1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.558 2008/12/12 14:44:25 nm4 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: Improvement 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.
71 NM/06 Bugzilla 744: EXPN did not work under TLS.
72 Patch provided by Phil Pennock
74 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
75 Patch provided by Richard Godbee
77 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
80 NM/09 Bugzilla 787: Potential buffer overflow in string_format
81 Patch provided by Eugene Bujak
83 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to accept()
84 Patch provided by Maxim Dounin
90 TK/01 Add preliminary DKIM support. Currently requires a forked version of
91 ALT-N's libdkim that I have put here:
92 http://duncanthrax.net/exim-experimental/
94 Note to Michael Haardt: I had to rename some vars in sieve.c. They
95 were called 'true' and it seems that C99 defines that as a reserved
96 keyword to be used with 'bool' variable types. That means you could
97 not include C99-style headers which use bools without triggering
98 build errors in sieve.c.
100 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
101 as mailq or other aliases. Changed the --help handling significantly
102 to do whats expected. exim_usage() emits usage/help information.
104 SC/01 Added the -bylocaldomain option to eximstats.
106 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
108 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
110 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
116 PH/01 Another patch from the Sieve maintainer.
118 PH/02 When an IPv6 address is converted to a string for single-key lookup
119 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
120 dots are used instead of colons so that keys in lsearch files need not
121 contain colons. This was done some time before quoting was made available
122 in lsearch files. However, iplsearch files do require colons in IPv6 keys
123 (notated using the quote facility) so as to distinguish them from IPv4
124 keys. This meant that lookups for IP addresses in host lists did not work
125 for iplsearch lookups.
127 This has been fixed by arranging for IPv6 addresses to be expressed with
128 colons if the lookup type is iplsearch. This is not incompatible, because
129 previously such lookups could never work.
131 The situation is now rather anomolous, since one *can* have colons in
132 ordinary lsearch keys. However, making the change in all cases is
133 incompatible and would probably break a number of configurations.
135 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
138 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
139 conversion specification without a maximum field width, thereby enabling
140 a rogue spamd server to cause a buffer overflow. While nobody in their
141 right mind would setup Exim to query an untrusted spamd server, an
142 attacker that gains access to a server running spamd could potentially
143 exploit this vulnerability to run arbitrary code as the Exim user.
145 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
146 $primary_hostname instead of what libspf2 thinks the hosts name is.
148 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
149 a directory entry by the name of the lookup key. Previously, if a
150 symlink pointed to a non-existing file or a file in a directory that
151 Exim lacked permissions to read, a lookup for a key matching that
152 symlink would fail. Now it is enough that a matching directory entry
153 exists, symlink or not. (Bugzilla 503.)
155 PH/03 The body_linecount and body_zerocount variables are now exported in the
158 PH/04 Added the $dnslist_matched variable.
160 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
161 This means they are set thereafter only if the connection becomes
164 PH/06 Added the client_condition to authenticators so that some can be skipped
165 by clients under certain conditions.
167 PH/07 The error message for a badly-placed control=no_multiline_responses left
168 "_responses" off the end of the name.
170 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
172 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
173 (without spaces) instead of just copying the configuration text.
175 PH/10 Added the /noupdate option to the ratelimit ACL condition.
177 PH/11 Added $max_received_linelength.
179 PH/12 Added +ignore_defer and +include_defer to host lists.
181 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
182 way in which PCRE > 7.0 is built.
184 PH/14 Implemented queue_only_load_latch.
186 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
187 MAIL command. The effect was to mangle the value on 64-bit systems.
189 PH/16 Another patch from the Sieve maintainer.
191 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
193 PH/18 If a system quota error occurred while trying to create the file for
194 a maildir delivery, the message "Mailbox is full" was not appended to the
195 bounce if the delivery eventually timed out. Change 4.67/27 below applied
196 only to a quota excession during the actual writing of the file.
198 PH/19 It seems that peer DN values may contain newlines (and other non-printing
199 characters?) which causes problems in log lines. The DN values are now
200 passed through string_printing() before being added to log lines.
202 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
203 and InterBase are left for another time.)
205 PH/21 Added message_body_newlines option.
207 PH/22 Guard against possible overflow in moan_check_errorcopy().
209 PH/23 POSIX allows open() to be a macro; guard against that.
211 PH/24 If the recipient of an error message contained an @ in the local part
212 (suitably quoted, of course), incorrect values were put in $domain and
213 $local_part during the evaluation of errors_copy.
219 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
220 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
223 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
224 issue a MAIL command.
226 PH/02 In an ACL statement such as
228 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
230 if a client was not listed at all, or was listed with a value other than
231 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
232 the condition was not true (as it should be), so access was not denied.
233 The bug was that the ! inversion was incorrectly passed on to the second
234 item. This has been fixed.
236 PH/03 Added additional dnslists conditions == and =& which are different from
237 = and & when the dns lookup returns more than one IP address.
239 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
240 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
242 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
243 FSYNC, which compiles an option called disable_fsync that allows for
244 bypassing fsync(). The documentation is heavily laced with warnings.
246 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
248 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
249 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
250 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
251 including adding "make clean"; (3) Added -fPIC when compiling the test
252 dynamically loaded module, to get rid of a warning.
254 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
255 message fails, move_frozen_messages = true and ignore_bounce_errors_after
256 = 0s. The bug is otherwise harmless.
258 PH/07 There was a bug in the dovecot authenticator such that the value of
259 $auth1 could be overwritten, and so not correctly preserved, after a
260 successful authentication. This usually meant that the value preserved by
261 the server_setid option was incorrect.
263 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
265 PH/09 Installed PCRE release 7.0.
267 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
268 run for batched SMTP input. It is now run at the start of every message
269 in the batch. While fixing this I discovered that the process information
270 (output by running exiwhat) was not always getting set for -bs and -bS
271 input. This is fixed, and it now also says "batched" for BSMTP.
273 PH/11 Added control=no_pipelining.
275 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
276 patch, slightly modified), and move the expansion of helo_data till after
277 the connection is made in the smtp transport (so it can use these
280 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
282 PH/14 Added log_selector = +pid.
284 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
286 PH/16 Add ${if forany and ${if forall.
288 PH/17 Added dsn_from option to vary the From: line in DSNs.
290 PH/18 Flush SMTP output before performing a callout, unless control =
291 no_callout_flush is set.
293 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
294 was true (the default) a successful delivery failed to delete the retry
295 item, thus causing premature timeout of the address. The bug is now
298 PH/20 Added hosts_avoid_pipelining to the smtp transport.
300 PH/21 Long custom messages for fakedefer and fakereject are now split up
301 into multiline reponses in the same way that messages for "deny" and
302 other ACL rejections are.
304 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
305 with slight modification.
307 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
308 draft, changing the syntax and factoring some duplicate code".
310 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
311 for deliveries of the second and subsequent messages over the same SMTP
314 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
315 ${reduce, with only minor "tidies".
317 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
319 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
320 expansion side effects.
322 PH/27 When a message times out after an over-quota error from an Exim-imposed
323 quota, the bounce message says "mailbox is full". This message was not
324 being given when it was a system quota that was exceeded. It now should
327 MH/03 Made $recipients available in local_scan(). local_scan() already has
328 better access to the recipient list through recipients_list[], but
329 $recipients can be useful in postmaster-provided expansion strings.
331 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
332 in the case of a MAIL command with additional options following the
333 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
334 were accidentally chopped off.
336 PH/29 SMTP synchronization checks are implemented when a command is read -
337 there is a check that no more input is waiting when there shouldn't be
338 any. However, for some commands, a delay in an ACL can mean that it is
339 some time before the response is written. In this time, more input might
340 arrive, invalidly. So now there are extra checks after an ACL has run for
341 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
342 pipelining has not been advertised.
344 PH/30 MH's patch to allow iscntrl() characters to be list separators.
346 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
347 returned in the SMTP response when smtp_return_error_details was false.
350 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
351 instead of the C I/O that was originally supplied, because problems were
354 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
355 Exim which did not show up earlier: it was assuming that a call to
356 SSL_CTX_set_info_callback() might give an error value. In fact, there is
357 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
358 was a macro that became an assignment, so it seemed to work. This has
359 changed to a proper function call with a void return, hence the compile
360 error. Exim's code has been fixed.
362 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
365 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
367 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
369 PH/37 If a message is not accepted after it has had an id assigned (e.g.
370 because it turns out to be too big or there is a timeout) there is no
371 "Completed" line in the log. When some messages of this type were
372 selected by exigrep, they were listed as "not completed". Others were
373 picked up by some special patterns. I have improved the selection
374 criteria to be more general.
376 PH/38 The host_find_failed option in the manualroute router can now be set
377 to "ignore", to completely ignore a host whose IP address cannot be
378 found. If all hosts are ignored, the behaviour is controlled by the new
379 host_all_ignored option.
381 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
382 homing or because of multiple MX records with /mx) generated more than
383 one IP address, and the following item turned out to be the local host,
384 all the secondary addresses of the first item were incorrectly removed
385 from the list, along with the local host and any following hosts (which
386 is what is supposed to happen).
388 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
389 whoever called Exim into the -H file. In the case of the daemon it was
390 behaving confusingly. When first started, it used values for whoever
391 started the daemon, but after a SIGHUP it used the Exim user (because it
392 calls itself on a restart). I have changed the code so that it now always
395 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
396 message are rejected with the same error (e.g. no authentication or bad
397 sender address), and a DATA command is nevertheless sent (as can happen
398 with PIPELINING or a stupid MUA), the error message that was given to the
399 RCPT commands is included in the rejection of the DATA command. This is
400 intended to be helpful for MUAs that show only the final error to their
403 PH/42 Another patch from the Sieve maintainer.
405 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
406 Eximstats - Fixed some broken HTML links and added missing column headers
408 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
409 columns for Rejects, Temp Rejects, Ham, and Spam rows.
411 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
413 PH/43 Yet another patch from the Sieve maintainer.
415 PH/44 I found a way to check for a TCP/IP connection going away before sending
416 the response to the final '.' that terminates a message, but only in the
417 case where the client has not sent further data following the '.'
418 (unfortunately, this is allowed). However, in many cases there won't be
419 any further data because there won't be any more messages to send. A call
420 to select() can be used: if it shows that the input is "ready", there is
421 either input waiting, or the socket has been closed. An attempt to read
422 the next input character can distinguish the two cases. Previously, Exim
423 would have sent an OK response which the client would never have see.
424 This could lead to message repetition. This fix should cure that, at
425 least in a lot of common cases.
427 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
428 advertised in response to EHLO.
434 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
435 fixed by 4.65/MH/01 (is this a record?) are fixed:
437 (i) An empty string was always treated as zero by the numeric comparison
438 operators. This behaviour has been restored.
440 (ii) It is documented that the numeric comparison operators always treat
441 their arguments as decimal numbers. This was broken in that numbers
442 starting with 0 were being interpreted as octal.
444 While fixing these problems I realized that there was another issue that
445 hadn't been noticed. Values of message_size_limit (both the global option
446 and the transport option) were treated as octal if they started with 0.
447 The documentation was vague. These values are now always treated as
448 decimal, and I will make that clear in the documentation.
454 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
455 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
458 MH/01 Don't check that the operands of numeric comparison operators are
459 integers when their expansion is in "skipping" mode (fixes bug
460 introduced by 4.64-PH/07).
462 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
463 child addresses, Exim now panics and dies. Previously, because the count
464 is held in a short int, deliveries were likely to be lost. As such a
465 large number of recipients for a single message is ridiculous
466 (performance will be very, very poor), I have chosen to impose a limit
467 rather than extend the field.
473 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
474 leftover -K file (the existence of which was triggered by #402).
475 While we were at it, introduced process PID as part of the -K
476 filename. This should rule out race conditions when creating
479 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
480 processing considerably. Previous code took too long for large mails,
481 triggering a timeout which in turn triggers #401.
483 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
484 in the DK code in transports.c. sendfile() is not really portable,
485 hence the _LINUX specificness.
487 TF/01 In the add_headers option to the mail command in an Exim filter,
488 there was a bug that Exim would claim a syntax error in any
489 header after the first one which had an odd number of characters
492 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
493 callout verification, Exim cached a "reject" for the entire domain. This
494 is correct for most verifications, but it is not correct for a recipient
495 verification with use_sender or use_postmaster set, because in that case
496 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
497 case of MAIL FROM:<> rejection from other early rejections (e.g.
498 rejection of HELO). When verifying a recipient using a non-null MAIL
499 address, the cache is ignored if it shows MAIL FROM:<> rejection.
500 Whatever the result of the callout, the value of the domain cache is
501 left unchanged (for any other kind of callout, getting as far as trying
502 RCPT means that the domain itself is ok).
504 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
507 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
508 manifest itself as EPIPE rather than ECONNECT. When tidying away a
509 session, the daemon ignores ECONNECT errors and logs others; it now
510 ignores EPIPE as well.
512 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
513 (quoted-printable decoding).
515 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
516 later the small subsequent patch to fix an introduced bug.
518 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
520 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
522 PH/08 An error is now given if message_size_limit is specified negative.
524 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
525 to be given (somewhat) arbitrary names.
527 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
530 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
531 miscellaneous code fixes
533 PH/10 Added the log_reject_target ACL modifier to specify where to log
536 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
537 hostname. This is wrong, because it relates to the incoming message (and
538 probably the interface on which it is arriving) and not to the outgoing
539 callout (which could be using a different interface). This has been
540 changed to use the value of the helo_data option from the smtp transport
541 instead - this is what is used when a message is actually being sent. If
542 there is no remote transport (possible with a router that sets up host
543 addresses), $smtp_active_hostname is used.
545 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
546 tweaks were necessary in order to get it to work (see also 21 below):
547 (a) The code assumed that strncpy() returns a negative number on buffer
548 overflow, which isn't the case. Replaced with Exim's string_format()
550 (b) There were several signed/unsigned issues. I just did the minimum
551 hacking in of casts. There is scope for a larger refactoring.
552 (c) The code used strcasecmp() which is not a standard C function.
553 Replaced with Exim's strcmpic() function.
554 (d) The code set only $1; it now sets $auth1 as well.
555 (e) A simple test gave the error "authentication client didn't specify
556 service in request". It would seem that Dovecot has changed its
557 interface. Fortunately there's a specification; I followed it and
558 changed what the client sends and it appears to be working now.
560 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
563 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
564 address A is aliased to B and C, where B exists and C does not. Without
565 -v the output is "A verified" because verification stops after a
566 successful redirection if more than one address is generated. However,
567 with -v the child addresses are also verified. Exim was outputting "A
568 failed to verify" and then showing the successful verification for C,
569 with its parentage. It now outputs "B failed to verify", showing B's
570 parentage before showing the successful verification of C.
572 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
573 look up a TXT record in a specific list after matching in a combined
576 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
577 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
578 they consult the DNS. I had assumed they would set it the way they
579 wanted; and indeed my experiments on Linux seem to show that in some
580 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
581 To be on the safe side, however, I have now made the interface to
582 host_find_byname() similar to host_find_bydns(), with an argument
583 containing the DNS resolver options. The host_find_byname() function now
584 sets these options at its start, just as host_find_bydns() does. The smtp
585 transport options dns_qualify_single and dns_search_parents are passed to
586 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
587 of host_find_byname() use the default settings of RES_DEFNAMES
588 (qualify_single) but not RES_DNSRCH (search_parents).
590 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
591 spool_read_header() do less string testing, by means of a preliminary
592 switch on the second character of optional "-foo" lines. (This is
593 overdue, caused by the large number of possibilities that now exist.
594 Originally there were few.) While I was there, I also converted the
595 str(n)cmp tests so they don't re-test the leading "-" and the first
596 character, in the hope this might squeeze out yet more improvement.
598 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
599 flag allowing group syntax was set by the header_syntax check but not
600 turned off, possible causing trouble later; (2) The flag was not being
601 set at all for the header_verify test, causing "group"-style headers to
602 be rejected. I have now set it in this case, and also caused header_
603 verify to ignore an empty address taken from a group. While doing this, I
604 came across some other cases where the code for allowing group syntax
605 while scanning a header line wasn't quite right (mostly, not resetting
606 the flag correctly in the right place). These bugs could have caused
607 trouble for malformed header lines. I hope it is now all correct.
609 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
610 with the "reply" argument non-NULL. The code, however (which originally
611 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
612 but it didn't always do it. This confused somebody who was copying the
613 code for some other use. I have removed all the tests.
615 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
616 feature that was used to support insecure browsers during the U.S. crypto
617 embargo. It requires special client support, and Exim is probably the
618 only MTA that supported it -- and would never use it because real RSA is
619 always available. This code has been removed, because it had the bad
620 effect of slowing Exim down by computing (never used) parameters for the
621 RSA_EXPORT functionality.
623 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
624 authenticator to fail if there's a tab character in the incoming data
625 (there should never be unless someone is messing about, as it's supposed
626 to be base64-encoded). Also added, on Timo's advice, the "secured" option
627 if the connection is using TLS or if the remote IP is the same as the
628 local IP, and the "valid-client-cert option" if a client certificate has
631 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
632 authenticators. This can be used for authorization after authentication
633 succeeds. (In the case of plaintext, it servers for both authentication
636 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
637 if any retry times were supplied.
639 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
640 connection, where there is no client IP address to check. In this
641 situation, the verify now always succeeds.
643 PH/25 Applied John Jetmore's -Mset patch.
645 PH/26 Added -bem to be like -Mset, but loading a message from a file.
647 PH/27 In a string expansion for a processed (not raw) header when multiple
648 headers of the same name were present, leading whitespace was being
649 removed from all of them, but trailing whitespace was being removed only
650 from the last one. Now trailing whitespace is removed from each header
651 before concatenation. Completely empty headers in a concatenation (as
654 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
655 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
657 PH/29 [Removed. This was a change that I later backed out, and forgot to
658 correct the ChangeLog entry (that I had efficiently created) before
659 committing the later change.]
661 PH/30 Exim was sometimes attempting to deliver messages that had suffered
662 address errors (4xx response to RCPT) over the same connection as other
663 messages routed to the same hosts. Such deliveries are always "forced",
664 so retry times are not inspected. This resulted in far too many retries
665 for the affected addresses. The effect occurred only when there were more
666 hosts than the hosts_max_try setting in the smtp transport when it had
667 the 4xx errors. Those hosts that it had tried were not added to the list
668 of hosts for which the message was waiting, so if all were tried, there
669 was no problem. Two fixes have been applied:
671 (i) If there are any address or message errors in an SMTP delivery, none
672 of the hosts (tried or untried) are now added to the list of hosts
673 for which the message is waiting, so the message should not be a
674 candidate for sending over the same connection that was used for a
675 successful delivery of some other message. This seems entirely
676 reasonable: after all the message is NOT "waiting for some host".
677 This is so "obvious" that I'm not sure why it wasn't done
678 previously. Hope I haven't missed anything, but it can't do any
679 harm, as the worst effect is to miss an optimization.
681 (ii) If, despite (i), such a delivery is accidentally attempted, the
682 routing retry time is respected, so at least it doesn't keep
683 hammering the server.
685 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
686 in ${readsocket because some servers need this prod.
688 PH/32 Added some extra debug output when updating a wait-xxx database.
690 PH/33 The hint "could be header name not terminated by colon", which has been
691 given for certain expansion errors for a long time, was not being given
692 for the ${if def:h_colon_omitted{... case.
694 PH/34 The spec says: "With one important exception, whenever a domain list is
695 being scanned, $domain contains the subject domain." There was at least
696 one case where this was not true.
698 PH/35 The error "getsockname() failed: connection reset by peer" was being
699 written to the panic log as well as the main log, but it isn't really
700 panic-worthy as it just means the connection died rather early on. I have
701 removed the panic log writing for the ECONNRESET error when getsockname()
704 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
705 runs only) independently of the message's sender address. This meant
706 that, if the 4xx error was in fact related to the sender, a different
707 message to the same recipient with a different sender could confuse
708 things. In particualar, this can happen when sending to a greylisting
709 server, but other circumstances could also provoke similar problems.
710 I have changed the default so that the retry time for these errors is now
711 based a combination of the sender and recipient addresses. This change
712 can be overridden by setting address_retry_include_sender=false in the
715 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
716 remote server are returned as part of bounce messages. This was not
717 happening for LMTP over a pipe (the lmtp transport), but now it is the
718 same for both kinds of LMTP.
720 PH/38 Despite being documented as not happening, Exim was rewriting addresses
721 in header lines that were in fact CNAMEs. This is no longer the case.
723 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
724 and queue runs started by the daemon processed all messages. This has
725 been fixed so that -R and -S can now usefully be given with -q<time>.
727 PH/40 Import PCRE release 6.7 (fixes some bugs).
729 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
731 PH/42 Give an error if -q is specified more than once.
733 PH/43 Renamed the variables $interface_address and $interface_port as
734 $received_ip_address and $received_port, to make it clear that these
735 values apply to message reception, and not to the outgoing interface when
736 a message is delivered. (The old names remain recognized, of course.)
738 PH/44 There was no timeout on the connect() call when using a Unix domain
739 socket in the ${readsocket expansion. There now is.
741 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
742 be meaningful with "accept".
744 SC/01 Eximstats V1.43
745 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
747 SC/02 Eximstats V1.44
748 Use a glob alias rather than an array ref in the generated
749 parser. This improves both readability and performance.
751 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
752 Collect SpamAssassin and rejection statistics.
753 Don't display local sender or destination tables unless
754 there is data to show.
755 Added average volumes into the top table text output.
757 SC/04 Eximstats V1.46
758 Collect data on the number of addresses (recipients)
759 as well as the number of messages.
761 SC/05 Eximstats V1.47
762 Added 'Message too big' to the list of mail rejection
763 reasons (thanks to Marco Gaiarin).
765 SC/06 Eximstats V1.48
766 Mainlog lines which have GMT offsets and are too short to
767 have a flag are now skipped.
769 SC/07 Eximstats V1.49 (Alain Williams)
770 Added the -emptyok flag.
772 SC/08 Eximstats V1.50
773 Fixes for obtaining the IP address from reject messages.
775 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
776 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
777 whitesspace changes from 4.64-PH/27
779 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
782 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
783 are found, allow negative numbers in numeric criteria)
785 JJ/06 exipick.20061117.2, added new $message_body_missing variable
787 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
788 to match changes made in 4.64-PH/43
790 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
792 PH/47 Put in an explicit test for a DNS lookup of an address record where the
793 "domain" is actually an IP address, and force a failure. This locks out
794 those revolvers/nameservers that support "A-for-A" lookups, in
795 contravention of the specifications.
797 PH/48 When a host name was looked up from an IP address, and the subsequent
798 forward lookup of the name timed out, the host name was left in
799 $sender_host_name, contrary to the specification.
801 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
802 restricted to single-key lookups, Exim was not diagnosing an error if
803 * or *@ was used with a query-style lookup.
805 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
807 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
808 long ago, but noone interested enough thought of it. Let's just say that
809 the "1.1" means that there are some new functions that weren't there at
810 some point in the past.
812 PH/51 Error processing for expansion failure of helo_data from an smtp
813 transport during callout processing was broken.
815 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
816 tested/used via the -bh/-bhc/-bs options.
818 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
819 bug, fixed in subsequent PCRE releases).
821 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
822 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
824 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
830 SC/01 Use a glob alias rather than an array ref in eximstats generated
831 parser. This improves both readability and performance.
833 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
834 Don't display local sender or destination tables in eximstats unless
835 there is data to show.
836 Added average volumes into the eximstats top table text output.
838 SC/03 Collect data on the number of addresses (recipients) as well
839 as the number of messages in eximstats.
841 TF/01 Correct an error in the documentation for the redirect router. Exim
842 does (usually) call initgroups() when daemonizing.
844 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
845 with consistent privilege compared to when running as a daemon.
847 TF/03 Note in the spec that $authenticated_id is not set for local
848 submissions from trusted users.
850 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
851 Thanks to Dean Brooks <dean@iglou.com> for the patch.
853 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
854 by adding some example configuration directives to the default
855 configuration file. A little bit of work is required to uncomment the
856 directives and define how usernames and passwords are checked, but
857 there is now a framework to start from.
859 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
860 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
861 without this. I don't know how relevant this is to other LDAP libraries.
863 PH/02 Add the verb name to the "unknown ACL verb" error.
865 PH/03 Magnus Holmgren's patch for filter_prepend_home.
867 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
869 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
870 directory not expanded when it should be if an expanded home directory
871 was set for the address (which is overridden by the transport).
873 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
876 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
877 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
878 because it is too late at that time, and has no effect.
880 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
881 security issue with \' (bugzilla #107). I could not use the
882 PQescapeStringConn() function, because it needs a PGconn value as one of
885 PH/08 When testing addresses using -bt, indicate those final addresses that
886 are duplicates that would not cause an additional delivery. At least one
887 person was confused, thinking that -bt output corresponded to deliveries.
888 (Suppressing duplicates isn't a good idea as you lose the information
889 about possibly different redirections that led to the duplicates.)
891 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
892 systems where poll() doesn't work, in particular OS X.
894 PH/10 Added more information to debugging output for retry time not reached.
896 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
897 operations in malware.c.
899 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
902 PH/13 If write_rejectlog was set false when logging was sent to syslog with
903 syslog_duplication set false, log lines that would normally be written
904 both the the main log and to the reject log were not written to syslog at
907 PH/14 In the default configuration, change the use of "message" in ACL warn
908 statements to "add_header".
910 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
911 not followed by a command (e.g. "seen endif").
913 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
914 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
917 PH/17 Added extra conditions to the default value of delay_warning_condition
921 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
922 { match{$h_precedence:}{(?i)bulk|list|junk} } \
923 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
926 The Auto-Submitted: and various List- headers are standardised, whereas I
927 don't think Precedence: ever was.
929 PH/18 Refactored debugging code in route_finduser() to show more information,
930 in particular, the error code if getpwnam() issues one.
932 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
933 This is apparently needed in addition to the PH/07 change above to avoid
934 any possible encoding problems.
936 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
937 but not after initializing Perl.
939 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
940 output them only if debugging. By default they are written stderr,
941 apparently, which is not desirable.
943 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
946 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
949 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
951 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
952 authenticated or an ident call has been made. Suppress the default
953 values for $authenticated_id and $authenticated_sender (but permit -oMai
954 and -oMas) when testing with -bh.
956 PH/24 Re-jigged the order of the tests in the default configuration so that the
957 tests for valid domains and recipients precede the DNS black list and CSA
958 tests, on the grounds that those ones are more expensive.
960 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
961 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
962 This bug exists in every version of Exim that I still have, right back to
965 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
966 However, an attempt to turn on case-sensitivity in a regex key by
967 including (?-i) didn't work because the subject string was already
968 lowercased, and the effects were non-intuitive. It turns out that a
969 one-line patch can be used to allow (?-i) to work as expected.
975 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
976 other effects) broke the use of negated acl sub-conditions.
978 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
981 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
982 "Deny" causes Exim to reject the incoming connection with a 554 error.
983 Unfortunately, if there is a major crisis, such as a disk failure,
984 tcp-wrappers gives "deny", whereas what one would like would be some
985 kind of temporary error. A kludge has been added to help with this.
986 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
987 554 error is used if errno is still zero or contains ENOENT (which occurs
988 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
991 PH/03 Add -lutil to the default FreeBSD LIBS setting.
993 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
994 errors. Otherwise a message that provokes a temporary error (when other
995 messages do not) can cause a whole host to time out.
997 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
998 addresses were routed directly to files or pipes from a redirect router.
999 File deliveries just didn't batch; pipe deliveries might have suffered
1002 PH/06 A failure to get a lock for a hints database would erroneously always say
1003 "Failed to get write lock", even when it was really a read lock.
1005 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1006 of 0600. This has been changed to use the value of the lockfile_mode
1007 option (which defaults to 0600).
1009 PH/08 Applied small patch from the Sieve maintainer.
1011 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1012 folder from quota calculations, a direct delivery into this folder messed
1013 up the contents of the maildirsize file. This was because the regex was
1014 used only to exclude .Trash (or whatever) when the size of the mailbox
1015 was calculated. There was no check that a delivery was happening into an
1016 excluded directory. This bug has been fixed by ignoring all quota
1017 processing for deliveries into excluded directories.
1019 PH/10 Added the maildirfolder_create_regex option to appendfile.
1025 PH/01 The code for finding all the local interface addresses on a FreeBSD
1026 system running IPv6 was broken. This may well have applied to all BSD
1027 systems, as well as to others that have similar system calls. The broken
1028 code found IPv4 interfaces correctly, but gave incorrect values for the
1029 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1030 that it would not match correctly against @[] and not recognize the IPv6
1033 PH/02 The ipliteral router was not recognizing addresses of the form user@
1034 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1036 PH/03 Added disable_ipv6.
1038 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1039 decoded form, because it is most often used to construct To: headers
1040 lines in autoreplies, and the decoded form may well be syntactically
1041 invalid. However, $reply_address has leading white space removed, and all
1042 newlines turned into spaces so that the autoreply transport does not
1045 PH/05 If group was specified without a user on a router, and no group or user
1046 was specified on a transport, the group from the router was ignored.
1048 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1049 for visible compile-time settings that can be used to change these
1050 numbers, for those that want even more. Backwards compatibility with old
1051 spool files has been maintained. However, going back to a previous Exim
1052 release will lost any variables that are in spool files.
1054 PH/07 Two small changes when running in the test harness: increase delay when
1055 passing a TCP/IP connection to a new process, in case the original
1056 process has to generate a bounce, and remove special handling of
1057 127.0.0.2 (sic), which is no longer necessary.
1059 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1060 be the same on different OS.
1062 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1065 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1066 whether --show-vars was specified or not
1068 JJ/02 exipick: Added support for new ACL variable spool format introduced
1071 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1072 syntactically invalid From: or Reply-to: line, and a filter used this to
1073 generate an autoreply, and therefore failed to obtain an address for the
1074 autoreply, Exim could try to deliver to a non-existent relative file
1075 name, causing unrelated and misleading errors. What now happens is that
1076 it logs this as a hard delivery error, but does not attempt to create a
1079 PH/11 The exinext utility has a -C option for testing purposes, but although
1080 the given file was scanned by exinext itself; it wasn't being passed on
1081 when Exim was called.
1083 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1084 an end-of-file indication when reading a command response.
1086 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1087 compiled. In many other places in Exim, IPv6 addresses are always
1088 recognized, so I have changed this. It also means that IPv4 domain
1089 literals of the form [IPV4:n.n.n.n] are now always recognized.
1091 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1092 used if the router is not running as root, for example, when verifying at
1093 ACL time, or when using -bh. The debugging output from this situation was
1094 non-existent - all you got was a failure to exec. I have made two
1097 (a) Failures to set uid/gid, the current directory, or a process leader
1098 in a subprocess such as that created by queryprogram now generate
1099 suitable debugging ouput when -d is set.
1101 (b) The queryprogram router detects when it is not running as root,
1102 outputs suitable debugging information if -d is set, and then runs
1103 the subprocess without attempting to change uid/gid.
1105 PH/15 Minor change to Makefile for building test_host (undocumented testing
1108 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1109 additional section of a DNS packet that returns MX or SRV records.
1110 Instead, it always explicitly searches for A/AAAA records. This avoids
1111 major problems that occur when a DNS server includes only records of one
1112 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1113 fixed another bug: if SRV records were looked up and the corresponding
1114 address records were *not* found in the additional section, the port
1115 values from the SRV records were lost.
1117 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1118 using the correct key (the original address) when searching the retry
1119 rules in order to find which one to use for generating the retry hint.
1121 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1122 adding the default one. Similarly, if it contains a Reply-To: header, the
1123 errors_reply_to option, if set, is not used.
1125 PH/19 When calculating a retry time, Exim used to measure the "time since
1126 failure" by looking at the "first failed" field in the retry record. Now
1127 it does not use this if it is later than than the arrival time of the
1128 message. Instead it uses the arrival time. This makes for better
1129 behaviour in cases where some deliveries succeed, thus re-setting the
1130 "first failed" field. An example is a quota failure for a huge message
1131 when small messages continue to be delivered. Without this change, the
1132 "time since failure" will always be short, possible causing more frequent
1133 delivery attempts for the huge message than are intended.
1134 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1136 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1137 $1, $2, $3) because the numerical variables can be reset during some
1138 expansion items (e.g. "match"), thereby losing the authentication data.
1140 PH/21 Make -bV show the size of off_t variables so that the test suite can
1141 decide whether to run tests for quotas > 2G.
1143 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1144 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1145 filecount value is greater than 2G or if a quota value is greater than 2G
1146 on a system where the size of off_t is not greater than 4, a panic error
1149 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1150 never match. The debug and -bh output now contains an explicit error
1151 message indicating a malformed IPv4 address or mask.
1153 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1154 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1155 PH/23 above applies.
1157 PH/25 Do not write to syslog when running in the test harness. The only
1158 occasion when this arises is a failure to open the main or panic logs
1159 (for which there is an explicit test).
1161 PH/26 Added the /no_tell option to "control=freeze".
1163 PH/27 If a host name lookup failed very early in a connection, for example, if
1164 the IP address matched host_lookup and the reverse lookup yielded a name
1165 that did not have a forward lookup, an error message of the form "no IP
1166 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1167 could be logged. Now it outputs the IP address instead of "NULL".
1169 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1170 allows the sender and the authenticated sender to be set when
1171 submitting a message from within Exim. Since child_open_exim() is
1172 documented for local_scan(), the new function should be too.
1174 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1175 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1176 results in an empty string is now treated as unset.
1178 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1180 PH/31 Added sender_verify_fail logging option.
1182 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1183 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1184 tidied the source and removed it altogether.
1186 PH/33 When a queue run was abandoned because the load average was too high, a
1187 log line was always written; now it is written only if the queue_run log
1188 selector is set. In addition, the log line for abandonment now contains
1189 information about the queue run such as the pid. This is always present
1190 in "start" and "stop" lines but was omitted from the "abandon" line.
1192 PH/34 Omit spaces between a header name and the colon in the error message that
1193 is given when verify = headers_syntax fails (if there are lots of them,
1194 the message gets confusing).
1196 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1197 names, as there are now some PTR records that contain slashes. This check
1198 is only to protect against broken name servers that fall over on strange
1199 characters, so the fact that it applies to all lookups doesn't matter.
1201 PH/36 Now that the new test suite is complete, we can remove some of the
1202 special code in Exim that was needed for the old test suite. For example,
1203 sorting DNS records because real resolvers return them in an arbitrary
1204 order. The new test suite's fake resolver always returns records in the
1207 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1208 bounces) except when queue_only is set, to avoid logging races between
1209 the different processes.
1211 PH/38 Panic-die if .include specifies a non-absolute path.
1213 PH/39 A tweak to the "H" retry rule from its user.
1215 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1216 a label. They prevented compilation on older perls.
1218 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1219 a warning to be raised on newish perls.
1221 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1222 on queue. Changes to match documented behaviour of showing count of
1223 messages matching specified criteria.
1225 PH/40 Changed the default ident timeout from 30s to 5s.
1227 PH/41 Added support for the use of login_cap features, on those BSD systems
1228 that have them, for controlling the resources used by pipe deliveries.
1230 PH/42 The content-scanning code uses fopen() to create files in which to put
1231 message data. Previously it was not paying any attention to the mode of
1232 the files. Exim runs with umask(0) because the rest of the code creates
1233 files with open(), and sets the required mode explicitly. Thus, these
1234 files were ending up world-writeable. This was not a big issue, because,
1235 being within the spool directory, they were not world-accessible. I have
1236 created a function called modefopen, which takes an additional mode
1237 argument. It sets umask(777), creates the file, chmods it to the required
1238 mode, then resets the umask. All the relevant calls to fopen() in the
1239 content scanning code have been changed to use this function.
1241 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1242 to 24 hours. This avoids potential overflow problems when processing G
1243 and H retry rules. I suspect nobody ever tinkers with this value.
1245 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1247 PH/45 When the plaintext authenticator is running as a client, the server's
1248 challenges are checked to ensure they are valid base64 strings. By
1249 default, the authentication attempt is cancelled if an invalid string is
1250 received. Setting client_ignore_invalid_base64 true ignores these errors.
1251 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1252 they are received. Thus, the responses can be made to depend on the
1253 challenges. If an invalid string is ignored, an empty string is placed in
1256 PH/46 Messages that are created by the autoreply transport now contains a
1257 References: header, in accordance with RFCs 2822 and 3834.
1259 PH/47 Added authenticated_sender_force to the smtp transport.
1261 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1263 PH/49 Installed latest patch from the Sieve maintainer.
1265 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1266 was also set, the appendfile transport was unnecessarily scanning a
1267 directory of message files (e.g. for maildir delivery) to find the count
1268 of files (along with the size), even though it did not need this
1269 information. It now does the scan only if it needs to find either the
1270 size of the count of files.
1272 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1274 PH/52 Two bugs concerned with error handling when the smtp transport is
1277 (i) Exim was not creating retry information for temporary errors given
1278 for individual recipients after the DATA command when the smtp transport
1279 was used in LMTP mode. This meant that they could be retried too
1280 frequently, and not timed out correctly.
1282 (ii) Exim was setting the flag that allows error details to be returned
1283 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1284 recipients that were returned after the DATA command.
1286 PH/53 This is related to PH/52, but is more general: for any failing address,
1287 when detailed error information was permitted to be returned to the
1288 sender, but the error was temporary, then after the final timeout, only
1289 "retry timeout exceeded" was returned. Now it returns the full error as
1290 well as "retry timeout exceeded".
1292 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1293 do this, and (what is worse) MTAs that accept it.
1295 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1296 will now be deprecated.
1298 PH/56 New os.c-cygwin from the Cygwin maintainer.
1300 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1301 formats (previously only available in exim formats via -bpr, -bpru,
1302 and -bpra. Now also available in native and exiqgrep formats)
1304 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1305 with very large, slow to parse queues
1307 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1309 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1311 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1312 responses to authentication challenges, though it was showing the
1313 challenges; (ii) I've removed the CR characters from the debug output for
1316 PH/58 Allow for the insertion of a newline as well as a space when a string
1317 is turned into more than one encoded-word during RFC 2047 encoding. The
1318 Sieve code now uses this.
1320 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1321 data_4xx, lost_connection, tls_required.
1323 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1324 message was being sent as an SMTP response.
1326 PH/61 Add -l and -k options to exicyclog.
1328 PH/62 When verifying, if an address was redirected to one new address, so that
1329 verification continued, and the new address failed or deferred after
1330 having set something in $address_data, the value of $address_data was not
1331 passed back to the ACL. This was different to the case when no
1332 redirection occurred. The value is now passed back in both cases.
1334 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1335 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1336 use login_cap.h, so on its own it isn't the distinguishing feature. The
1337 new name refers directly to the setclassresources() function.
1339 PH/65 Added configuration files for NetBSD3.
1341 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1343 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1344 is preferred over IPv4.
1346 PH/68 The bounce_return_message and bounce_return_body options were not being
1347 honoured for bounces generated during the reception of non-SMTP messages.
1348 In particular, this applied to messages rejected by the ACL. This bug has
1349 been fixed. However, if bounce_return_message is true and bounce_return_
1350 body is false, the headers that are returned for a non-SMTP message
1351 include only those that have been read before the error was detected.
1352 (In the case of an ACL rejection, they have all been read.)
1354 PH/69 The HTML version of the specification is now built in a directory called
1355 spec_html instead of spec.html, because the latter looks like a path with
1356 a MIME-type, and this confuses some software.
1358 PH/70 Catch two compiler warnings in sieve.c.
1360 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1361 function verify_get_ident() calls ip_connect() to connect a socket, but
1362 if the "connect()" function timed out, ip_connect() used to close the
1363 socket. However, verify_get_ident() also closes the socket later, and in
1364 between Exim writes to the log, which may get opened at this point. When
1365 the socket was closed in ip_connect(), the log could get the same file
1366 descriptor number as the socket. This naturally causes chaos. The fix is
1367 not to close the socket in ip_connect(); the socket should be closed by
1368 the function that creates it. There was only one place in the code where
1369 this was missing, in the iplookup router, which I don't think anybody now
1370 uses, but I've fixed it anyway.
1372 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1373 well as to direct DNS lookups. Otherwise the handling of names in host
1374 lists is inconsistent and therefore confusing.
1380 PH/01 Two changes to the default runtime configuration:
1382 (1) Move the checks for relay_from_hosts and authenticated clients from
1383 after to before the (commented out) DNS black list checks.
1385 (2) Add control=submission to the relay_from_hosts and authenticated
1386 clients checks, on the grounds that messages accepted by these
1387 statements are most likely to be submissions.
1389 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1391 (1) Generate an error if the third argument for the ${prvs expansion is
1394 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1397 (3) Reset the variables that are obtained from the first argument of
1398 ${prvscheck and used in the second argument before leaving the code,
1399 because their memory is reclaimed, so using them afterwards may do
1402 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1403 one (it's much easier than Tom thought :-).
1405 (5) Because of (4), we can now allow for the use of $prvscheck_result
1406 inside the third argument.
1408 PH/03 For some reason, the default setting of PATH when running a command from
1409 a pipe transport was just "/usr/bin". I have changed it to
1412 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1413 anything to be listed in the output from -bV.
1415 PH/05 When a filter generated an autoreply, the entire To: header line was
1416 quoted in the delivery log line, like this:
1418 => >A.N.Other <ano@some.domain> <original@ddress> ...
1420 This has been changed so that it extracts the operative address. There
1421 may be more than one such address. If so, they are comma-separated, like
1424 => >ano@some.domain,ona@other.domain <original@ddress> ...
1426 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1427 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1428 not being looked up in the rDNS to get a host name, Exim was showing the
1429 IP address twice in Received: lines, even though the IP addresses were
1430 identical. For example:
1432 Received: from [1.2.3.4] (helo=[1.2.3.4])
1434 However, if the real host name was known, it was omitting the HELO data
1435 if it matched the actual IP address. This has been tidied up so that it
1436 doesn't show the same IP address twice.
1438 PH/07 When both +timestamp and +memory debugging was on, the value given by
1439 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1440 called by the debug printing, thereby overwriting the timestamp buffer.
1441 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1443 PH/08 When the original message was included in an autoreply transport, it
1444 always said "this is a copy of the message, including all the headers",
1445 even if body_only or headers_only was set. It now gives an appropriate
1448 PH/09 Applied a patch from the Sieve maintainer which:
1450 o fixes some comments
1451 o adds the (disabled) notify extension core
1452 o adds some debug output for the result of if/elsif tests
1453 o points to the current vacation draft in the documentation
1454 and documents the missing references header update
1458 o fixes a bug in processing the envelope test (when testing
1459 multiple envelope elements, the last element determinted the
1462 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1463 Electronic Mail") by including:
1465 Auto-submitted: auto-generated
1467 in the messages that it generates (bounce messages and others, such as
1468 warnings). In the case of bounce messages for non-SMTP mesages, there was
1469 also a typo: it was using "Auto_submitted" (underscore instead of
1470 hyphen). Since every message generated by Exim is necessarily in response
1471 to another message, thes have all been changed to:
1473 Auto-Submitted: auto-replied
1475 in accordance with these statements in the RFC:
1477 The auto-replied keyword:
1479 - SHOULD be used on messages sent in direct response to another
1480 message by an automatic process,
1482 - MUST NOT be used on manually-generated messages,
1484 - MAY be used on Delivery Status Notifications (DSNs) and Message
1485 Disposition Notifications (MDNs),
1487 - MUST NOT be used on messages generated by automatic or periodic
1488 processes, except for messages which are automatic responses to
1491 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1492 to the default Received: header definition.
1494 PH/12 Added log selector acl_warn_skipped (default on).
1496 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1497 variables because (a) they are in the wrong storage pool and (b) even if
1498 they were copied, it wouldn't work properly because of the caching.
1500 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1501 checking when decoding. Apparently there are clients that generate
1502 overlong encoded strings. Why am I not surprised?
1504 PH/15 If the first argument of "${if match_address" was not empty, but did not
1505 contain an "@" character, Exim crashed. Now it writes a panic log message
1506 and treats the condition as false.
1508 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1510 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1511 extension "envelope-auth". The code is finished and in agreement with
1512 other implementations, but there is no documentation so far and in fact,
1513 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1514 not changing the active code.
1516 Print executed "if" and "elsif" statements when debugging is used. This
1517 helps a great deal to understand what a filter does.
1519 Document more things not specified clearly in RFC3028. I had all this
1520 sorted out, when out of a sudden new issues came to my mind. Oops."
1522 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1525 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1526 canonical form (as documented). However, after a host name lookup from
1527 the IP address, check_host() was doing a simple string comparison with
1528 addresses acquired from the DNS when checking that the found name did
1529 have the original IP as one of its addresses. Since any found IPv6
1530 addresses are likely to be in abbreviated form, the comparison could
1531 fail. Luckily, there already exists a function for doing the comparison
1532 by converting both addresses to binary, so now that is used instead of
1533 the text comparison.
1535 PH/20 There was another similar case to PH/19, when a complete host name was
1536 given in a host list; looking up its IP address could give an abbreviated
1537 form, whereas the current host's name might or might not be abbreviated.
1538 The same fix has been applied.
1544 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1545 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1548 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1549 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1551 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1553 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1554 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1555 Also, for "Auto-Submitted", treat anything other than "no" as
1556 non-personal, in accordance with RFC 3834. (Previously it treated
1557 anything starting "auto-" as non-personal.)
1559 TF/01 The control=submission/name=... option had a problem with syntax
1560 errors if the name included a slash character. The /name= option
1561 now slurps the rest of the string, so it can include any characters
1562 but it must come last in the list of options (after /sender_retain
1565 PH/05 Some modifications to the interface to the fake nameserver for the new
1573 TK/01 Added the "success_on_redirect" address verification option. See
1574 NewStuff for rationale and an example.
1576 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1578 PH/02 Patch to exigrep to allow it to work on syslog lines.
1580 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1581 fread() to skip over the body file's header line, because in Cygwin the
1582 header line is locked and is inaccessible.
1584 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1585 co-exist for some time) to make it clear that it is the Exim ID that is
1586 referenced, not the Message-ID: header line.
1588 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1589 string_format() function, because snprintf() does not exist on all
1592 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1593 ${stat: expansion item.
1595 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1596 protocol synchronization error", to keep the pedants happy.
1598 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1599 well as for IRIX systems, when gcc is being used. See the host.c source
1602 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1604 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1607 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1608 transport and to the smtp transport in LMTP mode.
1610 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1612 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1613 run a filter in a subprocess. This could lead to confusion in subsequent
1614 lookups in the parent process. There should also be a search_tidyup() at
1615 the end of the subprocess.
1617 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1618 only if the host matched helo_try_verify_hosts, which caused the
1619 verification to occur when the EHLO/HELO command was issued. The ACL just
1620 tested the remembered result. Now, if a previous verification attempt has
1621 not happened, "verify = helo" does it there and then.
1623 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1625 TK/03 Fix log output including CR from clamd.
1627 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1628 exist provoked a memory error which could cause a segfault.
1630 PH/15 Installed PCRE 6.2
1632 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1634 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1635 of the problem. Specifically, suggested +O2 rather than +O1 for the
1638 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1640 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1641 record, the port was not correctly logged when the outgoing_port log
1642 selector was set (it logged the transort's default port).
1644 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1645 fallback_hosts, and "hosts" in the smtp transport.
1647 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1648 host errors such as "Connection refused".
1650 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1651 authentication with radiusclient 0.4.9:
1653 - Error returned from rc_read_config was caught wrongly
1654 - Username/password not passed on to radius server due to wrong length.
1656 The presumption is that some radiusclient API changes for 4.51/PH/17
1657 were not taken care of correctly. The code is still untested by me (my
1658 Linux distribution still has 0.3.2 of radiusclient), but it was
1659 contributed by a Radius user.
1661 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1662 expanding the "port" option of the smtp transport.
1664 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1665 while reading a MIME header. Thanks to Tom Hughes for a patch.
1667 PH/24 Include config.h inside local_scan.h so that configuration settings are
1670 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1671 that in an ACL for RCPT (for example), you can examine exactly what was
1674 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1675 commands, but it was not correctly comparing the address with the actual
1676 client host address. Thus, it would show the EHLO address in Received:
1677 header lines when this was not necessary.
1679 PH/27 Added the % operator to ${eval:}.
1681 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1682 it should be ignoring failures (because with -C, for example, it has lost
1683 privilege). It wasn't ignoring creation failures other than "already
1686 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1689 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1690 because an input line was too long, either on its own, or by virtue of
1691 too many continuations, the temporary file was not being removed, and the
1692 return code was incorrect.
1694 PH/31 Missing "BOOL" in function definition in filtertest.c.
1696 PH/32 Applied Sieve patches from the maintainer.
1698 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1700 PH/33 Added "verify = not_blind".
1702 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1703 Local/Makefile (with some defaults set). These are used in built scripts
1704 such as exicyclog, but they have never been used in the exim_install
1705 script (though there are many overriding facilities there). I have
1706 arranged that the exim_install script now takes note of these two
1709 PH/35 Installed configuration files for Dragonfly.
1711 PH/36 When a locally submitted message by a trusted user did not contain a
1712 From: header, and the sender address was obtained from -f or from an SMTP
1713 MAIL command, and the trusted user did not use -F to supply a sender
1714 name, $originator_name was incorrectly used when constructing a From:
1715 header. Furthermore, $originator_name was used for submission mode
1716 messages from external hosts without From: headers in a similar way,
1717 which is clearly wrong.
1719 PH/37 Added control=suppress_local_fixups.
1721 PH/38 When log_selector = +received_sender was set, and the addition of the
1722 sender made the log line's construction buffer exactly full, or one byte
1723 less than full, an overflow happened when the terminating "\n" was
1726 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1727 when the result of a list match is failure because a DNS lookup failed.
1729 PH/40 RM_COMMAND is now used in the building process.
1731 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1732 the "build-* directories that it finds.
1734 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1735 address in a domain literal was a prefix of an interface address.
1737 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1738 when verifying a sender address, unless rewrite_headers is false.
1740 PH/44 Wrote a long comment about why errors_to addresses are verified as
1741 recipients, not senders.
1743 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1744 the ratelimit ACL was added.
1746 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1748 PH/46 Added extra information about PostgreSQL errors to the error string.
1750 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1751 avoiding the need to install special zones in a real server. This is
1752 backwards compatible; if it can't find the fake resolver, it drops back.
1753 Thus, both old and new test suites can be run.
1755 TF/02 Added util/ratelimit.pl
1757 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1758 clock is set back in time.
1760 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1761 Candler <B.Candler@pobox.com>.
1763 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1764 OK for addresses that are the result of redirections.
1766 PH/48 A number of further additions for the benefit of the new test suite,
1767 including a fake gethostbyname() that interfaces to the fake DNS resolver
1770 TF/06 The fix for widen_domains has also been applied to qualify_single and
1771 search_parents which are the other dnslookup options that can cause
1774 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1777 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1779 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1780 service" if it encountered an MX record with an empty target hostname.
1781 The message is now "an MX or SRV record indicated no SMTP service".
1783 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1784 if the DNS of the sending site is misconfigured. This is quite a
1785 common situation. This change restores the behaviour of treating a
1786 helo verification defer as a failure.
1788 PH/51 If self=fail was set on a router, the bounce message did not include the
1789 actual error message.
1795 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1797 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1798 command itself ended in error, the underlying message about the transport
1799 filter timeout was being overwritten with the pipe command error. Now the
1800 underlying error message should be appended to the second error message.
1802 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1804 PH/02 Reduce the amount of output that "make" produces by default. Full output
1805 can still be requested.
1807 PH/03 The warning log line about a condition test deferring for a "warn" verb
1808 was being output only once per connection, rather than after each
1809 occurrence (because it was using the same function as for successful
1810 "warn" verbs). This seems wrong, so I have changed it.
1812 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1813 it should not have, which might have caused a crash in the right
1814 circumstances, but probably never did.
1816 PH/04 Installed a modified version of Tony Finch's patch to make submission
1817 mode fix the return path as well as the Sender: header line, and to
1818 add a /name= option so that you can make the user's friendly name appear
1821 TF/03 Added the control = fakedefer ACL modifier.
1823 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1824 Mark Lowes for thorough testing.
1826 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1828 TK/03 Merged latest SRS patch from Miles Wilton.
1830 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1831 with the definition in sysexits.h (which is #included earlier).
1832 Fortunately, Exim does not actually use EX_OK. The code used to try to
1833 preserve the sysexits.h value, by assumimg that macro definitions were
1834 scanned for macro replacements. I have been disabused of this notion,
1835 so now the code just undefines EX_OK before #including unistd.h.
1837 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1838 in the smtp transport. When a block could not be written in a single
1839 write() function, the timeout was being re-applied to each part-write.
1840 This seems wrong - if the receiver was accepting one byte at a time it
1841 would take for ever. The timeout is now adjusted when this happens. It
1842 doesn't have to be particularly precise.
1844 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1845 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1847 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1848 without a domain if the check to <postmaster@domain> fails.
1850 SC/01 Eximstats: added -xls and the ability to specify output files
1851 (patch written by Frank Heydlauf).
1853 SC/02 Eximstats: use FileHandles for outputing results.
1855 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1857 SC/04 Eximstats: fixed display of large numbers with -nvr option
1859 SC/05 Eximstats: fixed merging of reports with empty tables.
1861 SC/06 Eximstats: added the -include_original_destination flag
1863 SC/07 Eximstats: removed tabs and trailing whitespace.
1865 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1867 TK/06 MBOX spool code: Add real "From " MBOX separator line
1868 so the .eml file is really in mbox format (even though
1869 most programs do not really care). Patch from Alex Miller.
1871 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1872 The latter is generated from $received_to and is only set if the
1873 message has one envelope recipient. SA can use these headers,
1874 obviously out-of-the-box. Patch from Alex Miller.
1876 PH/08 The ${def test on a variable was returning false if the variable's
1877 value was "0", contrary to what the specification has always said!
1878 The result should be true unless the variable is empty.
1880 PH/09 The syntax error of a character other than { following "${if
1881 def:variable_name" (after optional whitespace) was not being diagnosed.
1882 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1883 accidental colon was present, for example, could give incorrect results.
1885 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1886 result is used (not including appendfile, where other changes are about
1889 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1890 This involved changing a lot of size variables from int to off_t. It
1891 should work with maildirs and everything.
1893 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1894 spamd dying while we are connected to it.
1896 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1899 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1900 function for BDB 4.3.
1902 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1904 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1905 the sources for PCRE have been re-arranged and more files are now
1908 PH/15 The code I had for printing potentially long long variables in PH/11
1909 above was not the best (it lost precision). The length of off_t variables
1910 is now inspected at build time, and an appropriate printing format (%ld
1911 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1912 to be "long long int" or "long int". This is needed for the internal
1913 formatting function string_vformat().
1915 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1916 the configuration file to be ":syslog", then the script "guesses" where
1917 the logs files are, rather than using the compiled in default. In our
1918 case the guess is not the same as the compiled default, so the script
1919 suddenly stopped working when I started to use syslog. The patch checks
1920 to see if log_file_path is "". If so, it attempts to read it from exim
1921 with no configuration file to get the compiled in version, before it
1922 falls back to the previous guessing code."
1924 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1925 implementing BATV in an Exim configuration. See NewStuff for the gory
1928 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1929 Makefile that are specific to HP-UX.
1931 PH/18 If the "use_postmaster" option was set for a recipient callout together
1932 with the "random" option, the postmaster address was used as the MAIL
1933 FROM address for the random test, but not for the subsequent recipient
1934 test. It is now used for both.
1936 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1937 patch removes a few documentation additions to RFC 3028, because the
1938 latest draft now contains them. It adds the new en;ascii-case comparator
1939 and a new error check for 8bit text in MIME parts. Comparator and
1940 require names are now matched exactly. I enabled the subaddress
1941 extension, but it is not well tested yet (read: it works for me)."
1943 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1944 rework some of the code of TK/09 above to avoid the hardwired use of
1945 "%lld" and "long long". Replaced the call to snprintf() with a call to
1948 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1949 records point to non-existent hosts", "retry timeout exceeded", and
1950 "retry time not reached for any host after a long failure period".
1952 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1953 experimental DomainKeys support:
1955 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1956 (2) On an error such as an illegally used "control", the wrong name for
1957 the control was given.
1959 These problems did NOT occur unless DomainKeys support was compiled.
1961 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1963 PH/24 Added ${if match_ip condition.
1965 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1966 databases so that it will be absolutely obvious if a crash occurs in the
1967 DB library. This is a regular occurrence (often caused by mis-matched
1970 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1971 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1972 picked up on a user's system that detects such things. There doesn't seem
1973 to be a gcc warning option for this - only an attribute that has to be
1974 put on the function's prototype. It seems that in Fedora Core 4 they have
1975 set this on a number of new functions. No doubt there will be more in due
1978 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1979 specify a transport. However, if an address that was verified by such a
1980 router was the subject of a callout, Exim crashed because it tried to
1981 read the rcpt_include_affixes from the non-existent transport. Now it
1982 just assumes that the setting of that option is false. This bug was
1983 introduced by 4.51/PH/31.
1985 PH/28 Changed -d+all to exclude +memory, because that information is very
1986 rarely of interest, but it makes the output a lot bigger. People tend to
1987 do -d+all out of habit.
1989 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1990 code in os-type was giving problems when libc.so lives in lib64, like on
1993 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1994 aren't the modern standard, and it seems that some systems' include files
1995 don't always have them. Exim was already checking for some of the newer
1996 ones like T_AAAA, and defining it itself. I've added checks for all the
1997 record types that Exim uses.
1999 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2000 not automatically generating a new one, as it is supposed to. This
2001 prevented TLS from working. If the file did exist, but contained invalid
2002 data, a new version was generated, as expected. It was only the case of a
2003 non-existent file that was broken.
2005 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2006 with a change in libdomainkeys > 0.64.
2008 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2009 from DNS. If the selector record carries the flag, it now has
2010 precedence over the domain-wide flag.
2012 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2014 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2015 the use of a port name that isn't defined in /etc/services) occurred, the
2016 message was deferred as in a normal delivery, and thus remained on the
2017 spool, instead of being failed because of the mua_wrapper setting. This
2018 is now fixed, and I tidied up some of the mua_wrapper messages at the
2021 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2022 the messages in a hash of arrays rather than using individual hashes.
2023 This is a bit cleaner and results in dramatic memory savings, albeit
2024 at a slight CPU cost.
2026 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2027 as requested by Marc Sherman.
2029 SC/10 Eximstats: added histograms for user specified patterns as requested
2032 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2034 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2035 fopen() in the content-scanning modules that did not already have it.
2041 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2042 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2044 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2046 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2048 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2049 to test Sieve filters that use "vacation".
2051 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2052 that changes the way the GnuTLS parameters are stored in the cache file.
2053 The new format can be generated externally. For backward compatibility,
2054 if the data in the cache doesn't make sense, Exim assumes it has read an
2055 old-format file, and it generates new data and writes a new file. This
2056 means that you can't go back to an older release without removing the
2059 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2060 work if there are any delivery delays because "one_time" forces the
2061 parent to be marked "delivered", so its unseen clone is never tried
2062 again. For this reason, Exim now forbids the simultaneous setting of
2065 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2066 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2067 entry if you want to know the details. The fix, however, neglected to
2068 consider the case when local delivery batching is involved. The test for
2069 "previously delivered" was not happening when checking to see if an
2070 address could be batched with a previous (undelivered) one; under
2071 certain circumstances this could lead to multiple deliveries to the same
2074 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2075 in its include files, and this causes problems building Exim.
2077 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2078 header_syntax) but Exim was just ignoring anything given after a slash.
2079 In particular, this caused confusion with an attempt to use "verify =
2080 reverse_host_lookup/defer_ok". An error is now given when options are
2081 supplied for verify items that do not have them. (Maybe reverse_host_
2082 lookup should have a defer_ok option, but that's a different point.)
2084 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2085 defined by RFC 821) to 2048, because there were problems with some AUTH
2086 commands, and RFC 1869 says the size should be increased for extended
2087 SMTP commands that take arguments.
2089 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2092 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2093 "unknown" error; now it says that the functionality isn't in the binary.
2095 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2096 an address' error message when a string expansion fails (syntax or
2097 whatever). Otherwise the password may appear in the log. Following change
2098 PH/42 below, there is no longer a chance of it appearing in a bounce
2101 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2103 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2104 of its addresses was ever tried. (Bugzilla bug #2.)
2106 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2107 the result incorrectly in the debug output. (It correctly added a newline
2108 to what was transported.)
2110 TF/01 Added $received_time.
2112 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2113 commented out examples of how to interface to a virus scanner and to
2114 SpamAssassin. Also added commented examples of av_scanner and
2115 spamd_address settings.
2117 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2118 and controls are allowed in which ACLs. There were a couple of minor
2119 errors. Some of the entries in the conditions table (which is a table of
2120 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2121 negation of where the condition IS allowed.
2123 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2125 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2126 header file does not have a version number, so I've had to invent a new
2127 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2128 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2129 radiusclient), but it was contributed by a Radius user.
2131 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2132 files or directories, for OpenSSL.
2134 PH/19 When an Exim process that is running as root has to create an Exim log
2135 file, it does so in a subprocess that runs as exim:exim so as to get the
2136 ownership right at creation (otherwise, other Exim processes might see
2137 the file with the wrong ownership). There was no test for failure of this
2138 fork() call, which would lead to the process getting stuck as it waited
2139 for a non-existent subprocess. Forks do occasionally fail when resources
2140 run out. I reviewed all the other calls to fork(); they all seem to check
2143 PH/20 When checking for unexpected SMTP input at connect time (before writing
2144 the banner), Exim was not dealing correctly with a non-positive return
2145 from the read() function. If the client had disconnected by this time,
2146 the result was a log entry for a synchronization error with an empty
2147 string after "input=" when read() returned zero. If read() returned -1
2148 (an event I could not check), uninitialized data bytes were printed.
2149 There were reports of junk text (parts of files, etc) appearing after
2152 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2154 PH/22 Added support for macro redefinition, and (re)definition in between
2155 driver and ACL definitions.
2157 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2158 forgetting to use the resulting value; it was using the unexpanded value.
2160 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2161 hadn't been configured. The fix is from Juergen Kreileder, who
2162 understands it better than I do:
2164 "Here's what I see happening with three configured cyrus_sasl
2165 authenticators configured (plain, login, cram-md5):
2167 On startup auth_cyrus_sasl_init() gets called for each of these.
2168 This means three calls to sasl_listmech() without a specified mech_list.
2169 => SASL tests which mechs of all available mechs actually work
2170 => three warnings about OTP not working
2171 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2173 With the patch, sasl_listmech() also gets called three times. But now
2174 SASL's mech_list option is set to the server_mech specified in the the
2175 authenticator. Or in other words, the answer from sasl_listmech()
2176 gets limited to just the mech you're testing for (which is different
2178 => the return list contains just 'plain' or 'login', 'cram-md5' or
2179 nothing depending on the value of ob->server_mech.
2181 I've just tested the patch: Authentication still works fine,
2182 unavailable mechs specified in the exim configuration are still
2183 caught, and the auth.log warnings about OTP are gone."
2185 PH/25 When debugging is enabled, the contents of the command line are added
2186 to the debugging output, even when log_selector=+arguments is not
2189 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2190 answer is "GNU", and only if the return is "GNU/something" is the answer
2193 PH/27 $acl_verify_message is now set immediately after the failure of a
2194 verification in an ACL, and so is available in subsequent modifiers. In
2195 particular, the message can be preserved by coding like this:
2197 warn !verify = sender
2198 set acl_m0 = $acl_verify_message
2200 Previously, $acl_verify_message was set only while expanding "message"
2201 and "log_message" when a very denied access.
2203 PH/28 Modified OS/os.c-Linux with
2205 -#ifndef OS_LOAD_AVERAGE
2206 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2208 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2209 nomenclature these days.)
2211 PH/29 Installed patch from the Sieve maintainer that adds the options
2212 sieve_useraddress and sieve_subaddress to the redirect router.
2214 PH/30 In these circumstances:
2215 . Two addresses routed to the same list of hosts;
2216 . First host does not offer TLS;
2217 . First host accepts first address;
2218 . First host gives temporary error to second address;
2219 . Second host offers TLS and a TLS session is established;
2220 . Second host accepts second address.
2221 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2222 and peerdn, if requested) that were in fact used only for the second
2225 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2226 attention to any local part prefix or suffix that was matched by the
2227 router that accepted the address. It now behaves in the same way as it
2228 does for delivery: the affixes are removed from the local part unless
2229 rcpt_include_affixes is set on the transport.
2231 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2232 timeout during the DATA phase of an incoming message.
2234 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2235 applied a patch sanctioned by the Sieve maintainer.
2237 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2238 the uid or gid is negative. A case of a negative gid caused this to be
2239 noticed. The fix allows for either to be negative.
2241 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2242 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2245 PH/36 The change PH/12 above was broken. Fixed it.
2247 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2248 the grounds that routing the same address twice would always produce the
2249 same answer. This might have been true once, but it is certainly no
2250 longer true now. Routing a child address may depend on the previous
2251 routing that produced that child. Some complicated redirection strategies
2252 went wrong when messages had multiple recipients, and made Exim's
2253 behaviour dependent on the order in which the addresses were given.
2255 I have moved the duplicate checking until after the routing is complete.
2256 Exim scans the addresses that are assigned to local and remote
2257 transports, and removes any duplicates. This means that more work will be
2258 done, as duplicates will always all be routed, but duplicates are
2259 presumably rare, so I don't expect this is of any significance.
2261 For deliveries to pipes, files, and autoreplies, the duplicate checking
2262 still happens during the routing process, since they are not going to be
2265 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2266 It corrects a timeout issue with spamd. This is Ian's comment: "The
2267 background is that sometimes spamd either never reads data from a
2268 connection it has accepted, or it never writes response data. The exiscan
2269 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2270 blindly assumes that writes won't block so it may never time out."
2272 PH/39 Allow G after quota size as well as K and M.
2274 PH/40 The value set for $authenticated_id in an authenticator may not contain
2275 binary zeroes or newlines because the value is written to log lines and
2276 to spool files. There was no check on this. Now the value is run through
2277 the string_printing() function so that such characters are converted to
2278 printable escape sequences.
2280 PH/41 $message_linecount is a new variable that contains the total number of
2281 lines in the message. Compare $body_linecount, which is the count for the
2284 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2285 bounce and delay warning messages, except in certain special cases, which
2288 (a) An SMTP error message from a remote host;
2289 (b) A message specified in a :fail: redirection;
2290 (c) A message specified in a "fail" command in a system filter;
2291 (d) A message specified in a FAIL return from the queryprogram router;
2292 (e) A message specified by the cannot_route_message router option.
2294 In these cases only, Exim does include the error details in bounce and
2295 warning messages. There are also a few cases where bland messages such
2296 as "unrouteable address" or "local delivery error" are given.
2298 PH/43 $value is now also set for the "else" part of a ${run expansion.
2300 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2301 being worked on, but at least Exim now implements the latest version to
2304 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2305 process to complete was treated as a delivery failure, a timeout while
2306 writing the message to the pipe was logged, but erroneously treated as a
2307 successful delivery. Such timeouts include transport filter timeouts. For
2308 consistency with the overall process timeout, these timeouts are now
2309 treated as errors, giving rise to delivery failures by default. However,
2310 there is now a new Boolean option for the pipe transport called
2311 timeout_defer, which, if set TRUE, converts the failures into defers for
2312 both kinds of timeout. A transport filter timeout is now identified in
2315 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2316 systems where "make" and "gmake" are different, calling "gmake" at top
2317 level broke things. I've arranged for the value of $(MAKE) to be passed
2318 from the Makefile to this script so that it can call the same version of
2322 A note about Exim versions 4.44 and 4.50
2323 ----------------------------------------
2325 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2326 changes of various kinds. As a consequence, a big documentation update was
2327 needed. This delayed the release for rather longer than seemed good, especially
2328 in the light of a couple of (minor) security issues. Therefore, the changes
2329 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2330 release. So 4.44 and 4.50 are in effect two different branches that both start
2333 I have left the 4.50 change log unchanged; it contains all the changes since
2334 4.43. The change log for 4.44 is below; many of its items are identical to
2335 those for 4.50. This seems to be the most sensible way to preserve the
2336 historical information.
2342 1. Minor wording change to the doc/README.SIEVE file.
2344 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2345 computation of the current number of files was incorrect.
2347 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2348 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2351 4. Give more explanation in the error message when the command for a transport
2352 filter fails to execute.
2354 5. There are several places where Exim runs a non-Exim command in a
2355 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2356 was being done only for the command run by the queryprogram router. It is
2357 now done for all such subprocesses. The other cases are: ${run, transport
2358 filters, and the commands run by the lmtp and pipe transports.
2360 6. Added CONFIGURE_GROUP build-time option.
2362 7. Some older OS have a limit of 256 on the maximum number of file
2363 descriptors. Exim was using setrlimit() to set 1000 as a large value
2364 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2365 systems. I've change it so that if it can't get 1000, it tries for 256.
2367 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2368 was an oversight, and furthermore, ever since the addition of extra
2369 controls (e.g. 4.43/32), the checks on when to allow different forms of
2370 "control" were broken. There should now be diagnostics for all cases when a
2371 control that does not make sense is encountered.
2373 9. Added the /retain_sender option to "control=submission".
2375 10. $recipients is now available in the predata ACL (oversight).
2377 11. Tidy the search cache before the fork to do a delivery from a message
2378 received from the command line. Otherwise the child will trigger a lookup
2379 failure and thereby defer the delivery if it tries to use (for example) a
2380 cached ldap connection that the parent has called unbind on.
2382 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2383 of $address_data from the recipient verification was clobbered by the
2384 sender verification.
2386 13. The value of address_data from a sender verification is now available in
2387 $sender_address_data in subsequent conditions in the ACL statement.
2389 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2391 15. Added a new option "connect=<time>" to callout options, to set a different
2394 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2395 was its contents. (It was OK if the option was not defined at all.)
2397 17. A "Completed" log line is now written for messages that are removed from
2398 the spool by the -Mrm option.
2400 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2401 information about exactly what failed.
2403 19. Added -dd to debug only the daemon process.
2405 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2406 handles timeouts, both on the server side and network timeouts. Renamed the
2407 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2409 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2410 It is now set to "smtps".
2412 22. $host_address is now set to the target address during the checking of
2413 ignore_target_hosts.
2415 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2416 being passed; this would have caused $sender_host_name to have been used if
2417 matching the list had actually called for a host name (not very likely,
2418 since this list is usually IP addresses). A host name is now passed as
2421 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2422 code that specifies a non-restarting handler (typically sigaction() in
2423 modern systems) in an attempt to fix a rare and obscure crash bug.
2425 25. Narrowed the window for a race in the daemon that could cause it to ignore
2426 SIGCHLD signals. This is not a major problem, because they are used only to
2427 wake it up if nothing else does.
2429 26. A malformed maildirsize file could cause Exim to calculate negative values
2430 for the mailbox size or file count. Odd effects could occur as a result.
2431 The maildirsize information is now recalculated if the size or filecount
2434 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2435 support for a long time. Removed HAVE_SYS_VFS_H.
2437 28. Installed the latest version of exipick from John Jetmore.
2439 29. In an address list, if the pattern was not a regular expression, an empty
2440 subject address (from a bounce message) matched only if the pattern was an
2441 empty string. Non-empty patterns were not even tested. This was the wrong
2442 because it is perfectly reasonable to use an empty address as part of a
2443 database query. An empty address is now tested by patterns that are
2444 lookups. However, all the other forms of pattern expect the subject to
2445 contain a local part and a domain, and therefore, for them, an empty
2446 address still always fails if the pattern is not itself empty.
2448 30. Exim went into a mad DNS loop when attempting to do a callout where the
2449 host was specified on an smtp transport, and looking it up yielded more
2450 than one IP address.
2452 31. Re-factored the code for checking spool and log partition space into a
2453 function that finds that data and another that does the check. The former
2454 is then used to implement four new variables: $spool_space, $log_space,
2455 $spool_inodes, and $log_inodes.
2457 32. The RFC2047 encoding function was originally intended for short strings
2458 such as real names; it was not keeping to the 75-character limit for
2459 encoded words that the RFC imposes. It now respects the limit, and
2460 generates multiple encoded words if necessary. To be on the safe side, I
2461 have increased the buffer size for the ${rfc2047: expansion operator from
2464 33. It is now permitted to omit both strings after an "if" condition; if the
2465 condition is true, the result is "true". As before, when the second string
2466 is omitted, a false condition yields an empty string. This makes it less
2467 cumbersome to write custom ACL and router conditions.
2469 34. Failure to deliver a bounce message always caused it to be frozen, even if
2470 there was an errors_to setting on the router. The errors_to setting is now
2473 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2474 canonical form (fully expanded) before being placed in
2475 $sender_host_address.
2477 36. The table in the code that translates DNS record types into text (T_A to
2478 "A" for instance) was missing entries for NS and CNAME. It is just possible
2479 that this could have caused confusion if both these types were looked up
2480 for the same domain, because the text type is used as part of Exim's
2481 per-process caching. But the chance of anyone hitting this buglet seems
2484 37. The dnsdb lookup has been extended in a number of ways.
2486 (1) There is a new type, "zns", which walks up the domain tree until it
2487 finds some nameserver records. It should be used with care.
2489 (2) There is a new type, "mxh", which is like "mx" except that it returns
2490 just the host names, not the priorities.
2492 (3) It is now possible to give a list of domains (or IP addresses) to be
2493 looked up. The behaviour when one of the lookups defers can be
2494 controlled by a keyword.
2496 (4) It is now possible to specify the separator character for use when
2497 multiple records are returned.
2499 38. The dnslists ACL condition has been extended: it is now possible to supply
2500 a list of IP addresses and/or domains to be looked up in a particular DNS
2503 39. Added log_selector=+queue_time_overall.
2505 40. When running the queue in the test harness, wait just a tad after forking a
2506 delivery process, to get repeatability of debugging output.
2508 41. Include certificate and key file names in error message when GnuTLS fails
2509 to set them up, because the GnuTLS error message doesn't include the name
2510 of the failing file when there is a problem reading it.
2512 42. Allow both -bf and -bF in the same test run.
2514 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2516 44. The "Exiscan patch" is now merged into the mainline Exim source.
2518 45. Sometimes the final signoff response after QUIT could fail to get
2519 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2520 < 0 before doing a fflush(). This bug looks as though it goes back to the
2521 introduction of TLS in release 3.20, but "sometimes" must have been rare
2522 because the tests only now provoked it.
2524 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2525 (this can affect the format of dates).
2527 47. exim_tidydb, when checking for the continued existence of a message for
2528 which it has found a message-specific retry record, was not finding
2529 messages that were in split spool directories. Consequently, it was
2530 deleting retry records that should have stayed in existence.
2532 48. Steve fixed some bugs in eximstats.
2534 49. The SPA authentication driver was not abandoning authentication and moving
2535 on to the next authenticator when an expansion was forced to fail,
2536 contradicting the general specification for all authenticators. Instead it
2537 was generating a temporary error. It now behaves as specified.
2539 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2540 (the order specifies the preference for clients). The order is now AES256,
2541 AES128, 3DES, ARCFOUR128.
2543 51. Small patch to Sieve code - explicitly set From: when generating an
2546 52. Exim crashed if a remote delivery caused a very long error message to be
2547 recorded - for instance if somebody sent an entire SpamAssassin report back
2548 as a large number of 550 error lines. This bug was coincidentally fixed by
2549 increasing the size of one of Exim's internal buffers (big_buffer) that
2550 happened as part of the Exiscan merge. However, to be on the safe side, I
2551 have made the code more robust (and fixed the comments that describe what
2554 53. Now that there can be additional text after "Completed" in log lines (if
2555 the queue_time_overall log selector is set), a one-byte patch to exigrep
2556 was needed to allow it to recognize "Completed" as not the last thing in
2559 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2560 patch that reportedly fixes this has been added. I am not expert enough to
2561 create a test for it. This is what the patch creator wrote:
2563 "I found a little strange behaviour of ldap code when working with
2564 Windows 2003 AD Domain, where users was placed in more than one
2565 Organization Units. When I tried to give exim partial DN, the exit code
2566 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2567 But simultaneously result of request was absolutely normal ldap result,
2568 so I produce this patch..."
2570 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2571 REFERENCE, so I have modified the code to exclude the patch when that macro
2574 55. Some experimental protocols are using DNS PTR records for new purposes. The
2575 keys for these records are domain names, not reversed IP addresses. The
2576 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2577 leaves it alone. Component reversal etc. now happens only for IP addresses.
2580 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2582 57. Double the size of the debug message buffer (to 2048) so that more of very
2583 long debug lines gets shown.
2585 58. The exicyclog utility now does better if the number of log files to keep
2586 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2588 59. Two changes related to the smtp_active_hostname option:
2590 (1) $smtp_active_hostname is now available as a variable.
2591 (2) The default for smtp_banner uses $smtp_active_hostname instead
2592 of $primary_hostname.
2594 60. The host_aton() function is supposed to be passed a string that is known
2595 to be a valid IP address. However, in the case of IPv6 addresses, it was
2596 not checking this. This is a hostage to fortune. Exim now panics and dies
2597 if the condition is not met. A case was found where this could be provoked
2598 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2599 components; fortuitously, this particular loophole had already been fixed
2600 by change 4.50/55 above.
2602 If there are any other similar loopholes, the new check in host_aton()
2603 itself should stop them being exploited. The report I received stated that
2604 data on the command line could provoke the exploit when Exim was running as
2605 exim, but did not say which command line option was involved. All I could
2606 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2607 running as the user.
2610 61. There was a buffer overflow vulnerability in the SPA authentication code
2611 (which came originally from the Samba project). I have added a test to the
2612 spa_base64_to_bits() function which I hope fixes it.
2615 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2616 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2618 63. The daemon start-up calls getloadavg() while still root for those OS that
2619 need the first call to be done as root, but it missed one case: when
2620 deliver_queue_load_max is set with deliver_drop_privilege. This is
2621 necessary for the benefit of the queue runner, because there is no re-exec
2622 when deliver_drop_privilege is set.
2624 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2625 This has been fixed.
2627 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2628 was in use, was not putting the data itself into the right store pool;
2629 consequently, it could be overwritten for a subsequent message in the same
2630 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2633 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2635 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2636 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2637 the function were treating the return as a boolean value, which happened to
2638 work because 0=false and not-0=true, but is not correct code.
2640 68. The host_aton() function was not handling scoped IPv6 addresses (those
2641 with, for example, "%eth0" on the end) correctly.
2643 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2644 negated items (that is, ~something) in unsigned ints. Some compilers
2645 apparently mutter when there is no cast.
2647 70. If an address verification called from an ACL failed, and did not produce a
2648 user-specific message (i.e. there was only a "system" message), nothing was
2649 put in $acl_verify_message. In this situation, it now puts the system
2652 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2653 session; change 4.31/43 added the unwanted input to the log line - except
2654 that it did not do this in the start of session case. It now does.
2656 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2657 This is wrong and can cause the other end to generate a synchronization
2658 error if it is another Exim or anything else that does the synchronization
2659 check. A QUIT command is no longer sent after a timeout.
2661 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2662 during host lookups.
2664 74. The defer_ok option of callout verification was not working if it was used
2665 when verifying addresses in header lines, that is, for this case:
2667 verify = header_sender/callout=defer_ok
2669 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2670 those file descriptors could be used for SMTP connections. If anything
2671 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2672 could be sent to the SMTP client, causing chaos. The daemon now opens
2673 stdin, stdout, and stderr to /dev/null when it puts itself into the
2676 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2677 log by default. The user can override this with suitable Perl magic.
2679 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2680 the log message when discard triggers, was not working for the DATA ACL or
2681 for the non-SMTP ACL.
2683 78. Error message wording change in sieve.c.
2685 79. If smtp_accept_max_per_host was set, the number of connections could be
2686 restricted to fewer than expected, because the daemon was trying to set up
2687 a new connection before checking whether the processes handling previous
2688 connections had finished. The check for completed processes is now done
2689 earlier. On busy systems, this bug wouldn't be noticed because something
2690 else would have woken the daemon, and it would have reaped the completed
2693 80. If a message was submitted locally by a user whose login name contained one
2694 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2695 It caused a spool format error. I have fixed the spool reading code. A
2696 related problem was that the "from" clause in the Received: line became
2697 illegal because of the space(s). It is now covered by ${quote_local_part.
2699 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2702 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2704 83. Patch from Sieve maintainer to fix unterminated string problem in
2705 "vacation" handling.
2707 84. Some minor changes to the Linux configuration files to help with other
2708 OS variants using glibc.
2710 85. One more patch for Sieve to update vacation handling to latest spec.
2713 ----------------------------------------------------
2714 See the note above about the 4.44 and 4.50 releases.
2715 ----------------------------------------------------
2721 1. Change 4.43/35 introduced a bug that caused file counts to be
2722 incorrectly computed when quota_filecount was set in an appendfile
2725 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2726 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2729 3. Give more explanation in the error message when the command for a transport
2730 filter fails to execute.
2732 4. There are several places where Exim runs a non-Exim command in a
2733 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2734 was being done only for the command run by the queryprogram router. It is
2735 now done for all such subprocesses. The other cases are: ${run, transport
2736 filters, and the commands run by the lmtp and pipe transports.
2738 5. Some older OS have a limit of 256 on the maximum number of file
2739 descriptors. Exim was using setrlimit() to set 1000 as a large value
2740 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2741 systems. I've change it so that if it can't get 1000, it tries for 256.
2743 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2744 was an oversight, and furthermore, ever since the addition of extra
2745 controls (e.g. 4.43/32), the checks on when to allow different forms of
2746 "control" were broken. There should now be diagnostics for all cases when a
2747 control that does not make sense is encountered.
2749 7. $recipients is now available in the predata ACL (oversight).
2751 8. Tidy the search cache before the fork to do a delivery from a message
2752 received from the command line. Otherwise the child will trigger a lookup
2753 failure and thereby defer the delivery if it tries to use (for example) a
2754 cached ldap connection that the parent has called unbind on.
2756 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2757 of $address_data from the recipient verification was clobbered by the
2758 sender verification.
2760 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2761 was its contents. (It was OK if the option was not defined at all.)
2763 11. A "Completed" log line is now written for messages that are removed from
2764 the spool by the -Mrm option.
2766 12. $host_address is now set to the target address during the checking of
2767 ignore_target_hosts.
2769 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2770 being passed; this would have caused $sender_host_name to have been used if
2771 matching the list had actually called for a host name (not very likely,
2772 since this list is usually IP addresses). A host name is now passed as
2775 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2776 code that specifies a non-restarting handler (typically sigaction() in
2777 modern systems) in an attempt to fix a rare and obscure crash bug.
2779 15. Narrowed the window for a race in the daemon that could cause it to ignore
2780 SIGCHLD signals. This is not a major problem, because they are used only to
2781 wake it up if nothing else does.
2783 16. A malformed maildirsize file could cause Exim to calculate negative values
2784 for the mailbox size or file count. Odd effects could occur as a result.
2785 The maildirsize information is now recalculated if the size or filecount
2788 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2789 support for a long time. Removed HAVE_SYS_VFS_H.
2791 18. Updated exipick to current release from John Jetmore.
2793 19. Allow an empty sender to be matched against a lookup in an address list.
2794 Previously the only cases considered were a regular expression, or an
2797 20. Exim went into a mad DNS lookup loop when doing a callout where the
2798 host was specified on the transport, if the DNS lookup yielded more than
2801 21. The RFC2047 encoding function was originally intended for short strings
2802 such as real names; it was not keeping to the 75-character limit for
2803 encoded words that the RFC imposes. It now respects the limit, and
2804 generates multiple encoded words if necessary. To be on the safe side, I
2805 have increased the buffer size for the ${rfc2047: expansion operator from
2808 22. Failure to deliver a bounce message always caused it to be frozen, even if
2809 there was an errors_to setting on the router. The errors_to setting is now
2812 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2813 canonical form (fully expanded) before being placed in
2814 $sender_host_address.
2816 24. Updated eximstats to version 1.33
2818 25. Include certificate and key file names in error message when GnuTLS fails
2819 to set them up, because the GnuTLS error message doesn't include the name
2820 of the failing file when there is a problem reading it.
2822 26. Expand error message when OpenSSL has problems setting up cert/key files.
2825 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2826 (this can affect the format of dates).
2828 28. exim_tidydb, when checking for the continued existence of a message for
2829 which it has found a message-specific retry record, was not finding
2830 messages that were in split spool directories. Consequently, it was
2831 deleting retry records that should have stayed in existence.
2833 29. eximstats updated to version 1.35
2834 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2835 1.35 - bugfix such that pie charts by volume are generated correctly
2837 30. The SPA authentication driver was not abandoning authentication and moving
2838 on to the next authenticator when an expansion was forced to fail,
2839 contradicting the general specification for all authenticators. Instead it
2840 was generating a temporary error. It now behaves as specified.
2842 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2843 (the order specifies the preference for clients). The order is now AES256,
2844 AES128, 3DES, ARCFOUR128.
2846 31. Small patch to Sieve code - explicitly set From: when generating an
2849 32. Exim crashed if a remote delivery caused a very long error message to be
2850 recorded - for instance if somebody sent an entire SpamAssassin report back
2851 as a large number of 550 error lines. This bug was coincidentally fixed by
2852 increasing the size of one of Exim's internal buffers (big_buffer) that
2853 happened as part of the Exiscan merge. However, to be on the safe side, I
2854 have made the code more robust (and fixed the comments that describe what
2857 33. Some experimental protocols are using DNS PTR records for new purposes. The
2858 keys for these records are domain names, not reversed IP addresses. The
2859 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2860 leaves it alone. Component reversal etc. now happens only for IP addresses.
2863 34. The host_aton() function is supposed to be passed a string that is known
2864 to be a valid IP address. However, in the case of IPv6 addresses, it was
2865 not checking this. This is a hostage to fortune. Exim now panics and dies
2866 if the condition is not met. A case was found where this could be provoked
2867 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2868 components; fortuitously, this particular loophole had already been fixed
2869 by change 4.50/55 or 4.44/33 above.
2871 If there are any other similar loopholes, the new check in host_aton()
2872 itself should stop them being exploited. The report I received stated that
2873 data on the command line could provoke the exploit when Exim was running as
2874 exim, but did not say which command line option was involved. All I could
2875 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2876 running as the user.
2879 35. There was a buffer overflow vulnerability in the SPA authentication code
2880 (which came originally from the Samba project). I have added a test to the
2881 spa_base64_to_bits() function which I hope fixes it.
2884 36. The daemon start-up calls getloadavg() while still root for those OS that
2885 need the first call to be done as root, but it missed one case: when
2886 deliver_queue_load_max is set with deliver_drop_privilege. This is
2887 necessary for the benefit of the queue runner, because there is no re-exec
2888 when deliver_drop_privilege is set.
2890 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2891 was in use, was not putting the data itself into the right store pool;
2892 consequently, it could be overwritten for a subsequent message in the same
2893 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2896 38. Sometimes the final signoff response after QUIT could fail to get
2897 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2898 < 0 before doing a fflush(). This bug looks as though it goes back to the
2899 introduction of TLS in release 3.20, but "sometimes" must have been rare
2900 because the tests only now provoked it.
2906 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2907 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2908 Now it returns an integer. A number of calls were still expecting a T/F
2909 return. Fortuitously, in all cases, the tests worked in OK situations,
2910 which is the norm. However, things would have gone wrong on any write
2911 failures on the smtp file descriptor. This function is used when sending
2912 messages over SMTP and also when doing verify callouts.
2914 2. When Exim is called to do synchronous delivery of a locally submitted
2915 message (the -odf or -odi options), it no longer closes stderr before doing
2918 3. Implemented the mua_wrapper option.
2920 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2922 5. Implemented the functions header_remove(), header_testname(),
2923 header_add_at_position(), and receive_remove_recipient(), and exported them
2926 6. If an ACL "warn" statement specified the addition of headers, Exim already
2927 inserted X-ACL-Warn: at the start if there was no header name. However, it
2928 was not making this test for the second and subsequent header lines if
2929 there were newlines in the string. This meant that an invalid header could
2930 be inserted if Exim was badly configured.
2932 7. Allow an ACL "warn" statement to add header lines at the start or after all
2933 the Received: headers, as well as at the end.
2935 8. Added the rcpt_4xx retry error code.
2937 9. Added postmaster_mailfrom=xxx to callout verification option.
2939 10. Added mailfrom=xxxx to the callout verification option, for verify=
2942 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2943 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2945 12. Inserted some casts to stop certain compilers warning when using pointer
2946 differences as field lengths or precisions in printf-type calls (mostly
2947 affecting debugging statements).
2949 13. Added optional readline() support for -be (dynamically loaded).
2951 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2952 same clock tick as a message's arrival, so that its received time was the
2953 same as the "first fail" time on the retry record, and that message
2954 remained on the queue past the ultimate address timeout, every queue runner
2955 would try a delivery (because it was past the ultimate address timeout) but
2956 after another failure, the ultimate address timeout, which should have then
2957 bounced the address, did not kick in. This was a "< instead of <=" error;
2958 in most cases the first failure would have been in the next clock tick
2959 after the received time, and all would be well.
2961 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2962 being recognized when the domain list was tested by the match_domain
2963 condition in an expansion string.
2965 16. Added the ${str2b64: operator.
2967 17. Exim was always calling setrlimit() to set a large limit for the number of
2968 processes, without checking whether the existing limit was already
2969 adequate. (It did check for the limit on file descriptors.) Furthermore,
2970 errors from getrlimit() and setrlimit() were being ignored. Now they are
2971 logged to the main and panic logs, but Exim does carry on, to try to do its
2972 job under whatever limits there are.
2974 18. Imported PCRE 5.0.
2976 19. Trivial typo in log message " temporarily refused connection" (the leading
2979 20. If the log selector return_path_on_delivery was set and an address was
2980 redirected to /dev/null, the delivery process crashed because it assumed
2981 that a return path would always be set for a "successful" delivery. In this
2982 case, the whole delivery is bypassed as an optimization, and therefore no
2985 21. Internal re-arrangement: the function for sending a challenge and reading
2986 a response while authentication was assuming a zero-terminated challenge
2987 string. It's now changed to take a pointer and a length, to allow for
2988 binary data in such strings.
2990 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2992 23. Exim was not respecting finduser_retries when seeking the login of the
2993 uid under which it was called; it was always trying 10 times. (The default
2994 setting of finduser_retries is zero.) Also, it was sleeping after the final
2995 failure, which is pointless.
2997 24. Implemented tls_on_connect_ports.
2999 25. Implemented acl_smtp_predata.
3001 26. If the domain in control=submission is set empty, Exim assumes that the
3002 authenticated id is a complete email address when it generates From: or
3003 Sender: header lines.
3005 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3006 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3007 chown and chgrp in /bin and hostname in /usr/bin.
3009 28. Exim was keeping the "process log" file open after each use, just as it
3010 does for the main log. This opens the possibility of it remaining open for
3011 long periods when the USR1 signal hits a daemon. Occasional processlog
3012 errors were reported, that could have been caused by this. Anyway, it seems
3013 much more sensible not to leave this file open at all, so that is what now
3016 29. The long-running daemon process does not normally write to the log once it
3017 has entered its main loop, and it closes the log before doing so. This is
3018 so that log files can straightforwardly be renamed and moved. However,
3019 there are a couple of unusual error situations where the daemon does write
3020 log entries, and I had neglected to close the log afterwards.
3022 30. The text of an SMTP error response that was received during a remote
3023 delivery was being truncated at 512 bytes. This is too short for some of
3024 the long messages that one sometimes sees. I've increased the limit to
3027 31. It is now possible to make retry rules that apply only when a message has a
3028 specific sender, in particular, an empty sender.
3030 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3031 it possible to be selective about when SMTP synchronization is enforced.
3033 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3035 32. Implemented hosts_connection_nolog.
3037 33. Added an ACL for QUIT.
3039 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3042 35. Added mailbox_size and mailbox_filecount to appendfile.
3044 36. Added control = no_multiline_responses to ACLs.
3046 37. There was a bug in the logic of the code that waits for the clock to tick
3047 in the case where the clock went backwards by a substantial amount such
3048 that the microsecond fraction of "now" was more than the microsecond
3049 fraction of "then" (but the whole seconds number was less).
3051 38. Added support for the libradius Radius client library this is found on
3052 FreeBSD (previously only the radiusclient library was supported).
3058 1. When certain lookups returned multiple values in the form name=value, the
3059 quoting of the values was not always being done properly. Specifically:
3060 (a) If the value started with a double quote, but contained no whitespace,
3062 (b) If the value contained whitespace other than a space character (i.e.
3063 tabs or newlines or carriage returns) it was not quoted.
3064 This fix has been applied to the mysql and pgsql lookups by writing a
3065 separate quoting function and calling it from the lookup code. The fix
3066 should probably also be applied to nisplus, ibase and oracle lookups, but
3067 since I cannot test any of those, I have not disturbed their existing code.
3069 2. A hit in the callout cache for a specific address caused a log line with no
3070 reason for rejecting RCPT. Now it says "Previous (cached) callout
3071 verification failure".
3073 3. There was an off-by-one bug in the queryprogram router. An over-long
3074 return line was truncated at 256 instead of 255 characters, thereby
3075 overflowing its buffer with the terminating zero. As well as fixing this, I
3076 have increased the buffer size to 1024 (and made a note to document this).
3078 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3079 when Exim is waiting for an SMTP response from a remote server, Exim
3080 restarts its select() call on the socket, thereby resetting its timeout.
3081 This is not a problem when such interrupts are rare. Somebody set up a cron
3082 job to run exiwhat every 2 minutes, which is less than the normal select()
3083 timeout (5 or 10 minutes). This meant that the select() timeout never
3084 kicked in because it was always reset. I have fixed this by comparing the
3085 time when an interrupt arrives with the time at the start of the first call
3086 to select(). If more time than the timeout has elapsed, the interrupt is
3087 treated as a timeout.
3089 5. Some internal re-factoring in preparation for the addition of Sieve
3090 extensions (by MH). In particular, the "personal" test is moved to a
3091 separate function, and given an option for scanning Cc: and Bcc: (which is
3092 not set for Exim filters).
3094 6. When Exim created an email address using the login of the caller as the
3095 local part (e.g. when creating a From: or Sender: header line), it was not
3096 quoting the local part when it contained special characters such as @.
3098 7. Installed new OpenBSD configuration files.
3100 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3101 try to make them clearer.
3103 9. Callout options, other than the timeout value, were being ignored when
3104 verifying sender addresses in header lines. For example, when using
3106 verify = header_sender/callout=no_cache
3108 the cache was (incorrectly) being used.
3110 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3111 only the Cygwin environment.
3113 11. When return_path_on_delivery was set as a log selector, if different remote
3114 addresses in the same message used different return paths and parallel
3115 remote delivery occurred, the wrong values would sometimes be logged.
3116 (Whenever a remote delivery process finished, the return path value from
3117 the most recently started remote delivery process was logged.)
3119 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3120 lines when AUTH and/or TLS are in use. This is the "received protocol"
3121 field. Exim used to use "asmtp" for authenticated SMTP, without any
3122 indication (in the protocol name) for TLS use. Now it follows the RFC and
3123 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3124 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3125 names appear in log lines as well as in Received: header lines.
3127 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3128 extensions, and comparison tests, and to fix some bugs.
3130 14. Changes to the "personal" filter test:
3132 (1) The test was buggy in that it was just doing the equivalent of
3133 "contains" tests on header lines. For example, if a user's address was
3134 anne@some.where, the "personal" test would incorrectly be true for
3136 To: susanne@some.where
3138 This test is now done by extracting each address from the header in turn,
3139 and checking the entire address. Other tests that are part of "personal"
3140 are now done using regular expressions (for example, to check local parts
3141 of addresses in From: header lines).
3143 (2) The list of non-personal local parts in From: addresses has been
3144 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3145 taken from the Sieve specification recommendations.
3147 (3) If the message contains any header line starting with "List-" it is
3148 treated as non-personal.
3150 (4) The test for "circular" in the Subject: header line has been removed
3151 because it now seems ill-conceived.
3153 15. Minor typos in src/EDITME comments corrected.
3155 16. Installed latest exipick from John Jetmore.
3157 17. If headers_add on a router specified a text string that was too long for
3158 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3159 of string_sprintf() is now avoided.
3161 18. $message_body_size was not set (it was always zero) when running the DATA
3162 ACL and the local_scan() function.
3164 19. For the "mail" command in an Exim filter, no default was being set for
3165 the once_repeat time, causing a random time value to be used if "once" was
3166 specified. (If the value happened to be <= 0, no repeat happened.) The
3167 default is now 0s, meaning "never repeat". The "vacation" command was OK
3168 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3169 (I found it when inspecting the code).
3171 20. There is now an overall timeout for performing a callout verification. It
3172 defaults to 4 times the callout timeout, which applies to individual SMTP
3173 commands during the callout. The overall timeout applies when there is more
3174 than one host that can be tried. The timeout is checked before trying the
3175 next host. This prevents very long delays if there are a large number of
3176 hosts and all are timing out (e.g. when the network connections are timing
3177 out). The value of the overall timeout can be changed by specifying an
3178 additional sub-option for "callout", called "maxwait". For example:
3180 verify = sender/callout=5s,maxwait=20s
3182 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3183 to the end before writing, but this should make it even safer).
3185 22. Exim was forgetting that it had advertised PIPELINING for the second and
3186 subsequent messages on an SMTP connection. It was also not resetting its
3187 memory on STARTTLS and an internal HELO.
3189 23. When Exim logs an SMTP synchronization error within a session, it now
3190 records whether PIPELINING has been advertised or not.
3192 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3193 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3194 rather than long int.
3196 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3198 26. Added the never_mail option to autoreply.
3204 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3205 crash if the getsockname() call failed; this can happen if a connection is
3206 closed very soon after it is established. The problem was simply in the
3207 order in which certain operations were done, causing Exim to try to write
3208 to the SMTP stream before it had set up the file descriptor. The bug has
3209 been fixed by making things happen in the correct order.
3215 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3216 before the connection was closed, thus losing the rejection response.
3218 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3219 some early Solaris releases, but causes trouble in current releases where
3220 socklen_t is defined.
3222 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3225 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3228 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3229 that releasing the top of it at the end releases what was used for sub-
3230 expansions (unless the block got too big). However, discard this block if
3231 the first thing is a variable or header, so that we can use its block when
3232 it is dynamic (useful for very large $message_headers, for example).
3234 6. Lookups now cache *every* query, not just the most recent. A new, separate
3235 store pool is used for this. It can be recovered when all lookup caches are
3236 flushed. Lookups now release memory at the end of their result strings.
3237 This has involved some general refactoring of the lookup sources.
3239 7. Some code has been added to the store_xxx() functions to reduce the amount
3240 of flapping under certain conditions.
3242 8. log_incoming_interface used to affect only the <= reception log lines. Now
3243 it causes the local interface and port to be added to several more SMTP log
3244 lines, for example "SMTP connection from", and rejection lines.
3246 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3248 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3250 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3251 could be overwritten at the end of the current message (or the start of a
3252 new message if it was set in a HELO ACL). The value is now preserved for
3253 the duration of the SMTP connection.
3255 12. If a transport had a headers_rewrite setting, and a matching header line
3256 contained an unqualified address, that address was qualified, even if it
3257 did not match any rewriting rules. The underlying bug was that the values
3258 of the flags that permit the existence of unqualified sender and recipient
3259 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3260 non-local messages, and by -bnq for local messages) were not being
3261 preserved with the message after it was received.
3263 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3264 "next input=" as part of the text comprising the host identity instead of
3265 the correct text. The code was using the same buffer for two different
3266 strings. However, depending on which order the printing function evaluated
3267 its arguments, the bug did not always show up. Under Linux, for example, my
3268 test suite worked just fine.
3270 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3271 doesn't work with some older versions of Perl. It has been changed to "my",
3272 which in any case is probably the better facility to use.
3274 15. A really picky compiler found some instances of statements for creating
3275 error messages that either had too many or two few arguments for the format
3278 16. The size of the buffer for calls to the DNS resolver has been increased
3279 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3280 for addresses that have a lot of PTR records. This alleviates a problem; it
3281 does not fully solve it.
3283 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3284 buffer now truncates the list and logs the incident, which is the same
3285 action as happens when Exim is looking up a host name and its aliases.
3286 Previously in this situation something unpredictable would happen;
3287 sometimes it was "internal error: store_reset failed".
3289 18. If a server dropped the connection unexpectedly when an Exim client was
3290 using GnuTLS and trying to read a response, the client delivery process
3291 crashed while trying to generate an error log message.
3293 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3294 string, for example:
3296 warn message = H1: something\nH2: something
3298 the text was added as a single header line from Exim's point of view
3299 though it ended up OK in the delivered message. However, searching for the
3300 second and subsequent header lines using $h_h2: did not work. This has been
3301 fixed. Similarly, if a system filter added multiple headers in this way,
3302 the routers could not see them.
3304 20. Expanded the error message when iplsearch is called with an invalid key to
3305 suggest using net-iplsearch in a host list.
3307 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3308 ACLs are no longer actually imposed (and a message to that effect is
3311 22. If a "gecos" field in a passwd entry contained escaped characters, in
3312 particular, if it contained a \" sequence, Exim got it wrong when building
3313 a From: or a Sender: header from that name. A second bug also caused
3314 incorrect handling when an unquoted " was present following a character
3315 that needed quoting.
3317 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3318 was not being matched caselessly.
3320 24. Arranged for all hyphens in the exim.8 source to be escaped with
3323 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3324 properly. Recipient callout cache records were still being keyed to include
3325 the sender, even when use_sender was set false. This led to far more
3326 callouts that were necessary. The sender is no longer included in the key
3327 when use_sender is false.
3329 26. Added "control = submission" modifier to ACLs.
3331 27. Added the ${base62d: operator to decode base 62 numbers.
3333 28. dnsdb lookups can now access SRV records.
3335 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3336 the configuration file.
3338 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3339 (-v) mode. This makes the output for a verbose queue run more intelligible.
3341 31. Added a use_postmaster feature to recipient callouts.
3343 32. Added the $body_zerocount variable, containing the number of binary zero
3344 bytes in the message body.
3346 33. The time of last modification of the "new" subdirectory is now used as the
3347 "mailbox time last read" when there is a quota error for a maildir
3350 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3352 35. Added +ignore_unknown as a special item in host lists.
3354 36. Code for decoding IPv6 addresses in host lists is now included, even if
3355 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3356 address was recognized as an IP address, but was then not correctly decoded
3357 into binary, causing unexpected and incorrect effects when compared with
3364 1. Very minor rewording of debugging text in manualroute to say "list of
3365 hosts" instead of "hostlist".
3367 2. If verify=header_syntax was set, and a header line with an unqualified
3368 address (no domain) and a large number of spaces between the end of the
3369 name and the colon was received, the reception process suffered a buffer
3370 overflow, and (when I tested it) crashed. This was caused by some obsolete
3371 code that should have been removed. The fix is to remove it!
3373 3. When running in the test harness, delay a bit after writing a bounce
3374 message to get a bit more predictability in the log output.
3376 4. Added a call to search_tidyup() just before forking a reception process. In
3377 theory, someone could use a lookup in the expansion of smtp_accept_max_
3378 per_host which, without the tidyup, could leave open a database connection.
3380 5. Added the variables $recipient_data and $sender_data which get set from a
3381 lookup success in an ACL "recipients" or "senders" condition, or a router
3382 "senders" option, similar to $domain_data and $local_part_data.
3384 6. Moved the writing of debug_print from before to after the "senders" test
3387 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3388 problems for message scanning, either using a data ACL, or using
3389 local_scan() because the Received: header was not generated till after they
3390 were called (in order to set the time as the time of reception completion).
3391 I have revised the way this works. The header is now generated after the
3392 body is received, but before the ACL or local_scan() are called. After they
3393 are run, the timestamp in the header is updated.
3399 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3400 before starting a queue runner without re-exec. This happened only when
3401 deliver_drop_privilege was set or when the Exim user was set to root. The
3402 effect of the bug was that timeouts during subsequent deliveries caused
3403 crashes instead of being properly handled. The handler is now left at its
3404 default (and expected) setting.
3406 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3407 message, again when deliver_drop_privilege is set or Exim is run as root.
3408 The bug described in (1) was not present in this case, but the tidying up
3409 of the other signals was missing. I have made the two cases consistent.
3411 3. The ignore_target_hosts setting on a manualroute router was being ignored
3412 for hosts that were looked up using the /MX notation.
3414 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3417 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3418 operated on the sender address. After changing the $sender_address to <>
3419 for the sender address verify, Exim was re-instated it as the original
3420 (before rewriting) address, but remembering that it had rewritten it, so it
3421 wasn't rewriting it again. This bug also had the effect of breaking the
3422 sender address verification caching when the sender address was rewritten.
3424 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3425 This has been changed so that if the ip literal address matches
3426 ignore_target_hosts, the router declines.
3428 7. Added expansion conditions match_domain, match_address, and match_local_
3429 part (NOT match_host).
3431 8. The placeholder for the Received: header didn't have a length field set.
3433 9. Added code to Exim itself and to exim_lock to test for a specific race
3434 condition that could lead to file corruption when using MBX delivery. The
3435 issue is with the lockfile that is created in /tmp. If this file is removed
3436 after a process has opened it but before that process has acquired a lock,
3437 there is the potential for a second process to recreate the file and also
3438 acquire a lock. This could lead to two Exim processes writing to the file
3439 at the same time. The added code performs the same test as UW imapd; it
3440 checks after acquiring the lock that its file descriptor still refers to
3441 the same named file.
3443 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3444 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3447 11. Added the smtp_active_hostname option. If used, this will typically be made
3448 to depend on the incoming interface address. Because $interface_address is
3449 not set up until the daemon has forked a reception process, error responses
3450 that can happen earlier (such as "too many connections") no longer contain
3453 12. If an expansion in a condition on a "warn" statement fails because a lookup
3454 defers, the "warn" statement is abandoned, and the next ACL statement is
3455 processed. Previously this caused the whole ACL to be aborted.
3457 13. Added the iplsearch lookup type.
3459 14. Added ident_timeout as a log selector.
3461 15. Added tls_certificate_verified as a log selector.
3463 16. Added a global option tls_require_ciphers (compare the smtp transport
3464 option of the same name). This controls incoming TLS connections.
3466 17. I finally figured out how to make tls_require_ciphers do a similar thing
3467 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3468 before starting the TLS session.
3470 18. Tabs are now shown as \t in -bP output.
3472 19. If the log selector return_path_on_delivery was set, Exim crashed when
3473 bouncing a message because it had too many Received: header lines.
3475 20. If two routers both had headers_remove settings, and the first one included
3476 a superfluous trailing colon, the final name in the first list and the
3477 first name in the second list were incorrectly joined into one item (with a
3478 colon in the middle).
3484 1. Added -C and -D options to the exinext utility, mainly to make it easier
3485 to include in the automated testing, but these could be helpful when
3486 multiple configurations are in use.
3488 2. The exinext utility was not formatting the output nicely when there was
3489 an alternate port involved in the retry record key, nor when there was a
3490 message id as well (for retries that were specific to a specific message
3491 and a specific host). It was also confused by IPv6 addresses, because of
3492 the additional colons they contain. I have fixed the IPv4 problem, and
3493 patched it up to do a reasonable job for IPv6.
3495 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3496 delivery, the log line now contains "pipelined" if PIPELINING was used.
3498 4. An SMTP transport process used to panic and die if the bind() call to set
3499 an explicit outgoing interface failed. This has been changed; it is now
3500 treated in the same way as a connect() failure.
3502 5. A reference to $sender_host_name in the part of a conditional expansion
3503 that was being skipped was still causing a DNS lookup. This no longer
3506 6. The def: expansion condition was not recognizing references to header lines
3507 that used bh_ and bheader_.
3509 7. Added the _cache feature to named lists.
3511 8. The code for checking quota_filecount in the appendfile transport was
3512 allowing one more file than it should have been.
3514 9. For compatibility with Sendmail, the command line option
3522 and sets the incoming protocol and host name (for trusted callers). The
3523 host name and its colon can be omitted when only the protocol is to be set.
3524 Note the Exim already has two private options, -pd and -ps, that refer to
3525 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3526 "s", but I don't think that's a major issue.
3528 10. A number of refactoring changes to the code, none of which should affect
3531 (a) The number of logging options was getting close to filling up the
3532 32-bit word that was used as a bit map. I have split them into two classes:
3533 those that are passed in the argument to log_write(), and those that are
3534 only ever tested independently outside of that function. These are now in
3535 separate 32-bit words, so there is plenty of room for expansion again.
3536 There is no change in the user interface or the logging behaviour.
3538 (b) When building, for example, log lines, the code previously used a
3539 macro that called string_cat() twice, in order to add two strings. This is
3540 not really sufficiently general. Furthermore, there was one instance where
3541 it was actually wrong because one of the argument was used twice, and in
3542 one call a function was used. (As it happened, calling the function twice
3543 did not affect the overall behaviour.) The macro has been replaced by a
3544 function that can join an arbitrary number of extra strings onto a growing
3547 (c) The code for expansion conditions now uses a table and a binary chop
3548 instead of a serial search (which was left over from when there were very
3549 few conditions). Also, it now recognizes conditions like "pam" even when
3550 the relevant support is not compiled in: a suitably worded error message is
3551 given if an attempt is made to use such a condition.
3553 11. Added ${time_interval:xxxxx}.
3555 12. A bug was causing one of the ddress fields not to be passed back correctly
3556 from remote delivery subprocesses. The field in question was not being
3557 subsequently used, so this caused to problems in practice.
3559 13. Added new log selectors queue_time and deliver_time.
3561 14. Might have fixed a bug in maildirsizefile handling that threw up
3562 "unexpected character" debug warnings, and recalculated the data
3563 unnecessarily. In any case, I expanded the warning message to give more
3566 15. Added the message "Restricted characters in address" to the statements in
3567 the default ACL that block characters like @ and % in local parts.
3569 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3570 Three changes have been made:
3572 (a) There was a serious bug; a negative response to MAIL caused the whole
3573 recipient domain to be cached as invalid, thereby blocking all messages
3574 to all local parts at the same domain, from all senders. This bug has
3575 been fixed. The domain is no longer cached after a negative response to
3576 MAIL if the sender used is not empty.
3578 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3581 (c) A new callout option, "use_sender" has been added for people who want
3582 the modified behaviour.
3588 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3591 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3592 indeed breaks things for older releases.
3594 3. Added additional logging to the case where there is a problem reading data
3595 from a filter that is running in a subprocess using a pipe, in order to
3596 try to track down a specific problem.
3598 4. Testing facility fudge: when running in the test harness and attempting
3599 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3600 getting "No route to host". Convert this to a timeout.
3602 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3605 6. Some OS don't have socklen_t but use size_t instead. This affects the
3606 fifth argument of getsockopt() amongst other things. This is now
3607 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3608 can be set for individual OS. I have set it for SunOS5, OSF1, and
3609 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3610 some earlier ones do not.
3612 7. Change 4.30/15 was not doing the test caselessly.
3614 8. The standard form for an IPv6 address literal was being rejected by address
3615 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3616 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3617 this, as well as the form without the "IPv6" on the front (but only when
3618 address literals are enabled, of course).
3620 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3622 10. Exim crashed if a message with an empty sender address specified by -f
3623 encountered a router with an errors_to setting. This could be provoked only
3624 by a command such as
3628 where an empty string was supplied; "<>" did not hit this bug.
3630 11. Installed PCRE release 4.5.
3632 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3633 remained set. It is now erased.
3635 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3636 times from message ids (which are base 36 rather than the normal 62).
3638 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3639 were being counted as actual protocol errors, and logged if the log
3640 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3641 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3642 and DATA following a set of rejected RCPTs do not count as protocol errors.
3643 In other words, Exim assumes they were pipelined, though this may not
3644 actually be the case. Of course, in all cases the client gets an
3645 appropriate error code.
3647 15. If a lookup fails in an ACL condition, a message about the failure may
3648 be available; it is used if testing the ACL cannot continue, because most
3649 such messages specify what the cause of the deferral is. However, some
3650 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3651 that caused an old message to be retained and used if a later statement
3652 caused a defer, replacing the real cause of the deferral.
3654 16. If an IP address had so many PTR records that the DNS lookup buffer
3655 was not large enough to hold them, Exim could crash while trying to process
3656 the truncated data. It now detects and logs this case.
3658 17. Further to 4.21/58, another change has been made: if (and only if) the
3659 first line of a message (the first header line) ends with CRLF, a bare LF
3660 in a subsequent header line has a space inserted after it, so as not to
3661 terminate the header.
3663 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3664 unnecessarily, used atoi() instead of strtol(), and didn't check the
3665 termination when getting file sizes from file names by regex.
3667 19. Completely re-implemented the support for maildirsize files, in the light
3668 of a number of problems with the previous contributed implementation
3669 (4.30/29). In particular:
3671 . If the quota is zero, the maildirsize file is maintained, but no quota is
3674 . If the maildir directory does not exist, it is created before any attempt
3675 to write a maildirsize file.
3677 . The quota value in the file is just a cache; if the quota is changed in
3678 the transport, the new value overrides.
3680 . A regular expression is available for excluding directories from the
3683 20. The autoreply transport checks the characters in options that define the
3684 message's headers; it allows continued headers, but it was checking with
3685 isspace() after an embedded newline instead of explicitly looking for a
3688 21. If all the "regular" hosts to which an address was routed had passed their
3689 expiry times, and had not reached their retry times, the address was
3690 bounced, even if fallback hosts were defined. Now Exim should go on to try
3693 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3694 equivalent code in the SMTP transport. Some hosts send humungous responses
3695 to HELO/EHLO, more than 1024 it seems.
3697 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3698 gives compiler warnings in some environments. I've now done it "properly",
3701 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3702 (because of problems with the built-in one) was declared to return uschar *
3703 instead of char *, causing compiler failure.
3705 25. Fixed a file descriptor leak when processing alias/forward files.
3707 26. Fixed a minor format string issue in dbfn.c.
3709 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3711 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3712 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3715 29. When a sender address is verified, it is cached, to save repeating the test
3716 when there is more than one recipient in a message. However, when the
3717 verification involves a callout, it is possible for different callout
3718 options to be set for different recipients. It is too complicated to keep
3719 track of this in the cache, so now Exim always runs a verification when a
3720 callout is required, relying on the callout cache for the optimization.
3721 The overhead is duplication of the address routing, but this should not be
3724 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3725 to be verified with callout=postmaster, and the main callout worked but the
3726 postmaster check failed, the verification correctly failed. However, if a
3727 subsequent RCPT command asked for sender verification *without* the
3728 postmaster check, incorrect caching caused this verification also to fail,
3731 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3732 it was not caching the DNS options (qualify_single, search_parents) that
3733 were used when the lookup failed. A subsequent lookup with different
3734 options therefore always gave the same answer, though there were cases
3735 where it should not have. (Example: a "domains = !$mx_any" option on a
3736 dnslookup router: the "domains" option is always processed without any
3737 widening, but the router might have qualify_single set.) Now Exim uses the
3738 cached value only when the same options are set.
3740 32. Added John Jetmore's "exipick" utility to the distribution.
3742 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3743 than a timeout (e.g. a certificate is required, and is not provided), an
3744 Exim server now closes the connection immediately. Previously it waited for
3745 the client to close - but if the client is SSL, it seems that they each
3746 wait for each other, leading to a delay before one of them times out.
3748 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3749 maintained 0.8.x compatibility because I don't think many are using it, and
3750 it is clearly obsolete.
3752 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3755 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3756 client certificate was expired. A simple patch fixes this, though I don't
3757 understand the full logic of why the verify callback is called multiple
3760 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3761 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3762 which causes problems with some clients (such as the Certicom SSL Plus
3763 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3764 disables the coutermeasure allowing Eudora to connect."
3766 38. Exim was not checking that a write() to a log file succeeded. This could
3767 lead to Bad Things if a log got too big, in particular if it hit a file
3768 size limit. Exim now panics and dies if it cannot write to a log file, just
3769 as it does if it cannot open a log file.
3771 39. Modified OS/Makefile-Linux so that it now contains
3773 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3775 The two -D definitions ensure that Exim is compiled with large file
3776 support, which makes it possible to handle log files that are bigger than
3779 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3780 instance) a domain was checked against a named list that involved a lookup,
3781 causing $domain_data to be set, then another domain was checked against the
3782 same list, then the first domain was re-checked, the value of $domain_data
3783 after the final check could be wrong. In particular, if the second check
3784 failed, it could be set empty. This bug probably also applied to
3787 41. The strip_trailing_dot option was not being applied to the address given
3788 with the -f command-line option.
3790 42. The code for reading a message's header from the spool was incrementing
3791 $received_count, but never initializing it. This meant that the value was
3792 incorrect (doubled) while delivering a message in the same process in which
3793 it was received. In the most common configuration of Exim, this never
3794 happens - a fresh exec is done - but it can happen when
3795 deliver_drop_privilege is set.
3797 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3798 it now includes up to 150 characters of the unexpected data in the log
3801 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3802 and building data strings. The size of both of these buffers was 10 000
3803 bytes - far larger than anybody would *ever* want, thought I. Needless to
3804 say, somebody hit the limit. I have increased the maximum line length to
3805 20 000 and the maximum data length of concatenated lines to 100 000. I have
3806 also fixed two bugs, because there was no checking on these buffers. Tsk,
3807 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3808 buffer is too small.
3810 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3811 lsearch lookups. Now it does.
3813 46. When parsing a route_list item in a manualroute router, a fixed-length
3814 buffer was used for the list of hosts. I made this 1024 bytes long,
3815 thinking that nobody would ever have a list of hosts that long. Wrong.
3816 Somebody had a whole pile of complicated expansion conditions, and the
3817 string was silently truncated, leading to an expansion error. It turns out
3818 that it is easier to change to an unlimited length (owing to other changes
3819 that have happened since this code was originally written) than to build
3820 structure for giving a limitation error. The length of the item that
3821 expands into the list of hosts is now unlimited.
3823 47. The lsearch lookup could not handle data where the length of text line was
3824 more than 4095 characters. Such lines were truncated, leading to shortened
3825 data being returned. It should now handle lines of any length.
3827 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3828 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3831 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3832 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3833 no longer happens in comment lines. A list of replacements is now placed
3834 at the head of all of the source files, except those whose only change is
3835 to replace PERL_COMMAND in the very first #! line.
3837 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3838 messages on the queue, with a bottom-up merge sort, using code contributed
3839 by Michael Haardt. This should make operations like -bp somewhat faster on
3840 large queues. It won't affect queue runners, except when queue_run_in_order
3843 51. Installed eximstats 1.31 in the distribution.
3845 52. Added support for SRV lookups to the dnslookup router.
3847 53. If an ACL referred to $message_body or $message_body_end, the value was not
3848 reset for any messages that followed in the same SMTP session.
3850 54. The store-handling optimization for building very long strings was not
3851 differentiating between the different store pools. I don't think this
3852 actually made any difference in practice, but I've tidied it.
3854 55. While running the routers to verify a sender address, $sender_address
3855 was still set to the sender address. This is wrong, because when routing to
3856 send a bounce to the sender, it would be empty. Therefore, I have changed
3857 it so that, while verifying a sender address, $sender_address is set to <>.
3858 (There is no change to what happens when verifying a recipient address.)
3860 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3861 target A or AAAA records (if not already returned) without resetting the
3862 qualify_single or search_parents options of the DNS resolver. These are
3863 inappropriate in this case because the targets of MX and SRV records must
3864 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3865 target that, when qualified, matched something in the local domain. These
3866 two options are now turned off when doing these lookups.
3868 57. It seems that at least some releases of Reiserfs (which does not have the
3869 concept of a fixed number of inodes) returns zero and not -1 for the
3870 number of available inodes. This interacted badly with check_spool_inodes,
3871 which assumed that -1 was the "no such thing" setting. What I have done is
3872 to check that the total number of inodes is greater than zero before doing
3873 the test of how many are available.
3875 58. When a "warn" ACL statement has a log_message modifier, the message is
3876 remembered, and not repeated. This is to avoid a lot of repetition when a
3877 message has many recipients that cause the same warning to be written.
3878 Howewer, Exim was preserving the list of already written lines for an
3879 entire SMTP session, which doesn't seem right. The memory is now reset if a
3880 new message is started.
3882 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3883 debugging output unless log_rewrite was also set.
3885 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3886 of (int)(handle) when we know that handle contains (void *)(-1).
3888 61. The Exim daemon panic-logs an error return when it closes the incoming
3889 connection. However "connection reset by peer" seems to be common, and
3890 isn't really an error worthy of noting specially, so that particular error
3893 62. When Exim is trying to find all the local interfaces, it used to panic and
3894 die if the ioctl to get the interface flags failed. However, it seems that
3895 on at least one OS (Solaris 9) it is possible to have an interface that is
3896 included in the list of interfaces, but for which you get a failure error
3897 for this call. This happens when the interface is not "plumbed" into a
3898 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3899 failure of the "get flags" call assumes that the interface is down.
3901 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3902 makes life easier for people who are doing arithmetic on fields extracted
3903 from dates, where you often get leading zeros that should not be
3904 interpreted as octal.
3906 64. Added qualify_domain to the redirect router, to override the global
3909 65. If a pathologically long header line contained very many addresses (the
3910 report of this problem mentioned 10 000) and each of them was rewritten,
3911 Exim could use up a very large amount of memory. (It kept on making new
3912 copies of the header line as it rewrote, and never released the old ones.)
3913 At the expense of a bit more processing, the header rewriting function has
3914 been changed so that it no longer eats memory in this way.
3916 66. The generation of the Received: header has been moved from the time that a
3917 message starts to be received, to the time that it finishes. The timestamp
3918 in the Received: header should now be very close to that of the <= log
3919 line. There are two side-effects of this change:
3921 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3922 logged header lines no longer include the local Received: line, because
3923 it has not yet been created. The same applies to a copy of the message
3924 that is returned to a non-SMTP sender when a message is rejected.
3926 (b) When a filter file is tested using -bf, no additional Received: header
3927 is added to the test message. After some thought, I decided that this
3930 This change does not affect the value of $received_for. It is still set
3931 after address rewriting, but before local_scan() is called.
3933 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3935 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3936 gave an unhelpful panic error message, and a defer error. I have managed to
3937 change this behaviour so that it now rejects any supplied certificate,
3938 which seems right, as the list of acceptable certificates is empty.
3940 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3941 gave an unhelpful defer error. I have not managed to make this reject any
3942 supplied certificates, but the error message it gives is "no certificate
3943 supplied", which is not helpful.
3945 70. exigrep's output now also includes lines that are not associated with any
3946 message, but which match the given pattern. Implemented by a patch from
3947 Martin Sluka, which also tidied up the Perl a bit.
3949 71. Recipient callout verification, like sender verification, was using <> in
3950 the MAIL FROM command. This isn't really the right thing, since the actual
3951 sender may affect whether the remote host accepts the recipient or not. I
3952 have changed it to use the actual sender in the callout; this means that
3953 the cache record is now keyed on a recipient/sender pair, not just the
3954 recipient address. There doesn't seem to be a real danger of callout loops,
3955 since a callout by the remote host to check the sender would use <>.
3956 [SEE ABOVE: changed after hitting problems.]
3958 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3959 temporary errors. However, in the case of such a code being given after
3960 the end of a data transmission (i.e. after ".") Exim was failing to write
3961 a retry record for the message. (Yes, there was some broken host that was
3962 actually sending 8xx at this point.)
3964 73. An unknown lookup type in a host list could cause Exim to panic-die when
3965 the list was checked. (An example that provoked this was putting <; in the
3966 middle of a list instead of at the start.) If this happened during a DATA
3967 ACL check, a -D file could be left lying around. This kind of configuration
3968 error no longer causes Exim to die; instead it causes a defer errror. The
3969 incident is still logged to the main and panic logs.
3971 74. Buglet left over from Exim 3 conversion. The message "too many messages
3972 in one connection" was written to the rejectlog but not the mainlog, except
3973 when address rewriting (yes!) was being logged.
3975 75. Added write_rejectlog option.
3977 76. When a system filter was run not as root (that is, when system_filter_user
3978 was set), the values of the $n variables were not being returned to the
3979 main process; thus, they were not subsequently available in the $sn
3982 77. Added +return_path_on_delivery log selector.
3984 78. A connection timeout was being treated differently from recipients deferred
3985 when testing hosts_max_try with a message that was older than the host's
3986 retry timeout. (The host should not be counted, thus allowing all hosts to
3987 be tried at least once before bouncing.) This may have been the cause of an
3988 occasionally reported bug whereby a message would remain on the queue
3989 longer than the retry timeout, but would be bounced if a delivery was
3990 forced. I say "may" because I never totally pinned down the problem;
3991 setting up timeout/retry tests is difficult. See also the next item.
3993 79. The ultimate address timeout was not being applied to errors that involved
3994 a combination of host plus message (for example, a timeout on a MAIL
3995 command). When an address resolved to a number of possible hosts, and they
3996 were not all tried for each delivery (e.g. because of hosts_max_try), a
3997 message could remain on the queue longer than the retry timeout.
3999 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4002 81. Fixed an obscure SMTP outgoing bug which required at least the following
4003 conditions: (a) there was another message waiting for the same server;
4004 (b) the server returned 5xx to all RCPT commands in the first message so
4005 that the message was not completed; (c) the server dropped the connection
4006 or gave a negative response to the RSET that Exim sends to abort the
4007 transaction. The observed case was a dropped connection after DATA that had
4008 been sent in pipelining mode. That is, the server had advertised PIPELINING
4009 but was not implementing it correctly. The effect of the bug was incorrect
4010 behaviour, such as trying another host, and this could lead to a crash.
4016 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4017 and daemon.c were passed as pointers to ints; they should have been
4018 pointers to socklen_t variables (which are typically unsigned ints).
4020 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4023 3. Fixed a really odd bug that affected only the testing scheme; patching a
4024 certain fixed string in the binary changed the value of another string that
4025 happened to be identical to the end of the original first string.
4027 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4028 name", it returns that address as the IP address. On some operating
4029 systems (e.g. Solaris), it also passes back the IP address string as the
4030 "host name". However, on others (e.g. Linux), it passes back an empty
4031 string. Exim wasn't checking for this, and was changing the host name to an
4032 empty string, assuming it had been canonicized.
4034 5. Although rare, it is permitted to have more than one PTR record for a given
4035 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4036 all the names associated with an address, because they do in Solaris.
4037 However, it seems that they do not in Linux for data that comes from the
4038 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4039 I found this out when I moved to a new Linux workstation and tried to run
4040 the Exim test suite.
4042 To get round this problem I have changed the code so that it now does its
4043 own call to the DNS to look up PTR records when searching for a host name.
4044 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4045 addresses that are only in /etc/hosts are still found.
4047 This behaviour is, however, controlled by an option called host_lookup_
4048 order, which defaults to "bydns:byaddr". If people want to use the other
4049 order, or indeed, just use one or the other means of lookup, they can
4050 specify it in this variable.
4052 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4053 some operating systems, this comes back from gethostbyaddr() as an empty
4054 string, and this is what Exim used to test for. However, it seems that in
4055 other systems, "." is yielded. Exim now tests for this case too.
4057 7. The values of check_spool_space and check_log_space are now held internally
4058 as a number of kilobytes instead of an absolute number of bytes. If a
4059 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4060 kilobyte. This means that much larger values can be stored.
4062 8. Exim monitor: an attempt to get the action menu when not actually pointing
4063 at a message produces an empty menu entitled "No message selected". This
4064 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4065 no entries in it ("Shell widget menu has zero width and/or height"). So I
4066 have added a single, blank menu entry in this case.
4068 9. Added ${quote_local_part.
4070 10. MIME decoding is now applied to the contents of Subject: header lines when
4073 11. Now that a reference to $sender_host_address automatically causes a reverse
4074 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4075 host lookup before query-style lookups in lists that might use this
4076 variable. This has therefore been abolished, and the "net-" prefix is no
4077 longer necessary for query-style lookups.
4079 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4080 have been a typo for LFLAGS, so it has been changed.
4082 13. The install script calls Exim with "-C /dev/null" in order to find the
4083 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4084 to be output. Howeve, since Exim outputs its version number before the
4085 error, it didn't break the script. It just looked ugly. I fixed this by
4086 always allowing "-C /dev/null" if the caller is root.
4088 14. Ignore overlarge ACL variable number when reading spool file - insurance
4089 against a later release with more variables having written the file.
4091 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4092 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4093 this, as well as the form without the "IPv6" on the front.
4095 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4096 OS/Makefile-Darwin file.
4098 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4099 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4101 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4102 variable. However, some OS use ints for inodes. I've added cast to long int
4103 to get rid of the compiler warning.
4105 19. I had forgotten to lock out "/../" in configuration file names when
4106 ALT_CONFIG_PREFIX was set.
4108 20. Routers used for verification do not need to specify transports. However,
4109 if such a router generated a host list, and callout was configured, Exim
4110 crashed, because it could not find a port number from the (non-existent)
4111 transport. It now assumes port 25 in this circumstance.
4113 21. Added the -t option to exigrep.
4115 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4116 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4117 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4118 them is set, LOOKUP_LSEARCH is forced.
4120 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4122 24. Added sender and host information to the "rejected by local_scan()" log
4123 line; previously there was no indication of these.
4125 25. Added .include_if_exists.
4127 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4128 Linux. It turns out that not all file systems support this. Apparently some
4129 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4130 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4131 not supported on the file descriptor, is now ignored when Exim is trying to
4132 sync a directory. This applies only to Linux.
4134 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4136 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4137 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4140 29. Added support for maildirsize files from supplied patch (modified a bit).
4142 30. The use of :fail: followed by an empty string could lead Exim to respond to
4143 sender verification failures with (e.g.):
4145 550 Verification failed for <xxx>
4146 550 Sender verify failed
4148 where the first response line was missing the '-' that indicates it is not
4149 the final line of the response.
4151 31. The loop for finding the name of the user that called Exim had a hardwired
4152 limit of 10; it now uses the value of finduser_retries, which is used for
4153 all other user lookups.
4155 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4158 33. Exim was neglecting to zero errno before one call of strtol() when
4159 expanding a string and expecting an integer value. On some systems this
4160 resulted in spurious "integer overflow" errors. Also, it was casting the
4161 result into an int without checking.
4163 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4164 did not work. The code looks as if it has *never* worked, though it appears
4165 to have been documented since at least releast 1.62. I have made it work.
4167 35. The "timeout_DNS" error in retry rules, also documented since at least
4168 1.62, also never worked. As it isn't clear exactly what this means, and
4169 clearly it isn't a major issue, I have abolished the feature by treating it
4170 as "timeout", and writing a warning to the main and panic logs.
4172 36. The display of retry rules for -brt wasn't always showing the error code
4175 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4176 timeout_connect_A, timeout_connect_MX.
4178 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4179 to the empty sender.
4181 39. The daemon was not analysing the content of -oX till after it had closed
4182 stderr and disconnected from the controlling terminal. This meant that any
4183 syntax errors were only noted on the panic log, and the return code from
4184 the command was 0. By re-arranging the code a little, I've made the
4185 decoding happen first, so such errors now appear on stderr, and the return
4186 code is 1. However, the actual setting up of the sockets still happens in
4187 the disconnected process, so errors there are still only recorded on the
4190 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4191 connections (as happens on some IP stacks) was logged at start up time as
4192 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4193 it from "IPv6 and IPv4", which means that two separate sockets are being
4196 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4197 says whether AF_INET or AF_INET6 was passed as an argument.
4199 42. Exiwhat output was messed up when time zones were included in log
4202 43. Exiwhat now gives more information about the daemon's listening ports,
4203 and whether -tls-on-connect was used.
4205 44. The "port" option of the smtp transport is now expanded.
4207 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4208 silently ignored. Now an error message is written to the main and panic
4211 46. There's a new ACL modifier called "logwrite" which writes to a log file
4212 as soon as it is encountered.
4214 47. Added $local_user_uid and $local_user_gid at routing time.
4216 48. Exim crashed when trying to verify a sender address that was being
4219 49. Exim was recognizing only a space character after ".include". It now also
4220 recognizes a tab character.
4222 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4223 extracting the relevant information from the specification. The man page no
4224 longer contains scrambled data for the -d option, and I've added a section
4225 at the front about calling Exim under different names.
4227 51. Added "extra_headers" argument to the "mail" command in filter files.
4229 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4232 53. Installed eximstats 1.29.
4234 54. Added transport_filter_timeout as a generic transport option.
4236 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4237 Cc: header lines. This was required by RFC 822, but it not required by RFC
4240 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4241 incoming messages that did not have them. Now it does so only if the
4242 message originates locally, that is, if there is no associated remote host
4243 address. When Resent- header lines are present, this applies to the Resent-
4244 lines rather than the non-Resent- lines.
4246 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4247 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4249 58. Messages for configuration errors now include the name of the main
4250 configuration file - useful now that there may be more than one file in a
4251 list (.included file names were always shown).
4253 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4254 for those rare installations that do not start the daemon as root or run it
4255 setuid root. I've cut out the call to initgroups() if the daemon is not
4258 60. The Exim user and group can now be bound into the binary as text strings
4259 that are looked up at the start of Exim's processing.
4261 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4263 62. Added $mailstore_basename variable.
4265 63. Installed patch to sieve.c from Michael Haardt.
4267 64. When Exim failed to open the panic log after failing to open the main log,
4268 the original message it was trying to log was written to stderr and debug
4269 output, but if they were not available (the usual case in production), it
4270 was lost. Now it is written to syslog before the two lines that record the
4271 failures to open the logs.
4273 65. Users' Exim filters run in subprocesses under the user's uid. It is
4274 possible for a "deliver" command or an alias in a "personal" command to
4275 provoke an address rewrite. If logging of address rewriting is configured,
4276 this fails because the process is not running as root or exim. There may be
4277 a better way of dealing with this, but for the moment (because 4.30 needs
4278 to be released), I have disabled address rewrite logging when running a
4279 filter in a non-root, non-exim process.
4285 1. The buildconfig auxiliary program wasn't quoting the value set for
4286 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4287 not defined. This bug was masked in 4.22 by the effect that was fixed in
4290 2. Some messages that were rejected after a message id was allocated were
4291 shown as "incomplete" by exigrep. It no longer does this for messages that
4292 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4294 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4295 have allow_domain_literals set, the ID did not get logged in the <= line.
4296 Domain literals are now always recognized in Message-ID: header lines.
4298 4. The first argument for a ${extract expansion item is the key name or field
4299 number. Leading and trailing spaces in this item were not being ignored,
4300 causing some misleading effects.
4302 5. When deliver_drop_privilege was set, single queue runner processes started
4303 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4304 same command in the process it spins off) were not dropping privilege.
4306 6. When the daemon running as "exim" started a queue runner, it always
4307 re-executed Exim in the spun-off process. This is a waste of effort when
4308 deliver_drop_privilege is set. The new process now just calls the
4309 queue-runner function directly.
4315 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4318 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4319 ignored. Though the use of -oP was forcing the writing of a pid file, it
4320 was always written to the default place.
4322 3. If the message "no IP address found for host xxxx" is generated during
4323 incoming verification, it is now followed by identification of the incoming
4324 connection (so you can more easily find what provoked it).
4326 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4328 5. Added some features to "harden" Exim a bit more against certain attacks:
4330 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4331 be put in Local/Makefile. This is like the never_users runtime option,
4332 but it cannot be overridden. The default setting is "root".
4334 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4335 prefix string with which any file named in a -C command line option
4338 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4339 is retained for -C and -D only if the caller of Exim is root. Without
4340 it, the exim user may also use -C and -D and retain privilege.
4342 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4343 command line option is disabled.
4345 6. Macro names set by the -D option must start with an upper case letter, just
4346 like macro names defined in the configuration file.
4348 7. Added "dereference=" facility to LDAP.
4350 8. Two instances of the typo "uknown" in the source files are fixed.
4352 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4353 the Configure-Makefile script screwed up while processing it.
4355 10. Incorporated PCRE 4.4.
4357 11. The SMTP synchronization check was not operating right at the start of an
4358 SMTP session. For example, it could not catch a HELO sent before the client
4359 waited for the greeting. There is now a check for outstanding input at the
4360 point when the greeting is written. Because of the duplex, asynchronous
4361 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4362 way, but not yet received, when the check is performed.
4364 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4365 on TCP/IP sockets, because this apparently causes some broken clients to
4368 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4369 unchanged) from the Cygwin maintainer.
4371 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4372 was supported instead of testing for mbx. Effectively a typo.
4374 15. The spa authenticator server code was not checking that the input it
4375 received was valid base64.
4377 16. The debug output line for the "set" modifier in ACLs was not showing the
4378 name of the variable that was being set.
4380 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4382 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4383 lookup on its own. Something else was needed to trigger the lookup. For
4384 example, a match in host_lookup or the need for a host name in a host list.
4385 Now, if $sender_host_name is referenced and the host name has not yet been
4386 looked up, a lookup is performed. If the lookup fails, the variable remains
4387 empty, and $host_lookup_failed is set to "1".
4389 19. Added "eqi" as a case-independent comparison operator.
4391 20. The saslauthd authentication condition could segfault if neither service
4392 nor realm was specified.
4394 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4395 error message that was logged was misleading, and incoming SMTP
4396 connections were dropped. The message is now more accurate, and temporary
4397 errors are given to SMTP connections.
4399 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4400 (except RSET) with a 421 error, until QUIT is received. However, it was
4401 failing to send a response to QUIT.
4403 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4404 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4405 ACL is now run after the helo_try_verify_hosts code.
4407 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4408 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4409 case-independent, so other case variants are also recognized). Apparently
4410 some systems use these upper case variants.
4412 25. If more than two messages were waiting for the same host, and a transport
4413 filter was specified for the transport, Exim sent two messages over the
4414 same TCP/IP connection, and then failed with "socket operation on non-
4415 socket" when it tried to send the third.
4417 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4419 27. The extern definition of crypt16() in expand.c was not being excluded when
4420 the OS had its own crypt16() function.
4422 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4423 as a preferred synonym for return_size_limit, both as an option and as an
4426 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4428 30. Changed the default configuration ACL to relax the local part checking rule
4429 for addresses that are not in any local domains. For these addresses,
4430 slashes and pipe symbols are allowed within local parts, but the sequence
4431 /../ is explicitly forbidden.
4433 31. SPA server authentication was not clearing the challenge buffer before
4436 32. log_message in a "warn" ACL statement was writing to the reject log as
4437 well as to the main log, which contradicts the documentation and doesn't
4438 seem right (because no rejection is happening). So I have stopped it.
4440 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4441 However, I am unable to do any testing of this.
4443 34. Fixed an infelicity in the appendfile transport. When checking directories
4444 for a mailbox, to see if any needed to be created, it was accidentally
4445 using path names with one or more superfluous leading slashes; tracing
4446 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4448 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4449 contents are added to the log line that is written for every discarded
4450 recipient. (Previously a log_message setting was ignored.)
4452 36. The ${quote: operator now quotes the string if it is empty.
4454 37. The install script runs exim in order to find its version number. If for
4455 some reason other than non-existence or emptiness, which it checks, it
4456 could not run './exim', it was installing it with an empty version number,
4457 i.e. as "exim-". This error state is now caught, and the installation is
4460 38. An argument was missing from the function that creates an error message
4461 when Exim fails to connect to the socket for saslauthd authentication.
4462 This could cause Exim to crash, or give a corrupted message.
4464 39. Added isip, isip4, and isip6 to ${if conditions.
4466 40. The ACL variables $acl_xx are now saved with the message, and can be
4467 accessed later in routers, transports, and filters.
4469 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4470 strings in the file are not string-expanded.
4472 42. If a MAIL command specified a SIZE value that was too large to fit into an
4473 int variable, the check against message_size_limit failed. Such values are
4474 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4475 day this will have to be increased, but I don't think I want to be around
4476 when emails are that large.
4483 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4484 iconv() is not standard in FreeBSD.
4486 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4487 IPv6 enabled. The observed symptom was a segmentation fault on return from
4488 the function os_common_find_running_interfaces() in src/os.c.
4490 3. In the check_special_case() function in daemon.c I had used "errno" as an
4491 argument name, which causes warnings on some systems. This was basically a
4492 typo, since it was named "eno" in the comments!
4494 4. The code that waits for the clock to tick (at a resolution of some fraction
4495 of a second) so as to ensure message-id uniqueness was always waiting for
4496 at least one whole tick, when it could have waited for less. [This is
4497 almost certainly not relevant at current processor speeds, where it is
4498 unlikely to ever wait at all. But we try to future-proof.]
4500 5. The function that sleeps for a time interval that includes fractions of a
4501 second contained a race. It did not block SIGALRM between setting the
4502 timer, and suspending (a couple of lines later). If the interval was short
4503 and the sigsuspend() was delayed until after it had expired, the suspension
4504 never ended. On busy systems this could lead to processes getting stuck for
4507 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4508 process, before it forks any delivery processes. The open lookup caching
4509 mechanism meant that the open file or database connection was passed into
4510 the delivery process. The problem was that delivery processes always tidy
4511 up cached lookup data. This could cause a problem for the next delivery
4512 process started by the queue runner, because the external queue runner
4513 process does not know about the closure. So the next delivery process
4514 still has data in the lookup cache. In the case of a file lookup, there was
4515 no problem because closing a file descriptor in a subprocess doesn't affect
4516 the parent. However, if the lookup was caching a connection to a database,
4517 the connection was closed, and the second delivery process was likely to
4518 see errors such as "PGSQL: query failed: server closed the connection
4519 unexpectedly". The problem has been fixed by closing all cached lookups
4520 in a queue runner before running a delivery process.
4522 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4523 seem to have the "const" qualifier which it has on other OS. I've
4526 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4527 *of the same type* (client or server) with the same public name that an
4528 error should be diagnosed.
4530 9. When Exim looked up a host name for an IP address, but failed to find the
4531 original IP address when looking up the host name (a safety check), it
4532 output the message "<ip address> does not match any IP for NULL", which was
4533 confusing, to say the least. The bug was that the host name should have
4534 appeared instead of "NULL".
4536 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4537 user that is built into the binary, and the -C or -D options is used, root
4538 privilege is dropped before the configuration file is read. In addition,
4539 logging is switched to stderr instead of the normal log files. If the
4540 configuration then re-defines the Exim user, the unprivileged environment
4541 is probably not what is expected, so Exim logs a panic warning message (but
4544 However, if deliver_drop_privilege is set, the unprivileged state may well
4545 be exactly what is intended, so the warning has been cut out in that case,
4546 and Exim is allowed to try to write to its normal log files.
4552 1. smtp_return_error_details was not giving details for temporary sender
4553 or receiver verification errors.
4555 2. Diagnose a configuration error if two authenticators have the same public
4558 3. Exim used not to create the message log file for a message until the first
4559 delivery attempt. This could be confusing when incoming messages were held
4560 for policy or load reasons. The message log file is now created at the time
4561 the message is received, and an initial "Received" line is written to it.
4563 4. The automatically generated man page for command line options had a minor
4564 bug that caused no ill effects; however, a more serious problem was that
4565 the procedure for building the man page automatically didn't always
4566 operate. Consequently, release 4.20 contains an out-of-date version. This
4567 shouldn't happen again.
4569 5. When building Exim with embedded Perl support, the script that builds the
4570 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4571 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4573 6. The freeze_tell option was not being used for messages that were frozen on
4574 arrival, either by an ACL or by local_scan().
4576 7. Added the smtp_incomplete_transaction log selector.
4578 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4579 was accepting AUTH without a new EHLO.
4581 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4582 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4585 10. Logging of TCP/IP connections (when configured) now happens in the main
4586 daemon process instead of the child process, so that the TCP/IP connection
4587 count is more accurate (but it can never be perfect).
4589 11. The use of "drop" in a nested ACL was not being handled correctly in the
4590 outer ACL. Now, if condition failure induced by the nested "drop" causes
4591 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4592 or "require"), the connection is dropped.
4594 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4595 that yield "discard" can now be used with an "accept" or a "discard" verb,
4596 but an error is generated for any others (because I can't see a useful way
4597 to define what should happen).
4599 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4600 are now processed in the same way as lines in the Exim configuration file.
4601 In particular, continuation lines are supported.
4603 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4605 15. Added -ti meaning -t -i.
4607 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4608 domains, and warn by logging if others are found.
4610 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4611 returned from GIFCONF when Exim is trying to find the list of interfaces on
4612 a host. The code in os.c has been modified to copy each ifreq to an aligned
4613 structure in all cases.
4615 Also, in some cases, the returned ifreq's were being copied to a 'struct
4616 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4617 means the last couple of bytes of an IPv6 address could be chopped if the
4618 ifreq contained only a normal sockaddr (14 bytes storage).
4620 18. Named domain lists were not supported in the hosts_treat_as_local option.
4621 An entry such as +xxxx was not recognized, and was treated as a literal
4624 19. Ensure that header lines added by a DATA ACL are included in the reject log
4625 if the ACL subsequently rejects the message.
4627 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4628 MD5 (which is deprecated).
4630 21. When testing a filter file using -bf, Exim was writing a message when it
4631 took the sender from a "From " line in the message, but it was not doing so
4632 when it took $return_path from a Return-Path: header line. It now does.
4634 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4635 with a valid header line field name (a series of printing characters
4636 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4638 23. Changed "disc" in the source to "disk" to conform to the documentation and
4639 the book and for uniformity.
4641 24. Ignore Sendmail's -Ooption=value command line item.
4643 25. When execve() failed while trying to run a command in a pipe transport,
4644 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4645 could be confused with a return value of 69 from the command itself. This
4646 has been changed to 127, the value the shell returns if it is asked to run
4647 a non-existent command. The wording for the related log line suggests a
4648 non-existent command as the problem.
4650 26. If received_header_text expands to an empty string, do not add a Received:
4651 header line to the message. (Well, it adds a token one on the spool, but
4652 marks it "old" so that it doesn't get used or transmitted.)
4654 27. Installed eximstats 1.28 (addition of -nt option).
4656 28. There was no check for failure on the call to getsockname() in the daemon
4657 code. This can fail if there is a shortage of resources on the system, with
4658 ENOMEM, for example. A temporary error is now given on failure.
4660 29. Contrary to the C standard, it seems that in some environments, the
4661 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4662 program. Exim now does this explicitly; it affects the formatting of
4663 timestamps using strftime().
4665 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4666 complaints. I've now initialized all the variables, to avoid this.
4668 32. Header lines added by a system filter were not being "seen" during
4669 transport-time rewrites.
4671 33. The info_callback() function passed to OpenSSL is set up with type void
4672 (*)(SSL *, int, int), as described somewhere. However, when calling the
4673 function (actually a macro) that sets it up, the type void(*)() is
4674 expected. I've put in a cast to prevent warnings from picky compilers.
4676 34. If a DNS black list lookup found a CNAME record, but there were no A
4677 records associated with the domain it pointed at, Exim crashed.
4679 35. If a DNS black list lookup returned more than one A record, Exim ignored
4680 all but the first. It now scans all returned addresses if a particular IP
4681 value is being sought. In this situation, the contents of the
4682 $dnslist_value variable are a list of all the addresses, separated by a
4685 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4686 to accept a host name and all its aliases if the forward lookup for any of
4687 them yielded the IP address of the incoming connection. Now it accepts only
4688 those names whose forward lookup yields the correct IP address. Any other
4689 names are discarded. This closes a loophole whereby a rogue DNS
4690 administrator could create reverse DNS records to break through a
4691 wildcarded host restriction in an ACL.
4693 37. If a user filter or a system filter that ran in a subprocess used any of
4694 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4695 the wrong values were passed to the pipe command ($thisaddress had the
4696 value of $0, $0 had the value of $1, etc). This bug was introduced by
4697 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4699 38. Improved the line breaking for long SMTP error messages from ACLs.
4700 Previously, if there was no break point between 40 and 75 characters, Exim
4701 left the rest of the message alone. Two changes have been made: (a) I've
4702 reduced the minimum length to 35 characters; (b) if it can't find a break
4703 point between 35 and 75 characters, it looks ahead and uses the first one
4704 that it finds. This may give the occasional overlong line, but at least the
4705 remaining text gets split now.
4707 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4708 file descriptors might be low, and that setting 1000 would always raise it.
4709 It turns out that in some environments, the limit is already over 1000 and
4710 that lowering it causes trouble. So now Exim takes care not to decrease it.
4712 40. When delivering a message, the value of $return_path is set to $sender_
4713 address at the start of routing (routers may change the value). By an
4714 oversight, this default was not being set up when an address was tested by
4715 -bt or -bv, which affected the outcome if any router or filter referred to
4718 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4719 log only when "message" or "log_message" are set. However, if one of the
4720 conditions was an address verification, or a call to a nested ACL, the
4721 messages generated by the underlying test were being passed through. This
4722 no longer happens. The underlying message is available in $acl_verify_
4723 message for both "message" and "log_message" expansions, so it can be
4724 passed through if needed.
4726 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4727 new expansion $bh_ to give the encoded byte string without charset
4728 translation. Translation happens only if iconv() is available; HAVE_ICONV
4729 indicates this at build time. HEADERS_CHARSET gives the charset to
4730 translate to; headers_charset can change it in the configuration, and
4731 "headers charset" can change it in an individual filter file.
4733 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4734 that creates RFC 2047 encoded "words" labels them as that charset instead
4735 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4736 expansion operator; (ii) when Exim creates a From: line for a local
4737 message; (iii) when a header line is rewritten to include a "phrase" part.
4739 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4740 buggy, causing it to skip the first lines of messages whose message ID
4741 ended in 'D'. This would not have bitten before Exim release 4.14, because
4742 message IDs were unlikely to end in 'D' before then. The effect was to have
4743 incorrect size information for certain domains.
4745 45. #include "config.h" was missing at the start of the crypt16.c module. This
4746 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4749 46. If there was a timeout during a "random" callout check, Exim treated it as
4750 a failure of the random address, and carried on sending RSET and the real
4751 address. If the delay was just some slowness somewhere, the response to the
4752 original RCPT would be taken as a response to RSET and so on, causing
4753 mayhem of various kinds.
4755 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4756 when I implemented it. It didn't allow for the fact that some option values
4757 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4758 the right test for positive values.
4760 48. Domain names in DNS records are case-independent. Exim always looks them up
4761 in lower case. Some resolvers return domain names in exactly the case they
4762 appear in the zone file, that is, they may contain uppercase letters. Not
4763 all resolvers do this - some return always lower case. Exim was treating a
4764 change of case by a resolver as a change of domain, similar to a widening
4765 of a domain abbreviation. This triggered its re-routing code and so it was
4766 trying to route what was effectively the same domain again. This normally
4767 caused routing to fail (because the router wouldn't handle the domain
4768 twice). Now Exim checks for this case specially, and just changes the
4769 casing of the domain that it ultimately uses when it transmits the message
4772 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4775 50. If a filter generated a file delivery with a non-absolute name (possible if
4776 no home directory exists for the router), the forbid_file option was not
4779 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4780 the existing equality matching.
4782 52. Exim was using ints instead of ino_t variables in some places where it was
4783 dealing with inode numbers.
4785 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4786 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4787 called TMPDIR, and if it finds it is different, it changes its value.
4789 54. The smtp_printf() function is now made available to local_scan() so
4790 additional output lines can be written before returning. There is also an
4791 smtp_fflush() function to enable the detection of a dropped connection.
4792 The variables smtp_input and smtp_batched_input are exported to
4795 55. Changed the default runtime configuration: the message "Unknown user"
4796 has been removed from the ACL, and instead placed on the localuser router,
4797 using the cannot_route_message feature. This means that any verification
4798 failures that generate their own messages won't get overridden. Similarly,
4799 the "Unrouteable address" message that was in the ACL for unverifiable
4800 relay addresses has also been removed.
4802 56. Added hosts_avoid_esmtp to the smtp transport.
4804 57. The exicyclog script was not checking for the esoteric option
4805 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4806 will work only if exicyclog is run under the appropriate euid.
4808 58. Following a discussion on the list, the rules by which Exim recognises line
4809 endings on incoming messages have been changed. The -dropcr and drop_cr
4810 options are now no-ops, retained only for backwards compatibility. The
4811 following line terminators are recognized: LF CRLF CR. However, special
4812 processing applies to CR:
4814 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4815 nor a local message in the state where . is a terminator.
4817 (ii) If a bare CR is encountered in a header line, an extra space is added
4818 after the line terminator so as not to end the header. The reasoning
4819 behind this is that bare CRs in header lines are most likely either
4820 to be mistakes, or people trying to play silly games.
4822 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4823 was being incorrectly given as 18 bytes larger than it should have been.
4824 This is a VOB (very old bug).
4826 60. This may never have affected anything current, but just in case it has:
4827 When the local host is found other than at the start of a list of hosts,
4828 the local host, those with the same MX, and any that follow, are discarded.
4829 When the list in question was part of a longer list of hosts, the following
4830 hosts (not currently being processed) were also being discarded. This no
4831 longer happens. I'm not sure if this situation could ever has previously
4834 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4837 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4838 header. This is something that is recommended in a new Internet Draft, and
4839 is something that is documented as being done by Sendmail. There are two
4840 possible values. For messages generated by the autoreply transport, Exim
4843 Auto-Submitted: auto-replied
4845 whereas for all other generated messages (e.g. bounces) it adds
4847 Auto-Submitted: auto-generated
4849 63. The "personal" condition in filters now includes a test for the
4850 Auto-Submitted: header. If it contains the string "auto-" the message it
4851 not considered personal.
4853 64. Added rcpt_include_affixes as a generic transport option.
4855 65. Added queue_only_override (default true).
4857 66. Added the syslog_duplication option.
4859 67. If what should have been the first header line of a message consisted of
4860 a space followed by a colon, Exim was mis-interpreting it as a header line.
4861 It isn't of course - it is syntactically invalid and should therefore be
4862 treated as the start of the message body. The misbehaviour could have
4863 caused a number of strange effects, including loss of data in subsequent
4864 header lines, and spool format errors.
4866 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4867 client host had authenticated. This control can now be exercised by an ACL
4868 for more flexibility.
4870 69. By default, callouts do not happen when testing with -bh. There is now a
4871 variant, -bhc, which does actually run the callout code, including
4872 consulting and updating the callout cache.
4874 70. Added support for saslauthd authentication, courtesy of Alexander
4877 71. If statvfs() failed on the spool or log directories while checking their
4878 size for availability, Exim confusingly gave the error "space shortage".
4879 Furthermore, in debugging mode it crashed with a floating point exception.
4880 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4881 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4882 serious problem, Exim now writes to the main and panic logs when this
4883 happens, with details of the failure. It then refuses to accept the
4884 incoming message, giving the message "spool directory problem" or "log
4885 directory problem" with a 421 code for SMTP messages.
4887 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4888 0, 1, and 2 exist, because some OS complain for execs without them (see
4889 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4890 descriptors. However, the code omitted to check that the open succeeded,
4891 causing mysterious errors if for some reason the permissions on /dev/null
4892 got screwed. Now Exim writes a message to the main and panic logs, and
4893 bombs out if it can't open /dev/null.
4895 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4896 interact so that it is all more flexible. It is supposed to remain
4897 backwards compatible. Also added extra_local_interfaces.
4899 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4900 to bomb out with an assertion failure - to the client this appears as a
4901 connection drop. This problem occurs in the part of the code that was taken
4902 from the Samba project. Fortunately, the assertion is in a very simple
4903 function, so I have fixed this by reproducing the function inline in the
4904 one place where it is called, and arranging for authentication to fail
4905 instead of killing the process with assert().
4907 75. The SPA client code was not working when the server requested OEM rather
4908 than Unicode encoding.
4910 76. Added code to make require_files with a specific uid setting more usable in
4911 the case where statting the file as root fails - usually a non-root-mounted
4912 NFS file system. When this happens and the failure is EACCES, Exim now
4913 forks a subprocess and does the per-uid checking as the relevant uid.
4915 77. Added process_log_path.
4917 78. If log_file_path was not explicitly set, a setting of check_log_space or
4918 check_log_inodes was ignored.
4920 79. If a space check for the spool or log partitions fails, the incident is now
4921 logged. Of course, in the latter case the data may get lost...
4923 80. Added the %p formatting code to string_format() so that it can be used to
4924 print addresses in debug_print(). Adjusted all the address printing in the
4925 debugging in store.c to use %p rather than %d.
4927 81. There was a concern that a line of code in smtp_in.c could overflow a
4928 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4929 initially expressed, the concern was not well-founded, because trailing
4930 spaces are removed early. However, if the trailing spaces were followed by
4931 a NULL, they did not get removed, so the overflow was possible. Two fixes
4934 (a) I re-wrote the offending code in a cleaner fashion.
4935 (b) If an incoming SMTP command contains a NULL character, it is rejected
4938 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4939 runs initgroups(), so that if the Exim user is in any additional groups,
4940 they will be used during message reception.
4946 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob