Also, in some cases, the returned ifreq's were being copied to a 'struct
4601 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4602 means the last couple of bytes of an IPv6 address could be chopped if the
4603 ifreq contained only a normal sockaddr (14 bytes storage).
4605 18. Named domain lists were not supported in the hosts_treat_as_local option.
4606 An entry such as +xxxx was not recognized, and was treated as a literal
4609 19. Ensure that header lines added by a DATA ACL are included in the reject log
4610 if the ACL subsequently rejects the message.
4612 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4613 MD5 (which is deprecated).
4615 21. When testing a filter file using -bf, Exim was writing a message when it
4616 took the sender from a "From " line in the message, but it was not doing so
4617 when it took $return_path from a Return-Path: header line. It now does.
4619 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4620 with a valid header line field name (a series of printing characters
4621 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4623 23. Changed "disc" in the source to "disk" to conform to the documentation and
4624 the book and for uniformity.
4626 24. Ignore Sendmail's -Ooption=value command line item.
4628 25. When execve() failed while trying to run a command in a pipe transport,
4629 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4630 could be confused with a return value of 69 from the command itself. This
4631 has been changed to 127, the value the shell returns if it is asked to run
4632 a non-existent command. The wording for the related log line suggests a
4633 non-existent command as the problem.
4635 26. If received_header_text expands to an empty string, do not add a Received:
4636 header line to the message. (Well, it adds a token one on the spool, but
4637 marks it "old" so that it doesn't get used or transmitted.)
4639 27. Installed eximstats 1.28 (addition of -nt option).
4641 28. There was no check for failure on the call to getsockname() in the daemon
4642 code. This can fail if there is a shortage of resources on the system, with
4643 ENOMEM, for example. A temporary error is now given on failure.
4645 29. Contrary to the C standard, it seems that in some environments, the
4646 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4647 program. Exim now does this explicitly; it affects the formatting of
4648 timestamps using strftime().
4650 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4651 complaints. I've now initialized all the variables, to avoid this.
4653 32. Header lines added by a system filter were not being "seen" during
4654 transport-time rewrites.
4656 33. The info_callback() function passed to OpenSSL is set up with type void
4657 (*)(SSL *, int, int), as described somewhere. However, when calling the
4658 function (actually a macro) that sets it up, the type void(*)() is
4659 expected. I've put in a cast to prevent warnings from picky compilers.
4661 34. If a DNS black list lookup found a CNAME record, but there were no A
4662 records associated with the domain it pointed at, Exim crashed.
4664 35. If a DNS black list lookup returned more than one A record, Exim ignored
4665 all but the first. It now scans all returned addresses if a particular IP
4666 value is being sought. In this situation, the contents of the
4667 $dnslist_value variable are a list of all the addresses, separated by a
4670 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4671 to accept a host name and all its aliases if the forward lookup for any of
4672 them yielded the IP address of the incoming connection. Now it accepts only
4673 those names whose forward lookup yields the correct IP address. Any other
4674 names are discarded. This closes a loophole whereby a rogue DNS
4675 administrator could create reverse DNS records to break through a
4676 wildcarded host restriction in an ACL.
4678 37. If a user filter or a system filter that ran in a subprocess used any of
4679 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4680 the wrong values were passed to the pipe command ($thisaddress had the
4681 value of $0, $0 had the value of $1, etc). This bug was introduced by
4682 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4684 38. Improved the line breaking for long SMTP error messages from ACLs.
4685 Previously, if there was no break point between 40 and 75 characters, Exim
4686 left the rest of the message alone. Two changes have been made: (a) I've
4687 reduced the minimum length to 35 characters; (b) if it can't find a break
4688 point between 35 and 75 characters, it looks ahead and uses the first one
4689 that it finds. This may give the occasional overlong line, but at least the
4690 remaining text gets split now.
4692 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4693 file descriptors might be low, and that setting 1000 would always raise it.
4694 It turns out that in some environments, the limit is already over 1000 and
4695 that lowering it causes trouble. So now Exim takes care not to decrease it.
4697 40. When delivering a message, the value of $return_path is set to $sender_
4698 address at the start of routing (routers may change the value). By an
4699 oversight, this default was not being set up when an address was tested by
4700 -bt or -bv, which affected the outcome if any router or filter referred to
4703 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4704 log only when "message" or "log_message" are set. However, if one of the
4705 conditions was an address verification, or a call to a nested ACL, the
4706 messages generated by the underlying test were being passed through. This
4707 no longer happens. The underlying message is available in $acl_verify_
4708 message for both "message" and "log_message" expansions, so it can be
4709 passed through if needed.
4711 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4712 new expansion $bh_ to give the encoded byte string without charset
4713 translation. Translation happens only if iconv() is available; HAVE_ICONV
4714 indicates this at build time. HEADERS_CHARSET gives the charset to
4715 translate to; headers_charset can change it in the configuration, and
4716 "headers charset" can change it in an individual filter file.
4718 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4719 that creates RFC 2047 encoded "words" labels them as that charset instead
4720 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4721 expansion operator; (ii) when Exim creates a From: line for a local
4722 message; (iii) when a header line is rewritten to include a "phrase" part.
4724 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4725 buggy, causing it to skip the first lines of messages whose message ID
4726 ended in 'D'. This would not have bitten before Exim release 4.14, because
4727 message IDs were unlikely to end in 'D' before then. The effect was to have
4728 incorrect size information for certain domains.
4730 45. #include "config.h" was missing at the start of the crypt16.c module. This
4731 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4734 46. If there was a timeout during a "random" callout check, Exim treated it as
4735 a failure of the random address, and carried on sending RSET and the real
4736 address. If the delay was just some slowness somewhere, the response to the
4737 original RCPT would be taken as a response to RSET and so on, causing
4738 mayhem of various kinds.
4740 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4741 when I implemented it. It didn't allow for the fact that some option values
4742 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4743 the right test for positive values.
4745 48. Domain names in DNS records are case-independent. Exim always looks them up
4746 in lower case. Some resolvers return domain names in exactly the case they
4747 appear in the zone file, that is, they may contain uppercase letters. Not
4748 all resolvers do this - some return always lower case. Exim was treating a
4749 change of case by a resolver as a change of domain, similar to a widening
4750 of a domain abbreviation. This triggered its re-routing code and so it was
4751 trying to route what was effectively the same domain again. This normally
4752 caused routing to fail (because the router wouldn't handle the domain
4753 twice). Now Exim checks for this case specially, and just changes the
4754 casing of the domain that it ultimately uses when it transmits the message
4757 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4760 50. If a filter generated a file delivery with a non-absolute name (possible if
4761 no home directory exists for the router), the forbid_file option was not
4764 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4765 the existing equality matching.
4767 52. Exim was using ints instead of ino_t variables in some places where it was
4768 dealing with inode numbers.
4770 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4771 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4772 called TMPDIR, and if it finds it is different, it changes its value.
4774 54. The smtp_printf() function is now made available to local_scan() so
4775 additional output lines can be written before returning. There is also an
4776 smtp_fflush() function to enable the detection of a dropped connection.
4777 The variables smtp_input and smtp_batched_input are exported to
4780 55. Changed the default runtime configuration: the message "Unknown user"
4781 has been removed from the ACL, and instead placed on the localuser router,
4782 using the cannot_route_message feature. This means that any verification
4783 failures that generate their own messages won't get overridden. Similarly,
4784 the "Unrouteable address" message that was in the ACL for unverifiable
4785 relay addresses has also been removed.
4787 56. Added hosts_avoid_esmtp to the smtp transport.
4789 57. The exicyclog script was not checking for the esoteric option
4790 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4791 will work only if exicyclog is run under the appropriate euid.
4793 58. Following a discussion on the list, the rules by which Exim recognises line
4794 endings on incoming messages have been changed. The -dropcr and drop_cr
4795 options are now no-ops, retained only for backwards compatibility. The
4796 following line terminators are recognized: LF CRLF CR. However, special
4797 processing applies to CR:
4799 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4800 nor a local message in the state where . is a terminator.
4802 (ii) If a bare CR is encountered in a header line, an extra space is added
4803 after the line terminator so as not to end the header. The reasoning
4804 behind this is that bare CRs in header lines are most likely either
4805 to be mistakes, or people trying to play silly games.
4807 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4808 was being incorrectly given as 18 bytes larger than it should have been.
4809 This is a VOB (very old bug).
4811 60. This may never have affected anything current, but just in case it has:
4812 When the local host is found other than at the start of a list of hosts,
4813 the local host, those with the same MX, and any that follow, are discarded.
4814 When the list in question was part of a longer list of hosts, the following
4815 hosts (not currently being processed) were also being discarded. This no
4816 longer happens. I'm not sure if this situation could ever has previously
4819 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4822 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4823 header. This is something that is recommended in a new Internet Draft, and
4824 is something that is documented as being done by Sendmail. There are two
4825 possible values. For messages generated by the autoreply transport, Exim
4828 Auto-Submitted: auto-replied
4830 whereas for all other generated messages (e.g. bounces) it adds
4832 Auto-Submitted: auto-generated
4834 63. The "personal" condition in filters now includes a test for the
4835 Auto-Submitted: header. If it contains the string "auto-" the message it
4836 not considered personal.
4838 64. Added rcpt_include_affixes as a generic transport option.
4840 65. Added queue_only_override (default true).
4842 66. Added the syslog_duplication option.
4844 67. If what should have been the first header line of a message consisted of
4845 a space followed by a colon, Exim was mis-interpreting it as a header line.
4846 It isn't of course - it is syntactically invalid and should therefore be
4847 treated as the start of the message body. The misbehaviour could have
4848 caused a number of strange effects, including loss of data in subsequent
4849 header lines, and spool format errors.
4851 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4852 client host had authenticated. This control can now be exercised by an ACL
4853 for more flexibility.
4855 69. By default, callouts do not happen when testing with -bh. There is now a
4856 variant, -bhc, which does actually run the callout code, including
4857 consulting and updating the callout cache.
4859 70. Added support for saslauthd authentication, courtesy of Alexander
4862 71. If statvfs() failed on the spool or log directories while checking their
4863 size for availability, Exim confusingly gave the error "space shortage".
4864 Furthermore, in debugging mode it crashed with a floating point exception.
4865 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4866 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4867 serious problem, Exim now writes to the main and panic logs when this
4868 happens, with details of the failure. It then refuses to accept the
4869 incoming message, giving the message "spool directory problem" or "log
4870 directory problem" with a 421 code for SMTP messages.
4872 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4873 0, 1, and 2 exist, because some OS complain for execs without them (see
4874 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4875 descriptors. However, the code omitted to check that the open succeeded,
4876 causing mysterious errors if for some reason the permissions on /dev/null
4877 got screwed. Now Exim writes a message to the main and panic logs, and
4878 bombs out if it can't open /dev/null.
4880 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4881 interact so that it is all more flexible. It is supposed to remain
4882 backwards compatible. Also added extra_local_interfaces.
4884 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4885 to bomb out with an assertion failure - to the client this appears as a
4886 connection drop. This problem occurs in the part of the code that was taken
4887 from the Samba project. Fortunately, the assertion is in a very simple
4888 function, so I have fixed this by reproducing the function inline in the
4889 one place where it is called, and arranging for authentication to fail
4890 instead of killing the process with assert().
4892 75. The SPA client code was not working when the server requested OEM rather
4893 than Unicode encoding.
4895 76. Added code to make require_files with a specific uid setting more usable in
4896 the case where statting the file as root fails - usually a non-root-mounted
4897 NFS file system. When this happens and the failure is EACCES, Exim now
4898 forks a subprocess and does the per-uid checking as the relevant uid.
4900 77. Added process_log_path.
4902 78. If log_file_path was not explicitly set, a setting of check_log_space or
4903 check_log_inodes was ignored.
4905 79. If a space check for the spool or log partitions fails, the incident is now
4906 logged. Of course, in the latter case the data may get lost...
4908 80. Added the %p formatting code to string_format() so that it can be used to
4909 print addresses in debug_print(). Adjusted all the address printing in the
4910 debugging in store.c to use %p rather than %d.
4912 81. There was a concern that a line of code in smtp_in.c could overflow a
4913 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4914 initially expressed, the concern was not well-founded, because trailing
4915 spaces are removed early. However, if the trailing spaces were followed by
4916 a NULL, they did not get removed, so the overflow was possible. Two fixes
4919 (a) I re-wrote the offending code in a cleaner fashion.
4920 (b) If an incoming SMTP command contains a NULL character, it is rejected
4923 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4924 runs initgroups(), so that if the Exim user is in any additional groups,
4925 they will be used during message reception.
4931 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob