1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.563 2009/06/11 14:07:57 tom 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
78 acl_smtp_notquit, added index entry.
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
86 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
87 Patch provided by Phil Pennock
89 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
91 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
92 Patch provided by Brad "anomie" Jorsch
94 NM/14 Bugzilla 614: Added timeout setting to transport filter.
95 Patch provided by Dean Brooks
97 TK/05 Add native DKIM support (does not depend on external libraries).
103 TK/01 Add preliminary DKIM support. Currently requires a forked version of
104 ALT-N's libdkim that I have put here:
105 http://duncanthrax.net/exim-experimental/
107 Note to Michael Haardt: I had to rename some vars in sieve.c. They
108 were called 'true' and it seems that C99 defines that as a reserved
109 keyword to be used with 'bool' variable types. That means you could
110 not include C99-style headers which use bools without triggering
111 build errors in sieve.c.
113 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
114 as mailq or other aliases. Changed the --help handling significantly
115 to do whats expected. exim_usage() emits usage/help information.
117 SC/01 Added the -bylocaldomain option to eximstats.
119 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
121 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
123 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
129 PH/01 Another patch from the Sieve maintainer.
131 PH/02 When an IPv6 address is converted to a string for single-key lookup
132 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
133 dots are used instead of colons so that keys in lsearch files need not
134 contain colons. This was done some time before quoting was made available
135 in lsearch files. However, iplsearch files do require colons in IPv6 keys
136 (notated using the quote facility) so as to distinguish them from IPv4
137 keys. This meant that lookups for IP addresses in host lists did not work
138 for iplsearch lookups.
140 This has been fixed by arranging for IPv6 addresses to be expressed with
141 colons if the lookup type is iplsearch. This is not incompatible, because
142 previously such lookups could never work.
144 The situation is now rather anomolous, since one *can* have colons in
145 ordinary lsearch keys. However, making the change in all cases is
146 incompatible and would probably break a number of configurations.
148 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
151 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
152 conversion specification without a maximum field width, thereby enabling
153 a rogue spamd server to cause a buffer overflow. While nobody in their
154 right mind would setup Exim to query an untrusted spamd server, an
155 attacker that gains access to a server running spamd could potentially
156 exploit this vulnerability to run arbitrary code as the Exim user.
158 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
159 $primary_hostname instead of what libspf2 thinks the hosts name is.
161 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
162 a directory entry by the name of the lookup key. Previously, if a
163 symlink pointed to a non-existing file or a file in a directory that
164 Exim lacked permissions to read, a lookup for a key matching that
165 symlink would fail. Now it is enough that a matching directory entry
166 exists, symlink or not. (Bugzilla 503.)
168 PH/03 The body_linecount and body_zerocount variables are now exported in the
171 PH/04 Added the $dnslist_matched variable.
173 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
174 This means they are set thereafter only if the connection becomes
177 PH/06 Added the client_condition to authenticators so that some can be skipped
178 by clients under certain conditions.
180 PH/07 The error message for a badly-placed control=no_multiline_responses left
181 "_responses" off the end of the name.
183 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
185 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
186 (without spaces) instead of just copying the configuration text.
188 PH/10 Added the /noupdate option to the ratelimit ACL condition.
190 PH/11 Added $max_received_linelength.
192 PH/12 Added +ignore_defer and +include_defer to host lists.
194 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
195 way in which PCRE > 7.0 is built.
197 PH/14 Implemented queue_only_load_latch.
199 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
200 MAIL command. The effect was to mangle the value on 64-bit systems.
202 PH/16 Another patch from the Sieve maintainer.
204 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
206 PH/18 If a system quota error occurred while trying to create the file for
207 a maildir delivery, the message "Mailbox is full" was not appended to the
208 bounce if the delivery eventually timed out. Change 4.67/27 below applied
209 only to a quota excession during the actual writing of the file.
211 PH/19 It seems that peer DN values may contain newlines (and other non-printing
212 characters?) which causes problems in log lines. The DN values are now
213 passed through string_printing() before being added to log lines.
215 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
216 and InterBase are left for another time.)
218 PH/21 Added message_body_newlines option.
220 PH/22 Guard against possible overflow in moan_check_errorcopy().
222 PH/23 POSIX allows open() to be a macro; guard against that.
224 PH/24 If the recipient of an error message contained an @ in the local part
225 (suitably quoted, of course), incorrect values were put in $domain and
226 $local_part during the evaluation of errors_copy.
232 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
233 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
236 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
237 issue a MAIL command.
239 PH/02 In an ACL statement such as
241 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
243 if a client was not listed at all, or was listed with a value other than
244 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
245 the condition was not true (as it should be), so access was not denied.
246 The bug was that the ! inversion was incorrectly passed on to the second
247 item. This has been fixed.
249 PH/03 Added additional dnslists conditions == and =& which are different from
250 = and & when the dns lookup returns more than one IP address.
252 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
253 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
255 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
256 FSYNC, which compiles an option called disable_fsync that allows for
257 bypassing fsync(). The documentation is heavily laced with warnings.
259 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
261 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
262 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
263 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
264 including adding "make clean"; (3) Added -fPIC when compiling the test
265 dynamically loaded module, to get rid of a warning.
267 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
268 message fails, move_frozen_messages = true and ignore_bounce_errors_after
269 = 0s. The bug is otherwise harmless.
271 PH/07 There was a bug in the dovecot authenticator such that the value of
272 $auth1 could be overwritten, and so not correctly preserved, after a
273 successful authentication. This usually meant that the value preserved by
274 the server_setid option was incorrect.
276 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
278 PH/09 Installed PCRE release 7.0.
280 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
281 run for batched SMTP input. It is now run at the start of every message
282 in the batch. While fixing this I discovered that the process information
283 (output by running exiwhat) was not always getting set for -bs and -bS
284 input. This is fixed, and it now also says "batched" for BSMTP.
286 PH/11 Added control=no_pipelining.
288 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
289 patch, slightly modified), and move the expansion of helo_data till after
290 the connection is made in the smtp transport (so it can use these
293 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
295 PH/14 Added log_selector = +pid.
297 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
299 PH/16 Add ${if forany and ${if forall.
301 PH/17 Added dsn_from option to vary the From: line in DSNs.
303 PH/18 Flush SMTP output before performing a callout, unless control =
304 no_callout_flush is set.
306 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
307 was true (the default) a successful delivery failed to delete the retry
308 item, thus causing premature timeout of the address. The bug is now
311 PH/20 Added hosts_avoid_pipelining to the smtp transport.
313 PH/21 Long custom messages for fakedefer and fakereject are now split up
314 into multiline reponses in the same way that messages for "deny" and
315 other ACL rejections are.
317 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
318 with slight modification.
320 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
321 draft, changing the syntax and factoring some duplicate code".
323 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
324 for deliveries of the second and subsequent messages over the same SMTP
327 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
328 ${reduce, with only minor "tidies".
330 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
332 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
333 expansion side effects.
335 PH/27 When a message times out after an over-quota error from an Exim-imposed
336 quota, the bounce message says "mailbox is full". This message was not
337 being given when it was a system quota that was exceeded. It now should
340 MH/03 Made $recipients available in local_scan(). local_scan() already has
341 better access to the recipient list through recipients_list[], but
342 $recipients can be useful in postmaster-provided expansion strings.
344 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
345 in the case of a MAIL command with additional options following the
346 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
347 were accidentally chopped off.
349 PH/29 SMTP synchronization checks are implemented when a command is read -
350 there is a check that no more input is waiting when there shouldn't be
351 any. However, for some commands, a delay in an ACL can mean that it is
352 some time before the response is written. In this time, more input might
353 arrive, invalidly. So now there are extra checks after an ACL has run for
354 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
355 pipelining has not been advertised.
357 PH/30 MH's patch to allow iscntrl() characters to be list separators.
359 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
360 returned in the SMTP response when smtp_return_error_details was false.
363 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
364 instead of the C I/O that was originally supplied, because problems were
367 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
368 Exim which did not show up earlier: it was assuming that a call to
369 SSL_CTX_set_info_callback() might give an error value. In fact, there is
370 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
371 was a macro that became an assignment, so it seemed to work. This has
372 changed to a proper function call with a void return, hence the compile
373 error. Exim's code has been fixed.
375 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
378 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
380 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
382 PH/37 If a message is not accepted after it has had an id assigned (e.g.
383 because it turns out to be too big or there is a timeout) there is no
384 "Completed" line in the log. When some messages of this type were
385 selected by exigrep, they were listed as "not completed". Others were
386 picked up by some special patterns. I have improved the selection
387 criteria to be more general.
389 PH/38 The host_find_failed option in the manualroute router can now be set
390 to "ignore", to completely ignore a host whose IP address cannot be
391 found. If all hosts are ignored, the behaviour is controlled by the new
392 host_all_ignored option.
394 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
395 homing or because of multiple MX records with /mx) generated more than
396 one IP address, and the following item turned out to be the local host,
397 all the secondary addresses of the first item were incorrectly removed
398 from the list, along with the local host and any following hosts (which
399 is what is supposed to happen).
401 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
402 whoever called Exim into the -H file. In the case of the daemon it was
403 behaving confusingly. When first started, it used values for whoever
404 started the daemon, but after a SIGHUP it used the Exim user (because it
405 calls itself on a restart). I have changed the code so that it now always
408 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
409 message are rejected with the same error (e.g. no authentication or bad
410 sender address), and a DATA command is nevertheless sent (as can happen
411 with PIPELINING or a stupid MUA), the error message that was given to the
412 RCPT commands is included in the rejection of the DATA command. This is
413 intended to be helpful for MUAs that show only the final error to their
416 PH/42 Another patch from the Sieve maintainer.
418 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
419 Eximstats - Fixed some broken HTML links and added missing column headers
421 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
422 columns for Rejects, Temp Rejects, Ham, and Spam rows.
424 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
426 PH/43 Yet another patch from the Sieve maintainer.
428 PH/44 I found a way to check for a TCP/IP connection going away before sending
429 the response to the final '.' that terminates a message, but only in the
430 case where the client has not sent further data following the '.'
431 (unfortunately, this is allowed). However, in many cases there won't be
432 any further data because there won't be any more messages to send. A call
433 to select() can be used: if it shows that the input is "ready", there is
434 either input waiting, or the socket has been closed. An attempt to read
435 the next input character can distinguish the two cases. Previously, Exim
436 would have sent an OK response which the client would never have see.
437 This could lead to message repetition. This fix should cure that, at
438 least in a lot of common cases.
440 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
441 advertised in response to EHLO.
447 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
448 fixed by 4.65/MH/01 (is this a record?) are fixed:
450 (i) An empty string was always treated as zero by the numeric comparison
451 operators. This behaviour has been restored.
453 (ii) It is documented that the numeric comparison operators always treat
454 their arguments as decimal numbers. This was broken in that numbers
455 starting with 0 were being interpreted as octal.
457 While fixing these problems I realized that there was another issue that
458 hadn't been noticed. Values of message_size_limit (both the global option
459 and the transport option) were treated as octal if they started with 0.
460 The documentation was vague. These values are now always treated as
461 decimal, and I will make that clear in the documentation.
467 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
468 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
471 MH/01 Don't check that the operands of numeric comparison operators are
472 integers when their expansion is in "skipping" mode (fixes bug
473 introduced by 4.64-PH/07).
475 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
476 child addresses, Exim now panics and dies. Previously, because the count
477 is held in a short int, deliveries were likely to be lost. As such a
478 large number of recipients for a single message is ridiculous
479 (performance will be very, very poor), I have chosen to impose a limit
480 rather than extend the field.
486 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
487 leftover -K file (the existence of which was triggered by #402).
488 While we were at it, introduced process PID as part of the -K
489 filename. This should rule out race conditions when creating
492 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
493 processing considerably. Previous code took too long for large mails,
494 triggering a timeout which in turn triggers #401.
496 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
497 in the DK code in transports.c. sendfile() is not really portable,
498 hence the _LINUX specificness.
500 TF/01 In the add_headers option to the mail command in an Exim filter,
501 there was a bug that Exim would claim a syntax error in any
502 header after the first one which had an odd number of characters
505 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
506 callout verification, Exim cached a "reject" for the entire domain. This
507 is correct for most verifications, but it is not correct for a recipient
508 verification with use_sender or use_postmaster set, because in that case
509 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
510 case of MAIL FROM:<> rejection from other early rejections (e.g.
511 rejection of HELO). When verifying a recipient using a non-null MAIL
512 address, the cache is ignored if it shows MAIL FROM:<> rejection.
513 Whatever the result of the callout, the value of the domain cache is
514 left unchanged (for any other kind of callout, getting as far as trying
515 RCPT means that the domain itself is ok).
517 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
520 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
521 manifest itself as EPIPE rather than ECONNECT. When tidying away a
522 session, the daemon ignores ECONNECT errors and logs others; it now
523 ignores EPIPE as well.
525 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
526 (quoted-printable decoding).
528 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
529 later the small subsequent patch to fix an introduced bug.
531 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
533 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
535 PH/08 An error is now given if message_size_limit is specified negative.
537 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
538 to be given (somewhat) arbitrary names.
540 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
543 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
544 miscellaneous code fixes
546 PH/10 Added the log_reject_target ACL modifier to specify where to log
549 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
550 hostname. This is wrong, because it relates to the incoming message (and
551 probably the interface on which it is arriving) and not to the outgoing
552 callout (which could be using a different interface). This has been
553 changed to use the value of the helo_data option from the smtp transport
554 instead - this is what is used when a message is actually being sent. If
555 there is no remote transport (possible with a router that sets up host
556 addresses), $smtp_active_hostname is used.
558 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
559 tweaks were necessary in order to get it to work (see also 21 below):
560 (a) The code assumed that strncpy() returns a negative number on buffer
561 overflow, which isn't the case. Replaced with Exim's string_format()
563 (b) There were several signed/unsigned issues. I just did the minimum
564 hacking in of casts. There is scope for a larger refactoring.
565 (c) The code used strcasecmp() which is not a standard C function.
566 Replaced with Exim's strcmpic() function.
567 (d) The code set only $1; it now sets $auth1 as well.
568 (e) A simple test gave the error "authentication client didn't specify
569 service in request". It would seem that Dovecot has changed its
570 interface. Fortunately there's a specification; I followed it and
571 changed what the client sends and it appears to be working now.
573 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
576 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
577 address A is aliased to B and C, where B exists and C does not. Without
578 -v the output is "A verified" because verification stops after a
579 successful redirection if more than one address is generated. However,
580 with -v the child addresses are also verified. Exim was outputting "A
581 failed to verify" and then showing the successful verification for C,
582 with its parentage. It now outputs "B failed to verify", showing B's
583 parentage before showing the successful verification of C.
585 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
586 look up a TXT record in a specific list after matching in a combined
589 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
590 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
591 they consult the DNS. I had assumed they would set it the way they
592 wanted; and indeed my experiments on Linux seem to show that in some
593 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
594 To be on the safe side, however, I have now made the interface to
595 host_find_byname() similar to host_find_bydns(), with an argument
596 containing the DNS resolver options. The host_find_byname() function now
597 sets these options at its start, just as host_find_bydns() does. The smtp
598 transport options dns_qualify_single and dns_search_parents are passed to
599 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
600 of host_find_byname() use the default settings of RES_DEFNAMES
601 (qualify_single) but not RES_DNSRCH (search_parents).
603 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
604 spool_read_header() do less string testing, by means of a preliminary
605 switch on the second character of optional "-foo" lines. (This is
606 overdue, caused by the large number of possibilities that now exist.
607 Originally there were few.) While I was there, I also converted the
608 str(n)cmp tests so they don't re-test the leading "-" and the first
609 character, in the hope this might squeeze out yet more improvement.
611 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
612 flag allowing group syntax was set by the header_syntax check but not
613 turned off, possible causing trouble later; (2) The flag was not being
614 set at all for the header_verify test, causing "group"-style headers to
615 be rejected. I have now set it in this case, and also caused header_
616 verify to ignore an empty address taken from a group. While doing this, I
617 came across some other cases where the code for allowing group syntax
618 while scanning a header line wasn't quite right (mostly, not resetting
619 the flag correctly in the right place). These bugs could have caused
620 trouble for malformed header lines. I hope it is now all correct.
622 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
623 with the "reply" argument non-NULL. The code, however (which originally
624 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
625 but it didn't always do it. This confused somebody who was copying the
626 code for some other use. I have removed all the tests.
628 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
629 feature that was used to support insecure browsers during the U.S. crypto
630 embargo. It requires special client support, and Exim is probably the
631 only MTA that supported it -- and would never use it because real RSA is
632 always available. This code has been removed, because it had the bad
633 effect of slowing Exim down by computing (never used) parameters for the
634 RSA_EXPORT functionality.
636 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
637 authenticator to fail if there's a tab character in the incoming data
638 (there should never be unless someone is messing about, as it's supposed
639 to be base64-encoded). Also added, on Timo's advice, the "secured" option
640 if the connection is using TLS or if the remote IP is the same as the
641 local IP, and the "valid-client-cert option" if a client certificate has
644 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
645 authenticators. This can be used for authorization after authentication
646 succeeds. (In the case of plaintext, it servers for both authentication
649 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
650 if any retry times were supplied.
652 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
653 connection, where there is no client IP address to check. In this
654 situation, the verify now always succeeds.
656 PH/25 Applied John Jetmore's -Mset patch.
658 PH/26 Added -bem to be like -Mset, but loading a message from a file.
660 PH/27 In a string expansion for a processed (not raw) header when multiple
661 headers of the same name were present, leading whitespace was being
662 removed from all of them, but trailing whitespace was being removed only
663 from the last one. Now trailing whitespace is removed from each header
664 before concatenation. Completely empty headers in a concatenation (as
667 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
668 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
670 PH/29 [Removed. This was a change that I later backed out, and forgot to
671 correct the ChangeLog entry (that I had efficiently created) before
672 committing the later change.]
674 PH/30 Exim was sometimes attempting to deliver messages that had suffered
675 address errors (4xx response to RCPT) over the same connection as other
676 messages routed to the same hosts. Such deliveries are always "forced",
677 so retry times are not inspected. This resulted in far too many retries
678 for the affected addresses. The effect occurred only when there were more
679 hosts than the hosts_max_try setting in the smtp transport when it had
680 the 4xx errors. Those hosts that it had tried were not added to the list
681 of hosts for which the message was waiting, so if all were tried, there
682 was no problem. Two fixes have been applied:
684 (i) If there are any address or message errors in an SMTP delivery, none
685 of the hosts (tried or untried) are now added to the list of hosts
686 for which the message is waiting, so the message should not be a
687 candidate for sending over the same connection that was used for a
688 successful delivery of some other message. This seems entirely
689 reasonable: after all the message is NOT "waiting for some host".
690 This is so "obvious" that I'm not sure why it wasn't done
691 previously. Hope I haven't missed anything, but it can't do any
692 harm, as the worst effect is to miss an optimization.
694 (ii) If, despite (i), such a delivery is accidentally attempted, the
695 routing retry time is respected, so at least it doesn't keep
696 hammering the server.
698 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
699 in ${readsocket because some servers need this prod.
701 PH/32 Added some extra debug output when updating a wait-xxx database.
703 PH/33 The hint "could be header name not terminated by colon", which has been
704 given for certain expansion errors for a long time, was not being given
705 for the ${if def:h_colon_omitted{... case.
707 PH/34 The spec says: "With one important exception, whenever a domain list is
708 being scanned, $domain contains the subject domain." There was at least
709 one case where this was not true.
711 PH/35 The error "getsockname() failed: connection reset by peer" was being
712 written to the panic log as well as the main log, but it isn't really
713 panic-worthy as it just means the connection died rather early on. I have
714 removed the panic log writing for the ECONNRESET error when getsockname()
717 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
718 runs only) independently of the message's sender address. This meant
719 that, if the 4xx error was in fact related to the sender, a different
720 message to the same recipient with a different sender could confuse
721 things. In particualar, this can happen when sending to a greylisting
722 server, but other circumstances could also provoke similar problems.
723 I have changed the default so that the retry time for these errors is now
724 based a combination of the sender and recipient addresses. This change
725 can be overridden by setting address_retry_include_sender=false in the
728 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
729 remote server are returned as part of bounce messages. This was not
730 happening for LMTP over a pipe (the lmtp transport), but now it is the
731 same for both kinds of LMTP.
733 PH/38 Despite being documented as not happening, Exim was rewriting addresses
734 in header lines that were in fact CNAMEs. This is no longer the case.
736 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
737 and queue runs started by the daemon processed all messages. This has
738 been fixed so that -R and -S can now usefully be given with -q<time>.
740 PH/40 Import PCRE release 6.7 (fixes some bugs).
742 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
744 PH/42 Give an error if -q is specified more than once.
746 PH/43 Renamed the variables $interface_address and $interface_port as
747 $received_ip_address and $received_port, to make it clear that these
748 values apply to message reception, and not to the outgoing interface when
749 a message is delivered. (The old names remain recognized, of course.)
751 PH/44 There was no timeout on the connect() call when using a Unix domain
752 socket in the ${readsocket expansion. There now is.
754 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
755 be meaningful with "accept".
757 SC/01 Eximstats V1.43
758 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
760 SC/02 Eximstats V1.44
761 Use a glob alias rather than an array ref in the generated
762 parser. This improves both readability and performance.
764 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
765 Collect SpamAssassin and rejection statistics.
766 Don't display local sender or destination tables unless
767 there is data to show.
768 Added average volumes into the top table text output.
770 SC/04 Eximstats V1.46
771 Collect data on the number of addresses (recipients)
772 as well as the number of messages.
774 SC/05 Eximstats V1.47
775 Added 'Message too big' to the list of mail rejection
776 reasons (thanks to Marco Gaiarin).
778 SC/06 Eximstats V1.48
779 Mainlog lines which have GMT offsets and are too short to
780 have a flag are now skipped.
782 SC/07 Eximstats V1.49 (Alain Williams)
783 Added the -emptyok flag.
785 SC/08 Eximstats V1.50
786 Fixes for obtaining the IP address from reject messages.
788 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
789 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
790 whitesspace changes from 4.64-PH/27
792 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
795 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
796 are found, allow negative numbers in numeric criteria)
798 JJ/06 exipick.20061117.2, added new $message_body_missing variable
800 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
801 to match changes made in 4.64-PH/43
803 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
805 PH/47 Put in an explicit test for a DNS lookup of an address record where the
806 "domain" is actually an IP address, and force a failure. This locks out
807 those revolvers/nameservers that support "A-for-A" lookups, in
808 contravention of the specifications.
810 PH/48 When a host name was looked up from an IP address, and the subsequent
811 forward lookup of the name timed out, the host name was left in
812 $sender_host_name, contrary to the specification.
814 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
815 restricted to single-key lookups, Exim was not diagnosing an error if
816 * or *@ was used with a query-style lookup.
818 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
820 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
821 long ago, but noone interested enough thought of it. Let's just say that
822 the "1.1" means that there are some new functions that weren't there at
823 some point in the past.
825 PH/51 Error processing for expansion failure of helo_data from an smtp
826 transport during callout processing was broken.
828 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
829 tested/used via the -bh/-bhc/-bs options.
831 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
832 bug, fixed in subsequent PCRE releases).
834 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
835 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
837 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
843 SC/01 Use a glob alias rather than an array ref in eximstats generated
844 parser. This improves both readability and performance.
846 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
847 Don't display local sender or destination tables in eximstats unless
848 there is data to show.
849 Added average volumes into the eximstats top table text output.
851 SC/03 Collect data on the number of addresses (recipients) as well
852 as the number of messages in eximstats.
854 TF/01 Correct an error in the documentation for the redirect router. Exim
855 does (usually) call initgroups() when daemonizing.
857 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
858 with consistent privilege compared to when running as a daemon.
860 TF/03 Note in the spec that $authenticated_id is not set for local
861 submissions from trusted users.
863 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
864 Thanks to Dean Brooks <dean@iglou.com> for the patch.
866 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
867 by adding some example configuration directives to the default
868 configuration file. A little bit of work is required to uncomment the
869 directives and define how usernames and passwords are checked, but
870 there is now a framework to start from.
872 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
873 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
874 without this. I don't know how relevant this is to other LDAP libraries.
876 PH/02 Add the verb name to the "unknown ACL verb" error.
878 PH/03 Magnus Holmgren's patch for filter_prepend_home.
880 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
882 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
883 directory not expanded when it should be if an expanded home directory
884 was set for the address (which is overridden by the transport).
886 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
889 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
890 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
891 because it is too late at that time, and has no effect.
893 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
894 security issue with \' (bugzilla #107). I could not use the
895 PQescapeStringConn() function, because it needs a PGconn value as one of
898 PH/08 When testing addresses using -bt, indicate those final addresses that
899 are duplicates that would not cause an additional delivery. At least one
900 person was confused, thinking that -bt output corresponded to deliveries.
901 (Suppressing duplicates isn't a good idea as you lose the information
902 about possibly different redirections that led to the duplicates.)
904 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
905 systems where poll() doesn't work, in particular OS X.
907 PH/10 Added more information to debugging output for retry time not reached.
909 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
910 operations in malware.c.
912 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
915 PH/13 If write_rejectlog was set false when logging was sent to syslog with
916 syslog_duplication set false, log lines that would normally be written
917 both the the main log and to the reject log were not written to syslog at
920 PH/14 In the default configuration, change the use of "message" in ACL warn
921 statements to "add_header".
923 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
924 not followed by a command (e.g. "seen endif").
926 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
927 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
930 PH/17 Added extra conditions to the default value of delay_warning_condition
934 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
935 { match{$h_precedence:}{(?i)bulk|list|junk} } \
936 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
939 The Auto-Submitted: and various List- headers are standardised, whereas I
940 don't think Precedence: ever was.
942 PH/18 Refactored debugging code in route_finduser() to show more information,
943 in particular, the error code if getpwnam() issues one.
945 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
946 This is apparently needed in addition to the PH/07 change above to avoid
947 any possible encoding problems.
949 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
950 but not after initializing Perl.
952 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
953 output them only if debugging. By default they are written stderr,
954 apparently, which is not desirable.
956 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
959 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
962 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
964 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
965 authenticated or an ident call has been made. Suppress the default
966 values for $authenticated_id and $authenticated_sender (but permit -oMai
967 and -oMas) when testing with -bh.
969 PH/24 Re-jigged the order of the tests in the default configuration so that the
970 tests for valid domains and recipients precede the DNS black list and CSA
971 tests, on the grounds that those ones are more expensive.
973 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
974 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
975 This bug exists in every version of Exim that I still have, right back to
978 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
979 However, an attempt to turn on case-sensitivity in a regex key by
980 including (?-i) didn't work because the subject string was already
981 lowercased, and the effects were non-intuitive. It turns out that a
982 one-line patch can be used to allow (?-i) to work as expected.
988 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
989 other effects) broke the use of negated acl sub-conditions.
991 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
994 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
995 "Deny" causes Exim to reject the incoming connection with a 554 error.
996 Unfortunately, if there is a major crisis, such as a disk failure,
997 tcp-wrappers gives "deny", whereas what one would like would be some
998 kind of temporary error. A kludge has been added to help with this.
999 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1000 554 error is used if errno is still zero or contains ENOENT (which occurs
1001 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1004 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1006 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1007 errors. Otherwise a message that provokes a temporary error (when other
1008 messages do not) can cause a whole host to time out.
1010 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1011 addresses were routed directly to files or pipes from a redirect router.
1012 File deliveries just didn't batch; pipe deliveries might have suffered
1015 PH/06 A failure to get a lock for a hints database would erroneously always say
1016 "Failed to get write lock", even when it was really a read lock.
1018 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1019 of 0600. This has been changed to use the value of the lockfile_mode
1020 option (which defaults to 0600).
1022 PH/08 Applied small patch from the Sieve maintainer.
1024 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1025 folder from quota calculations, a direct delivery into this folder messed
1026 up the contents of the maildirsize file. This was because the regex was
1027 used only to exclude .Trash (or whatever) when the size of the mailbox
1028 was calculated. There was no check that a delivery was happening into an
1029 excluded directory. This bug has been fixed by ignoring all quota
1030 processing for deliveries into excluded directories.
1032 PH/10 Added the maildirfolder_create_regex option to appendfile.
1038 PH/01 The code for finding all the local interface addresses on a FreeBSD
1039 system running IPv6 was broken. This may well have applied to all BSD
1040 systems, as well as to others that have similar system calls. The broken
1041 code found IPv4 interfaces correctly, but gave incorrect values for the
1042 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1043 that it would not match correctly against @[] and not recognize the IPv6
1046 PH/02 The ipliteral router was not recognizing addresses of the form user@
1047 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1049 PH/03 Added disable_ipv6.
1051 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1052 decoded form, because it is most often used to construct To: headers
1053 lines in autoreplies, and the decoded form may well be syntactically
1054 invalid. However, $reply_address has leading white space removed, and all
1055 newlines turned into spaces so that the autoreply transport does not
1058 PH/05 If group was specified without a user on a router, and no group or user
1059 was specified on a transport, the group from the router was ignored.
1061 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1062 for visible compile-time settings that can be used to change these
1063 numbers, for those that want even more. Backwards compatibility with old
1064 spool files has been maintained. However, going back to a previous Exim
1065 release will lost any variables that are in spool files.
1067 PH/07 Two small changes when running in the test harness: increase delay when
1068 passing a TCP/IP connection to a new process, in case the original
1069 process has to generate a bounce, and remove special handling of
1070 127.0.0.2 (sic), which is no longer necessary.
1072 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1073 be the same on different OS.
1075 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1078 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1079 whether --show-vars was specified or not
1081 JJ/02 exipick: Added support for new ACL variable spool format introduced
1084 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1085 syntactically invalid From: or Reply-to: line, and a filter used this to
1086 generate an autoreply, and therefore failed to obtain an address for the
1087 autoreply, Exim could try to deliver to a non-existent relative file
1088 name, causing unrelated and misleading errors. What now happens is that
1089 it logs this as a hard delivery error, but does not attempt to create a
1092 PH/11 The exinext utility has a -C option for testing purposes, but although
1093 the given file was scanned by exinext itself; it wasn't being passed on
1094 when Exim was called.
1096 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1097 an end-of-file indication when reading a command response.
1099 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1100 compiled. In many other places in Exim, IPv6 addresses are always
1101 recognized, so I have changed this. It also means that IPv4 domain
1102 literals of the form [IPV4:n.n.n.n] are now always recognized.
1104 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1105 used if the router is not running as root, for example, when verifying at
1106 ACL time, or when using -bh. The debugging output from this situation was
1107 non-existent - all you got was a failure to exec. I have made two
1110 (a) Failures to set uid/gid, the current directory, or a process leader
1111 in a subprocess such as that created by queryprogram now generate
1112 suitable debugging ouput when -d is set.
1114 (b) The queryprogram router detects when it is not running as root,
1115 outputs suitable debugging information if -d is set, and then runs
1116 the subprocess without attempting to change uid/gid.
1118 PH/15 Minor change to Makefile for building test_host (undocumented testing
1121 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1122 additional section of a DNS packet that returns MX or SRV records.
1123 Instead, it always explicitly searches for A/AAAA records. This avoids
1124 major problems that occur when a DNS server includes only records of one
1125 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1126 fixed another bug: if SRV records were looked up and the corresponding
1127 address records were *not* found in the additional section, the port
1128 values from the SRV records were lost.
1130 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1131 using the correct key (the original address) when searching the retry
1132 rules in order to find which one to use for generating the retry hint.
1134 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1135 adding the default one. Similarly, if it contains a Reply-To: header, the
1136 errors_reply_to option, if set, is not used.
1138 PH/19 When calculating a retry time, Exim used to measure the "time since
1139 failure" by looking at the "first failed" field in the retry record. Now
1140 it does not use this if it is later than than the arrival time of the
1141 message. Instead it uses the arrival time. This makes for better
1142 behaviour in cases where some deliveries succeed, thus re-setting the
1143 "first failed" field. An example is a quota failure for a huge message
1144 when small messages continue to be delivered. Without this change, the
1145 "time since failure" will always be short, possible causing more frequent
1146 delivery attempts for the huge message than are intended.
1147 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1149 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1150 $1, $2, $3) because the numerical variables can be reset during some
1151 expansion items (e.g. "match"), thereby losing the authentication data.
1153 PH/21 Make -bV show the size of off_t variables so that the test suite can
1154 decide whether to run tests for quotas > 2G.
1156 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1157 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1158 filecount value is greater than 2G or if a quota value is greater than 2G
1159 on a system where the size of off_t is not greater than 4, a panic error
1162 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1163 never match. The debug and -bh output now contains an explicit error
1164 message indicating a malformed IPv4 address or mask.
1166 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1167 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1168 PH/23 above applies.
1170 PH/25 Do not write to syslog when running in the test harness. The only
1171 occasion when this arises is a failure to open the main or panic logs
1172 (for which there is an explicit test).
1174 PH/26 Added the /no_tell option to "control=freeze".
1176 PH/27 If a host name lookup failed very early in a connection, for example, if
1177 the IP address matched host_lookup and the reverse lookup yielded a name
1178 that did not have a forward lookup, an error message of the form "no IP
1179 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1180 could be logged. Now it outputs the IP address instead of "NULL".
1182 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1183 allows the sender and the authenticated sender to be set when
1184 submitting a message from within Exim. Since child_open_exim() is
1185 documented for local_scan(), the new function should be too.
1187 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1188 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1189 results in an empty string is now treated as unset.
1191 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1193 PH/31 Added sender_verify_fail logging option.
1195 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1196 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1197 tidied the source and removed it altogether.
1199 PH/33 When a queue run was abandoned because the load average was too high, a
1200 log line was always written; now it is written only if the queue_run log
1201 selector is set. In addition, the log line for abandonment now contains
1202 information about the queue run such as the pid. This is always present
1203 in "start" and "stop" lines but was omitted from the "abandon" line.
1205 PH/34 Omit spaces between a header name and the colon in the error message that
1206 is given when verify = headers_syntax fails (if there are lots of them,
1207 the message gets confusing).
1209 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1210 names, as there are now some PTR records that contain slashes. This check
1211 is only to protect against broken name servers that fall over on strange
1212 characters, so the fact that it applies to all lookups doesn't matter.
1214 PH/36 Now that the new test suite is complete, we can remove some of the
1215 special code in Exim that was needed for the old test suite. For example,
1216 sorting DNS records because real resolvers return them in an arbitrary
1217 order. The new test suite's fake resolver always returns records in the
1220 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1221 bounces) except when queue_only is set, to avoid logging races between
1222 the different processes.
1224 PH/38 Panic-die if .include specifies a non-absolute path.
1226 PH/39 A tweak to the "H" retry rule from its user.
1228 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1229 a label. They prevented compilation on older perls.
1231 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1232 a warning to be raised on newish perls.
1234 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1235 on queue. Changes to match documented behaviour of showing count of
1236 messages matching specified criteria.
1238 PH/40 Changed the default ident timeout from 30s to 5s.
1240 PH/41 Added support for the use of login_cap features, on those BSD systems
1241 that have them, for controlling the resources used by pipe deliveries.
1243 PH/42 The content-scanning code uses fopen() to create files in which to put
1244 message data. Previously it was not paying any attention to the mode of
1245 the files. Exim runs with umask(0) because the rest of the code creates
1246 files with open(), and sets the required mode explicitly. Thus, these
1247 files were ending up world-writeable. This was not a big issue, because,
1248 being within the spool directory, they were not world-accessible. I have
1249 created a function called modefopen, which takes an additional mode
1250 argument. It sets umask(777), creates the file, chmods it to the required
1251 mode, then resets the umask. All the relevant calls to fopen() in the
1252 content scanning code have been changed to use this function.
1254 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1255 to 24 hours. This avoids potential overflow problems when processing G
1256 and H retry rules. I suspect nobody ever tinkers with this value.
1258 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1260 PH/45 When the plaintext authenticator is running as a client, the server's
1261 challenges are checked to ensure they are valid base64 strings. By
1262 default, the authentication attempt is cancelled if an invalid string is
1263 received. Setting client_ignore_invalid_base64 true ignores these errors.
1264 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1265 they are received. Thus, the responses can be made to depend on the
1266 challenges. If an invalid string is ignored, an empty string is placed in
1269 PH/46 Messages that are created by the autoreply transport now contains a
1270 References: header, in accordance with RFCs 2822 and 3834.
1272 PH/47 Added authenticated_sender_force to the smtp transport.
1274 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1276 PH/49 Installed latest patch from the Sieve maintainer.
1278 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1279 was also set, the appendfile transport was unnecessarily scanning a
1280 directory of message files (e.g. for maildir delivery) to find the count
1281 of files (along with the size), even though it did not need this
1282 information. It now does the scan only if it needs to find either the
1283 size of the count of files.
1285 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1287 PH/52 Two bugs concerned with error handling when the smtp transport is
1290 (i) Exim was not creating retry information for temporary errors given
1291 for individual recipients after the DATA command when the smtp transport
1292 was used in LMTP mode. This meant that they could be retried too
1293 frequently, and not timed out correctly.
1295 (ii) Exim was setting the flag that allows error details to be returned
1296 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1297 recipients that were returned after the DATA command.
1299 PH/53 This is related to PH/52, but is more general: for any failing address,
1300 when detailed error information was permitted to be returned to the
1301 sender, but the error was temporary, then after the final timeout, only
1302 "retry timeout exceeded" was returned. Now it returns the full error as
1303 well as "retry timeout exceeded".
1305 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1306 do this, and (what is worse) MTAs that accept it.
1308 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1309 will now be deprecated.
1311 PH/56 New os.c-cygwin from the Cygwin maintainer.
1313 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1314 formats (previously only available in exim formats via -bpr, -bpru,
1315 and -bpra. Now also available in native and exiqgrep formats)
1317 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1318 with very large, slow to parse queues
1320 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1322 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1324 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1325 responses to authentication challenges, though it was showing the
1326 challenges; (ii) I've removed the CR characters from the debug output for
1329 PH/58 Allow for the insertion of a newline as well as a space when a string
1330 is turned into more than one encoded-word during RFC 2047 encoding. The
1331 Sieve code now uses this.
1333 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1334 data_4xx, lost_connection, tls_required.
1336 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1337 message was being sent as an SMTP response.
1339 PH/61 Add -l and -k options to exicyclog.
1341 PH/62 When verifying, if an address was redirected to one new address, so that
1342 verification continued, and the new address failed or deferred after
1343 having set something in $address_data, the value of $address_data was not
1344 passed back to the ACL. This was different to the case when no
1345 redirection occurred. The value is now passed back in both cases.
1347 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1348 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1349 use login_cap.h, so on its own it isn't the distinguishing feature. The
1350 new name refers directly to the setclassresources() function.
1352 PH/65 Added configuration files for NetBSD3.
1354 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1356 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1357 is preferred over IPv4.
1359 PH/68 The bounce_return_message and bounce_return_body options were not being
1360 honoured for bounces generated during the reception of non-SMTP messages.
1361 In particular, this applied to messages rejected by the ACL. This bug has
1362 been fixed. However, if bounce_return_message is true and bounce_return_
1363 body is false, the headers that are returned for a non-SMTP message
1364 include only those that have been read before the error was detected.
1365 (In the case of an ACL rejection, they have all been read.)
1367 PH/69 The HTML version of the specification is now built in a directory called
1368 spec_html instead of spec.html, because the latter looks like a path with
1369 a MIME-type, and this confuses some software.
1371 PH/70 Catch two compiler warnings in sieve.c.
1373 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1374 function verify_get_ident() calls ip_connect() to connect a socket, but
1375 if the "connect()" function timed out, ip_connect() used to close the
1376 socket. However, verify_get_ident() also closes the socket later, and in
1377 between Exim writes to the log, which may get opened at this point. When
1378 the socket was closed in ip_connect(), the log could get the same file
1379 descriptor number as the socket. This naturally causes chaos. The fix is
1380 not to close the socket in ip_connect(); the socket should be closed by
1381 the function that creates it. There was only one place in the code where
1382 this was missing, in the iplookup router, which I don't think anybody now
1383 uses, but I've fixed it anyway.
1385 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1386 well as to direct DNS lookups. Otherwise the handling of names in host
1387 lists is inconsistent and therefore confusing.
1393 PH/01 Two changes to the default runtime configuration:
1395 (1) Move the checks for relay_from_hosts and authenticated clients from
1396 after to before the (commented out) DNS black list checks.
1398 (2) Add control=submission to the relay_from_hosts and authenticated
1399 clients checks, on the grounds that messages accepted by these
1400 statements are most likely to be submissions.
1402 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1404 (1) Generate an error if the third argument for the ${prvs expansion is
1407 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1410 (3) Reset the variables that are obtained from the first argument of
1411 ${prvscheck and used in the second argument before leaving the code,
1412 because their memory is reclaimed, so using them afterwards may do
1415 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1416 one (it's much easier than Tom thought :-).
1418 (5) Because of (4), we can now allow for the use of $prvscheck_result
1419 inside the third argument.
1421 PH/03 For some reason, the default setting of PATH when running a command from
1422 a pipe transport was just "/usr/bin". I have changed it to
1425 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1426 anything to be listed in the output from -bV.
1428 PH/05 When a filter generated an autoreply, the entire To: header line was
1429 quoted in the delivery log line, like this:
1431 => >A.N.Other <ano@some.domain> <original@ddress> ...
1433 This has been changed so that it extracts the operative address. There
1434 may be more than one such address. If so, they are comma-separated, like
1437 => >ano@some.domain,ona@other.domain <original@ddress> ...
1439 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1440 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1441 not being looked up in the rDNS to get a host name, Exim was showing the
1442 IP address twice in Received: lines, even though the IP addresses were
1443 identical. For example:
1445 Received: from [1.2.3.4] (helo=[1.2.3.4])
1447 However, if the real host name was known, it was omitting the HELO data
1448 if it matched the actual IP address. This has been tidied up so that it
1449 doesn't show the same IP address twice.
1451 PH/07 When both +timestamp and +memory debugging was on, the value given by
1452 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1453 called by the debug printing, thereby overwriting the timestamp buffer.
1454 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1456 PH/08 When the original message was included in an autoreply transport, it
1457 always said "this is a copy of the message, including all the headers",
1458 even if body_only or headers_only was set. It now gives an appropriate
1461 PH/09 Applied a patch from the Sieve maintainer which:
1463 o fixes some comments
1464 o adds the (disabled) notify extension core
1465 o adds some debug output for the result of if/elsif tests
1466 o points to the current vacation draft in the documentation
1467 and documents the missing references header update
1471 o fixes a bug in processing the envelope test (when testing
1472 multiple envelope elements, the last element determinted the
1475 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1476 Electronic Mail") by including:
1478 Auto-submitted: auto-generated
1480 in the messages that it generates (bounce messages and others, such as
1481 warnings). In the case of bounce messages for non-SMTP mesages, there was
1482 also a typo: it was using "Auto_submitted" (underscore instead of
1483 hyphen). Since every message generated by Exim is necessarily in response
1484 to another message, thes have all been changed to:
1486 Auto-Submitted: auto-replied
1488 in accordance with these statements in the RFC:
1490 The auto-replied keyword:
1492 - SHOULD be used on messages sent in direct response to another
1493 message by an automatic process,
1495 - MUST NOT be used on manually-generated messages,
1497 - MAY be used on Delivery Status Notifications (DSNs) and Message
1498 Disposition Notifications (MDNs),
1500 - MUST NOT be used on messages generated by automatic or periodic
1501 processes, except for messages which are automatic responses to
1504 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1505 to the default Received: header definition.
1507 PH/12 Added log selector acl_warn_skipped (default on).
1509 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1510 variables because (a) they are in the wrong storage pool and (b) even if
1511 they were copied, it wouldn't work properly because of the caching.
1513 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1514 checking when decoding. Apparently there are clients that generate
1515 overlong encoded strings. Why am I not surprised?
1517 PH/15 If the first argument of "${if match_address" was not empty, but did not
1518 contain an "@" character, Exim crashed. Now it writes a panic log message
1519 and treats the condition as false.
1521 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1523 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1524 extension "envelope-auth". The code is finished and in agreement with
1525 other implementations, but there is no documentation so far and in fact,
1526 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1527 not changing the active code.
1529 Print executed "if" and "elsif" statements when debugging is used. This
1530 helps a great deal to understand what a filter does.
1532 Document more things not specified clearly in RFC3028. I had all this
1533 sorted out, when out of a sudden new issues came to my mind. Oops."
1535 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1538 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1539 canonical form (as documented). However, after a host name lookup from
1540 the IP address, check_host() was doing a simple string comparison with
1541 addresses acquired from the DNS when checking that the found name did
1542 have the original IP as one of its addresses. Since any found IPv6
1543 addresses are likely to be in abbreviated form, the comparison could
1544 fail. Luckily, there already exists a function for doing the comparison
1545 by converting both addresses to binary, so now that is used instead of
1546 the text comparison.
1548 PH/20 There was another similar case to PH/19, when a complete host name was
1549 given in a host list; looking up its IP address could give an abbreviated
1550 form, whereas the current host's name might or might not be abbreviated.
1551 The same fix has been applied.
1557 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1558 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1561 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1562 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1564 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1566 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1567 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1568 Also, for "Auto-Submitted", treat anything other than "no" as
1569 non-personal, in accordance with RFC 3834. (Previously it treated
1570 anything starting "auto-" as non-personal.)
1572 TF/01 The control=submission/name=... option had a problem with syntax
1573 errors if the name included a slash character. The /name= option
1574 now slurps the rest of the string, so it can include any characters
1575 but it must come last in the list of options (after /sender_retain
1578 PH/05 Some modifications to the interface to the fake nameserver for the new
1586 TK/01 Added the "success_on_redirect" address verification option. See
1587 NewStuff for rationale and an example.
1589 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1591 PH/02 Patch to exigrep to allow it to work on syslog lines.
1593 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1594 fread() to skip over the body file's header line, because in Cygwin the
1595 header line is locked and is inaccessible.
1597 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1598 co-exist for some time) to make it clear that it is the Exim ID that is
1599 referenced, not the Message-ID: header line.
1601 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1602 string_format() function, because snprintf() does not exist on all
1605 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1606 ${stat: expansion item.
1608 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1609 protocol synchronization error", to keep the pedants happy.
1611 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1612 well as for IRIX systems, when gcc is being used. See the host.c source
1615 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1617 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1620 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1621 transport and to the smtp transport in LMTP mode.
1623 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1625 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1626 run a filter in a subprocess. This could lead to confusion in subsequent
1627 lookups in the parent process. There should also be a search_tidyup() at
1628 the end of the subprocess.
1630 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1631 only if the host matched helo_try_verify_hosts, which caused the
1632 verification to occur when the EHLO/HELO command was issued. The ACL just
1633 tested the remembered result. Now, if a previous verification attempt has
1634 not happened, "verify = helo" does it there and then.
1636 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1638 TK/03 Fix log output including CR from clamd.
1640 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1641 exist provoked a memory error which could cause a segfault.
1643 PH/15 Installed PCRE 6.2
1645 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1647 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1648 of the problem. Specifically, suggested +O2 rather than +O1 for the
1651 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1653 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1654 record, the port was not correctly logged when the outgoing_port log
1655 selector was set (it logged the transort's default port).
1657 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1658 fallback_hosts, and "hosts" in the smtp transport.
1660 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1661 host errors such as "Connection refused".
1663 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1664 authentication with radiusclient 0.4.9:
1666 - Error returned from rc_read_config was caught wrongly
1667 - Username/password not passed on to radius server due to wrong length.
1669 The presumption is that some radiusclient API changes for 4.51/PH/17
1670 were not taken care of correctly. The code is still untested by me (my
1671 Linux distribution still has 0.3.2 of radiusclient), but it was
1672 contributed by a Radius user.
1674 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1675 expanding the "port" option of the smtp transport.
1677 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1678 while reading a MIME header. Thanks to Tom Hughes for a patch.
1680 PH/24 Include config.h inside local_scan.h so that configuration settings are
1683 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1684 that in an ACL for RCPT (for example), you can examine exactly what was
1687 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1688 commands, but it was not correctly comparing the address with the actual
1689 client host address. Thus, it would show the EHLO address in Received:
1690 header lines when this was not necessary.
1692 PH/27 Added the % operator to ${eval:}.
1694 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1695 it should be ignoring failures (because with -C, for example, it has lost
1696 privilege). It wasn't ignoring creation failures other than "already
1699 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1702 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1703 because an input line was too long, either on its own, or by virtue of
1704 too many continuations, the temporary file was not being removed, and the
1705 return code was incorrect.
1707 PH/31 Missing "BOOL" in function definition in filtertest.c.
1709 PH/32 Applied Sieve patches from the maintainer.
1711 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1713 PH/33 Added "verify = not_blind".
1715 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1716 Local/Makefile (with some defaults set). These are used in built scripts
1717 such as exicyclog, but they have never been used in the exim_install
1718 script (though there are many overriding facilities there). I have
1719 arranged that the exim_install script now takes note of these two
1722 PH/35 Installed configuration files for Dragonfly.
1724 PH/36 When a locally submitted message by a trusted user did not contain a
1725 From: header, and the sender address was obtained from -f or from an SMTP
1726 MAIL command, and the trusted user did not use -F to supply a sender
1727 name, $originator_name was incorrectly used when constructing a From:
1728 header. Furthermore, $originator_name was used for submission mode
1729 messages from external hosts without From: headers in a similar way,
1730 which is clearly wrong.
1732 PH/37 Added control=suppress_local_fixups.
1734 PH/38 When log_selector = +received_sender was set, and the addition of the
1735 sender made the log line's construction buffer exactly full, or one byte
1736 less than full, an overflow happened when the terminating "\n" was
1739 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1740 when the result of a list match is failure because a DNS lookup failed.
1742 PH/40 RM_COMMAND is now used in the building process.
1744 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1745 the "build-* directories that it finds.
1747 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1748 address in a domain literal was a prefix of an interface address.
1750 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1751 when verifying a sender address, unless rewrite_headers is false.
1753 PH/44 Wrote a long comment about why errors_to addresses are verified as
1754 recipients, not senders.
1756 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1757 the ratelimit ACL was added.
1759 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1761 PH/46 Added extra information about PostgreSQL errors to the error string.
1763 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1764 avoiding the need to install special zones in a real server. This is
1765 backwards compatible; if it can't find the fake resolver, it drops back.
1766 Thus, both old and new test suites can be run.
1768 TF/02 Added util/ratelimit.pl
1770 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1771 clock is set back in time.
1773 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1774 Candler <B.Candler@pobox.com>.
1776 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1777 OK for addresses that are the result of redirections.
1779 PH/48 A number of further additions for the benefit of the new test suite,
1780 including a fake gethostbyname() that interfaces to the fake DNS resolver
1783 TF/06 The fix for widen_domains has also been applied to qualify_single and
1784 search_parents which are the other dnslookup options that can cause
1787 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1790 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1792 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1793 service" if it encountered an MX record with an empty target hostname.
1794 The message is now "an MX or SRV record indicated no SMTP service".
1796 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1797 if the DNS of the sending site is misconfigured. This is quite a
1798 common situation. This change restores the behaviour of treating a
1799 helo verification defer as a failure.
1801 PH/51 If self=fail was set on a router, the bounce message did not include the
1802 actual error message.
1808 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1810 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1811 command itself ended in error, the underlying message about the transport
1812 filter timeout was being overwritten with the pipe command error. Now the
1813 underlying error message should be appended to the second error message.
1815 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1817 PH/02 Reduce the amount of output that "make" produces by default. Full output
1818 can still be requested.
1820 PH/03 The warning log line about a condition test deferring for a "warn" verb
1821 was being output only once per connection, rather than after each
1822 occurrence (because it was using the same function as for successful
1823 "warn" verbs). This seems wrong, so I have changed it.
1825 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1826 it should not have, which might have caused a crash in the right
1827 circumstances, but probably never did.
1829 PH/04 Installed a modified version of Tony Finch's patch to make submission
1830 mode fix the return path as well as the Sender: header line, and to
1831 add a /name= option so that you can make the user's friendly name appear
1834 TF/03 Added the control = fakedefer ACL modifier.
1836 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1837 Mark Lowes for thorough testing.
1839 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1841 TK/03 Merged latest SRS patch from Miles Wilton.
1843 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1844 with the definition in sysexits.h (which is #included earlier).
1845 Fortunately, Exim does not actually use EX_OK. The code used to try to
1846 preserve the sysexits.h value, by assumimg that macro definitions were
1847 scanned for macro replacements. I have been disabused of this notion,
1848 so now the code just undefines EX_OK before #including unistd.h.
1850 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1851 in the smtp transport. When a block could not be written in a single
1852 write() function, the timeout was being re-applied to each part-write.
1853 This seems wrong - if the receiver was accepting one byte at a time it
1854 would take for ever. The timeout is now adjusted when this happens. It
1855 doesn't have to be particularly precise.
1857 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1858 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1860 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1861 without a domain if the check to <postmaster@domain> fails.
1863 SC/01 Eximstats: added -xls and the ability to specify output files
1864 (patch written by Frank Heydlauf).
1866 SC/02 Eximstats: use FileHandles for outputing results.
1868 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1870 SC/04 Eximstats: fixed display of large numbers with -nvr option
1872 SC/05 Eximstats: fixed merging of reports with empty tables.
1874 SC/06 Eximstats: added the -include_original_destination flag
1876 SC/07 Eximstats: removed tabs and trailing whitespace.
1878 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1880 TK/06 MBOX spool code: Add real "From " MBOX separator line
1881 so the .eml file is really in mbox format (even though
1882 most programs do not really care). Patch from Alex Miller.
1884 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1885 The latter is generated from $received_to and is only set if the
1886 message has one envelope recipient. SA can use these headers,
1887 obviously out-of-the-box. Patch from Alex Miller.
1889 PH/08 The ${def test on a variable was returning false if the variable's
1890 value was "0", contrary to what the specification has always said!
1891 The result should be true unless the variable is empty.
1893 PH/09 The syntax error of a character other than { following "${if
1894 def:variable_name" (after optional whitespace) was not being diagnosed.
1895 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1896 accidental colon was present, for example, could give incorrect results.
1898 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1899 result is used (not including appendfile, where other changes are about
1902 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1903 This involved changing a lot of size variables from int to off_t. It
1904 should work with maildirs and everything.
1906 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1907 spamd dying while we are connected to it.
1909 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1912 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1913 function for BDB 4.3.
1915 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1917 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1918 the sources for PCRE have been re-arranged and more files are now
1921 PH/15 The code I had for printing potentially long long variables in PH/11
1922 above was not the best (it lost precision). The length of off_t variables
1923 is now inspected at build time, and an appropriate printing format (%ld
1924 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1925 to be "long long int" or "long int". This is needed for the internal
1926 formatting function string_vformat().
1928 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1929 the configuration file to be ":syslog", then the script "guesses" where
1930 the logs files are, rather than using the compiled in default. In our
1931 case the guess is not the same as the compiled default, so the script
1932 suddenly stopped working when I started to use syslog. The patch checks
1933 to see if log_file_path is "". If so, it attempts to read it from exim
1934 with no configuration file to get the compiled in version, before it
1935 falls back to the previous guessing code."
1937 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1938 implementing BATV in an Exim configuration. See NewStuff for the gory
1941 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1942 Makefile that are specific to HP-UX.
1944 PH/18 If the "use_postmaster" option was set for a recipient callout together
1945 with the "random" option, the postmaster address was used as the MAIL
1946 FROM address for the random test, but not for the subsequent recipient
1947 test. It is now used for both.
1949 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1950 patch removes a few documentation additions to RFC 3028, because the
1951 latest draft now contains them. It adds the new en;ascii-case comparator
1952 and a new error check for 8bit text in MIME parts. Comparator and
1953 require names are now matched exactly. I enabled the subaddress
1954 extension, but it is not well tested yet (read: it works for me)."
1956 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1957 rework some of the code of TK/09 above to avoid the hardwired use of
1958 "%lld" and "long long". Replaced the call to snprintf() with a call to
1961 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1962 records point to non-existent hosts", "retry timeout exceeded", and
1963 "retry time not reached for any host after a long failure period".
1965 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1966 experimental DomainKeys support:
1968 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1969 (2) On an error such as an illegally used "control", the wrong name for
1970 the control was given.
1972 These problems did NOT occur unless DomainKeys support was compiled.
1974 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1976 PH/24 Added ${if match_ip condition.
1978 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1979 databases so that it will be absolutely obvious if a crash occurs in the
1980 DB library. This is a regular occurrence (often caused by mis-matched
1983 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1984 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1985 picked up on a user's system that detects such things. There doesn't seem
1986 to be a gcc warning option for this - only an attribute that has to be
1987 put on the function's prototype. It seems that in Fedora Core 4 they have
1988 set this on a number of new functions. No doubt there will be more in due
1991 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1992 specify a transport. However, if an address that was verified by such a
1993 router was the subject of a callout, Exim crashed because it tried to
1994 read the rcpt_include_affixes from the non-existent transport. Now it
1995 just assumes that the setting of that option is false. This bug was
1996 introduced by 4.51/PH/31.
1998 PH/28 Changed -d+all to exclude +memory, because that information is very
1999 rarely of interest, but it makes the output a lot bigger. People tend to
2000 do -d+all out of habit.
2002 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2003 code in os-type was giving problems when libc.so lives in lib64, like on
2006 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2007 aren't the modern standard, and it seems that some systems' include files
2008 don't always have them. Exim was already checking for some of the newer
2009 ones like T_AAAA, and defining it itself. I've added checks for all the
2010 record types that Exim uses.
2012 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2013 not automatically generating a new one, as it is supposed to. This
2014 prevented TLS from working. If the file did exist, but contained invalid
2015 data, a new version was generated, as expected. It was only the case of a
2016 non-existent file that was broken.
2018 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2019 with a change in libdomainkeys > 0.64.
2021 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2022 from DNS. If the selector record carries the flag, it now has
2023 precedence over the domain-wide flag.
2025 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2027 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2028 the use of a port name that isn't defined in /etc/services) occurred, the
2029 message was deferred as in a normal delivery, and thus remained on the
2030 spool, instead of being failed because of the mua_wrapper setting. This
2031 is now fixed, and I tidied up some of the mua_wrapper messages at the
2034 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2035 the messages in a hash of arrays rather than using individual hashes.
2036 This is a bit cleaner and results in dramatic memory savings, albeit
2037 at a slight CPU cost.
2039 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2040 as requested by Marc Sherman.
2042 SC/10 Eximstats: added histograms for user specified patterns as requested
2045 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2047 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2048 fopen() in the content-scanning modules that did not already have it.
2054 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2055 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2057 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2059 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2061 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2062 to test Sieve filters that use "vacation".
2064 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2065 that changes the way the GnuTLS parameters are stored in the cache file.
2066 The new format can be generated externally. For backward compatibility,
2067 if the data in the cache doesn't make sense, Exim assumes it has read an
2068 old-format file, and it generates new data and writes a new file. This
2069 means that you can't go back to an older release without removing the
2072 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2073 work if there are any delivery delays because "one_time" forces the
2074 parent to be marked "delivered", so its unseen clone is never tried
2075 again. For this reason, Exim now forbids the simultaneous setting of
2078 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2079 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2080 entry if you want to know the details. The fix, however, neglected to
2081 consider the case when local delivery batching is involved. The test for
2082 "previously delivered" was not happening when checking to see if an
2083 address could be batched with a previous (undelivered) one; under
2084 certain circumstances this could lead to multiple deliveries to the same
2087 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2088 in its include files, and this causes problems building Exim.
2090 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2091 header_syntax) but Exim was just ignoring anything given after a slash.
2092 In particular, this caused confusion with an attempt to use "verify =
2093 reverse_host_lookup/defer_ok". An error is now given when options are
2094 supplied for verify items that do not have them. (Maybe reverse_host_
2095 lookup should have a defer_ok option, but that's a different point.)
2097 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2098 defined by RFC 821) to 2048, because there were problems with some AUTH
2099 commands, and RFC 1869 says the size should be increased for extended
2100 SMTP commands that take arguments.
2102 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2105 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2106 "unknown" error; now it says that the functionality isn't in the binary.
2108 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2109 an address' error message when a string expansion fails (syntax or
2110 whatever). Otherwise the password may appear in the log. Following change
2111 PH/42 below, there is no longer a chance of it appearing in a bounce
2114 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2116 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2117 of its addresses was ever tried. (Bugzilla bug #2.)
2119 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2120 the result incorrectly in the debug output. (It correctly added a newline
2121 to what was transported.)
2123 TF/01 Added $received_time.
2125 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2126 commented out examples of how to interface to a virus scanner and to
2127 SpamAssassin. Also added commented examples of av_scanner and
2128 spamd_address settings.
2130 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2131 and controls are allowed in which ACLs. There were a couple of minor
2132 errors. Some of the entries in the conditions table (which is a table of
2133 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2134 negation of where the condition IS allowed.
2136 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2138 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2139 header file does not have a version number, so I've had to invent a new
2140 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2141 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2142 radiusclient), but it was contributed by a Radius user.
2144 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2145 files or directories, for OpenSSL.
2147 PH/19 When an Exim process that is running as root has to create an Exim log
2148 file, it does so in a subprocess that runs as exim:exim so as to get the
2149 ownership right at creation (otherwise, other Exim processes might see
2150 the file with the wrong ownership). There was no test for failure of this
2151 fork() call, which would lead to the process getting stuck as it waited
2152 for a non-existent subprocess. Forks do occasionally fail when resources
2153 run out. I reviewed all the other calls to fork(); they all seem to check
2156 PH/20 When checking for unexpected SMTP input at connect time (before writing
2157 the banner), Exim was not dealing correctly with a non-positive return
2158 from the read() function. If the client had disconnected by this time,
2159 the result was a log entry for a synchronization error with an empty
2160 string after "input=" when read() returned zero. If read() returned -1
2161 (an event I could not check), uninitialized data bytes were printed.
2162 There were reports of junk text (parts of files, etc) appearing after
2165 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2167 PH/22 Added support for macro redefinition, and (re)definition in between
2168 driver and ACL definitions.
2170 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2171 forgetting to use the resulting value; it was using the unexpanded value.
2173 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2174 hadn't been configured. The fix is from Juergen Kreileder, who
2175 understands it better than I do:
2177 "Here's what I see happening with three configured cyrus_sasl
2178 authenticators configured (plain, login, cram-md5):
2180 On startup auth_cyrus_sasl_init() gets called for each of these.
2181 This means three calls to sasl_listmech() without a specified mech_list.
2182 => SASL tests which mechs of all available mechs actually work
2183 => three warnings about OTP not working
2184 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2186 With the patch, sasl_listmech() also gets called three times. But now
2187 SASL's mech_list option is set to the server_mech specified in the the
2188 authenticator. Or in other words, the answer from sasl_listmech()
2189 gets limited to just the mech you're testing for (which is different
2191 => the return list contains just 'plain' or 'login', 'cram-md5' or
2192 nothing depending on the value of ob->server_mech.
2194 I've just tested the patch: Authentication still works fine,
2195 unavailable mechs specified in the exim configuration are still
2196 caught, and the auth.log warnings about OTP are gone."
2198 PH/25 When debugging is enabled, the contents of the command line are added
2199 to the debugging output, even when log_selector=+arguments is not
2202 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2203 answer is "GNU", and only if the return is "GNU/something" is the answer
2206 PH/27 $acl_verify_message is now set immediately after the failure of a
2207 verification in an ACL, and so is available in subsequent modifiers. In
2208 particular, the message can be preserved by coding like this:
2210 warn !verify = sender
2211 set acl_m0 = $acl_verify_message
2213 Previously, $acl_verify_message was set only while expanding "message"
2214 and "log_message" when a very denied access.
2216 PH/28 Modified OS/os.c-Linux with
2218 -#ifndef OS_LOAD_AVERAGE
2219 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2221 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2222 nomenclature these days.)
2224 PH/29 Installed patch from the Sieve maintainer that adds the options
2225 sieve_useraddress and sieve_subaddress to the redirect router.
2227 PH/30 In these circumstances:
2228 . Two addresses routed to the same list of hosts;
2229 . First host does not offer TLS;
2230 . First host accepts first address;
2231 . First host gives temporary error to second address;
2232 . Second host offers TLS and a TLS session is established;
2233 . Second host accepts second address.
2234 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2235 and peerdn, if requested) that were in fact used only for the second
2238 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2239 attention to any local part prefix or suffix that was matched by the
2240 router that accepted the address. It now behaves in the same way as it
2241 does for delivery: the affixes are removed from the local part unless
2242 rcpt_include_affixes is set on the transport.
2244 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2245 timeout during the DATA phase of an incoming message.
2247 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2248 applied a patch sanctioned by the Sieve maintainer.
2250 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2251 the uid or gid is negative. A case of a negative gid caused this to be
2252 noticed. The fix allows for either to be negative.
2254 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2255 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2258 PH/36 The change PH/12 above was broken. Fixed it.
2260 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2261 the grounds that routing the same address twice would always produce the
2262 same answer. This might have been true once, but it is certainly no
2263 longer true now. Routing a child address may depend on the previous
2264 routing that produced that child. Some complicated redirection strategies
2265 went wrong when messages had multiple recipients, and made Exim's
2266 behaviour dependent on the order in which the addresses were given.
2268 I have moved the duplicate checking until after the routing is complete.
2269 Exim scans the addresses that are assigned to local and remote
2270 transports, and removes any duplicates. This means that more work will be
2271 done, as duplicates will always all be routed, but duplicates are
2272 presumably rare, so I don't expect this is of any significance.
2274 For deliveries to pipes, files, and autoreplies, the duplicate checking
2275 still happens during the routing process, since they are not going to be
2278 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2279 It corrects a timeout issue with spamd. This is Ian's comment: "The
2280 background is that sometimes spamd either never reads data from a
2281 connection it has accepted, or it never writes response data. The exiscan
2282 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2283 blindly assumes that writes won't block so it may never time out."
2285 PH/39 Allow G after quota size as well as K and M.
2287 PH/40 The value set for $authenticated_id in an authenticator may not contain
2288 binary zeroes or newlines because the value is written to log lines and
2289 to spool files. There was no check on this. Now the value is run through
2290 the string_printing() function so that such characters are converted to
2291 printable escape sequences.
2293 PH/41 $message_linecount is a new variable that contains the total number of
2294 lines in the message. Compare $body_linecount, which is the count for the
2297 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2298 bounce and delay warning messages, except in certain special cases, which
2301 (a) An SMTP error message from a remote host;
2302 (b) A message specified in a :fail: redirection;
2303 (c) A message specified in a "fail" command in a system filter;
2304 (d) A message specified in a FAIL return from the queryprogram router;
2305 (e) A message specified by the cannot_route_message router option.
2307 In these cases only, Exim does include the error details in bounce and
2308 warning messages. There are also a few cases where bland messages such
2309 as "unrouteable address" or "local delivery error" are given.
2311 PH/43 $value is now also set for the "else" part of a ${run expansion.
2313 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2314 being worked on, but at least Exim now implements the latest version to
2317 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2318 process to complete was treated as a delivery failure, a timeout while
2319 writing the message to the pipe was logged, but erroneously treated as a
2320 successful delivery. Such timeouts include transport filter timeouts. For
2321 consistency with the overall process timeout, these timeouts are now
2322 treated as errors, giving rise to delivery failures by default. However,
2323 there is now a new Boolean option for the pipe transport called
2324 timeout_defer, which, if set TRUE, converts the failures into defers for
2325 both kinds of timeout. A transport filter timeout is now identified in
2328 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2329 systems where "make" and "gmake" are different, calling "gmake" at top
2330 level broke things. I've arranged for the value of $(MAKE) to be passed
2331 from the Makefile to this script so that it can call the same version of
2335 A note about Exim versions 4.44 and 4.50
2336 ----------------------------------------
2338 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2339 changes of various kinds. As a consequence, a big documentation update was
2340 needed. This delayed the release for rather longer than seemed good, especially
2341 in the light of a couple of (minor) security issues. Therefore, the changes
2342 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2343 release. So 4.44 and 4.50 are in effect two different branches that both start
2346 I have left the 4.50 change log unchanged; it contains all the changes since
2347 4.43. The change log for 4.44 is below; many of its items are identical to
2348 those for 4.50. This seems to be the most sensible way to preserve the
2349 historical information.
2355 1. Minor wording change to the doc/README.SIEVE file.
2357 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2358 computation of the current number of files was incorrect.
2360 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2361 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2364 4. Give more explanation in the error message when the command for a transport
2365 filter fails to execute.
2367 5. There are several places where Exim runs a non-Exim command in a
2368 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2369 was being done only for the command run by the queryprogram router. It is
2370 now done for all such subprocesses. The other cases are: ${run, transport
2371 filters, and the commands run by the lmtp and pipe transports.
2373 6. Added CONFIGURE_GROUP build-time option.
2375 7. Some older OS have a limit of 256 on the maximum number of file
2376 descriptors. Exim was using setrlimit() to set 1000 as a large value
2377 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2378 systems. I've change it so that if it can't get 1000, it tries for 256.
2380 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2381 was an oversight, and furthermore, ever since the addition of extra
2382 controls (e.g. 4.43/32), the checks on when to allow different forms of
2383 "control" were broken. There should now be diagnostics for all cases when a
2384 control that does not make sense is encountered.
2386 9. Added the /retain_sender option to "control=submission".
2388 10. $recipients is now available in the predata ACL (oversight).
2390 11. Tidy the search cache before the fork to do a delivery from a message
2391 received from the command line. Otherwise the child will trigger a lookup
2392 failure and thereby defer the delivery if it tries to use (for example) a
2393 cached ldap connection that the parent has called unbind on.
2395 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2396 of $address_data from the recipient verification was clobbered by the
2397 sender verification.
2399 13. The value of address_data from a sender verification is now available in
2400 $sender_address_data in subsequent conditions in the ACL statement.
2402 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2404 15. Added a new option "connect=<time>" to callout options, to set a different
2407 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2408 was its contents. (It was OK if the option was not defined at all.)
2410 17. A "Completed" log line is now written for messages that are removed from
2411 the spool by the -Mrm option.
2413 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2414 information about exactly what failed.
2416 19. Added -dd to debug only the daemon process.
2418 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2419 handles timeouts, both on the server side and network timeouts. Renamed the
2420 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2422 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2423 It is now set to "smtps".
2425 22. $host_address is now set to the target address during the checking of
2426 ignore_target_hosts.
2428 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2429 being passed; this would have caused $sender_host_name to have been used if
2430 matching the list had actually called for a host name (not very likely,
2431 since this list is usually IP addresses). A host name is now passed as
2434 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2435 code that specifies a non-restarting handler (typically sigaction() in
2436 modern systems) in an attempt to fix a rare and obscure crash bug.
2438 25. Narrowed the window for a race in the daemon that could cause it to ignore
2439 SIGCHLD signals. This is not a major problem, because they are used only to
2440 wake it up if nothing else does.
2442 26. A malformed maildirsize file could cause Exim to calculate negative values
2443 for the mailbox size or file count. Odd effects could occur as a result.
2444 The maildirsize information is now recalculated if the size or filecount
2447 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2448 support for a long time. Removed HAVE_SYS_VFS_H.
2450 28. Installed the latest version of exipick from John Jetmore.
2452 29. In an address list, if the pattern was not a regular expression, an empty
2453 subject address (from a bounce message) matched only if the pattern was an
2454 empty string. Non-empty patterns were not even tested. This was the wrong
2455 because it is perfectly reasonable to use an empty address as part of a
2456 database query. An empty address is now tested by patterns that are
2457 lookups. However, all the other forms of pattern expect the subject to
2458 contain a local part and a domain, and therefore, for them, an empty
2459 address still always fails if the pattern is not itself empty.
2461 30. Exim went into a mad DNS loop when attempting to do a callout where the
2462 host was specified on an smtp transport, and looking it up yielded more
2463 than one IP address.
2465 31. Re-factored the code for checking spool and log partition space into a
2466 function that finds that data and another that does the check. The former
2467 is then used to implement four new variables: $spool_space, $log_space,
2468 $spool_inodes, and $log_inodes.
2470 32. The RFC2047 encoding function was originally intended for short strings
2471 such as real names; it was not keeping to the 75-character limit for
2472 encoded words that the RFC imposes. It now respects the limit, and
2473 generates multiple encoded words if necessary. To be on the safe side, I
2474 have increased the buffer size for the ${rfc2047: expansion operator from
2477 33. It is now permitted to omit both strings after an "if" condition; if the
2478 condition is true, the result is "true". As before, when the second string
2479 is omitted, a false condition yields an empty string. This makes it less
2480 cumbersome to write custom ACL and router conditions.
2482 34. Failure to deliver a bounce message always caused it to be frozen, even if
2483 there was an errors_to setting on the router. The errors_to setting is now
2486 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2487 canonical form (fully expanded) before being placed in
2488 $sender_host_address.
2490 36. The table in the code that translates DNS record types into text (T_A to
2491 "A" for instance) was missing entries for NS and CNAME. It is just possible
2492 that this could have caused confusion if both these types were looked up
2493 for the same domain, because the text type is used as part of Exim's
2494 per-process caching. But the chance of anyone hitting this buglet seems
2497 37. The dnsdb lookup has been extended in a number of ways.
2499 (1) There is a new type, "zns", which walks up the domain tree until it
2500 finds some nameserver records. It should be used with care.
2502 (2) There is a new type, "mxh", which is like "mx" except that it returns
2503 just the host names, not the priorities.
2505 (3) It is now possible to give a list of domains (or IP addresses) to be
2506 looked up. The behaviour when one of the lookups defers can be
2507 controlled by a keyword.
2509 (4) It is now possible to specify the separator character for use when
2510 multiple records are returned.
2512 38. The dnslists ACL condition has been extended: it is now possible to supply
2513 a list of IP addresses and/or domains to be looked up in a particular DNS
2516 39. Added log_selector=+queue_time_overall.
2518 40. When running the queue in the test harness, wait just a tad after forking a
2519 delivery process, to get repeatability of debugging output.
2521 41. Include certificate and key file names in error message when GnuTLS fails
2522 to set them up, because the GnuTLS error message doesn't include the name
2523 of the failing file when there is a problem reading it.
2525 42. Allow both -bf and -bF in the same test run.
2527 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2529 44. The "Exiscan patch" is now merged into the mainline Exim source.
2531 45. Sometimes the final signoff response after QUIT could fail to get
2532 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2533 < 0 before doing a fflush(). This bug looks as though it goes back to the
2534 introduction of TLS in release 3.20, but "sometimes" must have been rare
2535 because the tests only now provoked it.
2537 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2538 (this can affect the format of dates).
2540 47. exim_tidydb, when checking for the continued existence of a message for
2541 which it has found a message-specific retry record, was not finding
2542 messages that were in split spool directories. Consequently, it was
2543 deleting retry records that should have stayed in existence.
2545 48. Steve fixed some bugs in eximstats.
2547 49. The SPA authentication driver was not abandoning authentication and moving
2548 on to the next authenticator when an expansion was forced to fail,
2549 contradicting the general specification for all authenticators. Instead it
2550 was generating a temporary error. It now behaves as specified.
2552 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2553 (the order specifies the preference for clients). The order is now AES256,
2554 AES128, 3DES, ARCFOUR128.
2556 51. Small patch to Sieve code - explicitly set From: when generating an
2559 52. Exim crashed if a remote delivery caused a very long error message to be
2560 recorded - for instance if somebody sent an entire SpamAssassin report back
2561 as a large number of 550 error lines. This bug was coincidentally fixed by
2562 increasing the size of one of Exim's internal buffers (big_buffer) that
2563 happened as part of the Exiscan merge. However, to be on the safe side, I
2564 have made the code more robust (and fixed the comments that describe what
2567 53. Now that there can be additional text after "Completed" in log lines (if
2568 the queue_time_overall log selector is set), a one-byte patch to exigrep
2569 was needed to allow it to recognize "Completed" as not the last thing in
2572 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2573 patch that reportedly fixes this has been added. I am not expert enough to
2574 create a test for it. This is what the patch creator wrote:
2576 "I found a little strange behaviour of ldap code when working with
2577 Windows 2003 AD Domain, where users was placed in more than one
2578 Organization Units. When I tried to give exim partial DN, the exit code
2579 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2580 But simultaneously result of request was absolutely normal ldap result,
2581 so I produce this patch..."
2583 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2584 REFERENCE, so I have modified the code to exclude the patch when that macro
2587 55. Some experimental protocols are using DNS PTR records for new purposes. The
2588 keys for these records are domain names, not reversed IP addresses. The
2589 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2590 leaves it alone. Component reversal etc. now happens only for IP addresses.
2593 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2595 57. Double the size of the debug message buffer (to 2048) so that more of very
2596 long debug lines gets shown.
2598 58. The exicyclog utility now does better if the number of log files to keep
2599 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2601 59. Two changes related to the smtp_active_hostname option:
2603 (1) $smtp_active_hostname is now available as a variable.
2604 (2) The default for smtp_banner uses $smtp_active_hostname instead
2605 of $primary_hostname.
2607 60. The host_aton() function is supposed to be passed a string that is known
2608 to be a valid IP address. However, in the case of IPv6 addresses, it was
2609 not checking this. This is a hostage to fortune. Exim now panics and dies
2610 if the condition is not met. A case was found where this could be provoked
2611 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2612 components; fortuitously, this particular loophole had already been fixed
2613 by change 4.50/55 above.
2615 If there are any other similar loopholes, the new check in host_aton()
2616 itself should stop them being exploited. The report I received stated that
2617 data on the command line could provoke the exploit when Exim was running as
2618 exim, but did not say which command line option was involved. All I could
2619 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2620 running as the user.
2623 61. There was a buffer overflow vulnerability in the SPA authentication code
2624 (which came originally from the Samba project). I have added a test to the
2625 spa_base64_to_bits() function which I hope fixes it.
2628 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2629 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2631 63. The daemon start-up calls getloadavg() while still root for those OS that
2632 need the first call to be done as root, but it missed one case: when
2633 deliver_queue_load_max is set with deliver_drop_privilege. This is
2634 necessary for the benefit of the queue runner, because there is no re-exec
2635 when deliver_drop_privilege is set.
2637 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2638 This has been fixed.
2640 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2641 was in use, was not putting the data itself into the right store pool;
2642 consequently, it could be overwritten for a subsequent message in the same
2643 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2646 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2648 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2649 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2650 the function were treating the return as a boolean value, which happened to
2651 work because 0=false and not-0=true, but is not correct code.
2653 68. The host_aton() function was not handling scoped IPv6 addresses (those
2654 with, for example, "%eth0" on the end) correctly.
2656 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2657 negated items (that is, ~something) in unsigned ints. Some compilers
2658 apparently mutter when there is no cast.
2660 70. If an address verification called from an ACL failed, and did not produce a
2661 user-specific message (i.e. there was only a "system" message), nothing was
2662 put in $acl_verify_message. In this situation, it now puts the system
2665 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2666 session; change 4.31/43 added the unwanted input to the log line - except
2667 that it did not do this in the start of session case. It now does.
2669 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2670 This is wrong and can cause the other end to generate a synchronization
2671 error if it is another Exim or anything else that does the synchronization
2672 check. A QUIT command is no longer sent after a timeout.
2674 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2675 during host lookups.
2677 74. The defer_ok option of callout verification was not working if it was used
2678 when verifying addresses in header lines, that is, for this case:
2680 verify = header_sender/callout=defer_ok
2682 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2683 those file descriptors could be used for SMTP connections. If anything
2684 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2685 could be sent to the SMTP client, causing chaos. The daemon now opens
2686 stdin, stdout, and stderr to /dev/null when it puts itself into the
2689 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2690 log by default. The user can override this with suitable Perl magic.
2692 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2693 the log message when discard triggers, was not working for the DATA ACL or
2694 for the non-SMTP ACL.
2696 78. Error message wording change in sieve.c.
2698 79. If smtp_accept_max_per_host was set, the number of connections could be
2699 restricted to fewer than expected, because the daemon was trying to set up
2700 a new connection before checking whether the processes handling previous
2701 connections had finished. The check for completed processes is now done
2702 earlier. On busy systems, this bug wouldn't be noticed because something
2703 else would have woken the daemon, and it would have reaped the completed
2706 80. If a message was submitted locally by a user whose login name contained one
2707 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2708 It caused a spool format error. I have fixed the spool reading code. A
2709 related problem was that the "from" clause in the Received: line became
2710 illegal because of the space(s). It is now covered by ${quote_local_part.
2712 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2715 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2717 83. Patch from Sieve maintainer to fix unterminated string problem in
2718 "vacation" handling.
2720 84. Some minor changes to the Linux configuration files to help with other
2721 OS variants using glibc.
2723 85. One more patch for Sieve to update vacation handling to latest spec.
2726 ----------------------------------------------------
2727 See the note above about the 4.44 and 4.50 releases.
2728 ----------------------------------------------------
2734 1. Change 4.43/35 introduced a bug that caused file counts to be
2735 incorrectly computed when quota_filecount was set in an appendfile
2738 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2739 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2742 3. Give more explanation in the error message when the command for a transport
2743 filter fails to execute.
2745 4. There are several places where Exim runs a non-Exim command in a
2746 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2747 was being done only for the command run by the queryprogram router. It is
2748 now done for all such subprocesses. The other cases are: ${run, transport
2749 filters, and the commands run by the lmtp and pipe transports.
2751 5. Some older OS have a limit of 256 on the maximum number of file
2752 descriptors. Exim was using setrlimit() to set 1000 as a large value
2753 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2754 systems. I've change it so that if it can't get 1000, it tries for 256.
2756 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2757 was an oversight, and furthermore, ever since the addition of extra
2758 controls (e.g. 4.43/32), the checks on when to allow different forms of
2759 "control" were broken. There should now be diagnostics for all cases when a
2760 control that does not make sense is encountered.
2762 7. $recipients is now available in the predata ACL (oversight).
2764 8. Tidy the search cache before the fork to do a delivery from a message
2765 received from the command line. Otherwise the child will trigger a lookup
2766 failure and thereby defer the delivery if it tries to use (for example) a
2767 cached ldap connection that the parent has called unbind on.
2769 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2770 of $address_data from the recipient verification was clobbered by the
2771 sender verification.
2773 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2774 was its contents. (It was OK if the option was not defined at all.)
2776 11. A "Completed" log line is now written for messages that are removed from
2777 the spool by the -Mrm option.
2779 12. $host_address is now set to the target address during the checking of
2780 ignore_target_hosts.
2782 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2783 being passed; this would have caused $sender_host_name to have been used if
2784 matching the list had actually called for a host name (not very likely,
2785 since this list is usually IP addresses). A host name is now passed as
2788 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2789 code that specifies a non-restarting handler (typically sigaction() in
2790 modern systems) in an attempt to fix a rare and obscure crash bug.
2792 15. Narrowed the window for a race in the daemon that could cause it to ignore
2793 SIGCHLD signals. This is not a major problem, because they are used only to
2794 wake it up if nothing else does.
2796 16. A malformed maildirsize file could cause Exim to calculate negative values
2797 for the mailbox size or file count. Odd effects could occur as a result.
2798 The maildirsize information is now recalculated if the size or filecount
2801 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2802 support for a long time. Removed HAVE_SYS_VFS_H.
2804 18. Updated exipick to current release from John Jetmore.
2806 19. Allow an empty sender to be matched against a lookup in an address list.
2807 Previously the only cases considered were a regular expression, or an
2810 20. Exim went into a mad DNS lookup loop when doing a callout where the
2811 host was specified on the transport, if the DNS lookup yielded more than
2814 21. The RFC2047 encoding function was originally intended for short strings
2815 such as real names; it was not keeping to the 75-character limit for
2816 encoded words that the RFC imposes. It now respects the limit, and
2817 generates multiple encoded words if necessary. To be on the safe side, I
2818 have increased the buffer size for the ${rfc2047: expansion operator from
2821 22. Failure to deliver a bounce message always caused it to be frozen, even if
2822 there was an errors_to setting on the router. The errors_to setting is now
2825 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2826 canonical form (fully expanded) before being placed in
2827 $sender_host_address.
2829 24. Updated eximstats to version 1.33
2831 25. Include certificate and key file names in error message when GnuTLS fails
2832 to set them up, because the GnuTLS error message doesn't include the name
2833 of the failing file when there is a problem reading it.
2835 26. Expand error message when OpenSSL has problems setting up cert/key files.
2838 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2839 (this can affect the format of dates).
2841 28. exim_tidydb, when checking for the continued existence of a message for
2842 which it has found a message-specific retry record, was not finding
2843 messages that were in split spool directories. Consequently, it was
2844 deleting retry records that should have stayed in existence.
2846 29. eximstats updated to version 1.35
2847 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2848 1.35 - bugfix such that pie charts by volume are generated correctly
2850 30. The SPA authentication driver was not abandoning authentication and moving
2851 on to the next authenticator when an expansion was forced to fail,
2852 contradicting the general specification for all authenticators. Instead it
2853 was generating a temporary error. It now behaves as specified.
2855 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2856 (the order specifies the preference for clients). The order is now AES256,
2857 AES128, 3DES, ARCFOUR128.
2859 31. Small patch to Sieve code - explicitly set From: when generating an
2862 32. Exim crashed if a remote delivery caused a very long error message to be
2863 recorded - for instance if somebody sent an entire SpamAssassin report back
2864 as a large number of 550 error lines. This bug was coincidentally fixed by
2865 increasing the size of one of Exim's internal buffers (big_buffer) that
2866 happened as part of the Exiscan merge. However, to be on the safe side, I
2867 have made the code more robust (and fixed the comments that describe what
2870 33. Some experimental protocols are using DNS PTR records for new purposes. The
2871 keys for these records are domain names, not reversed IP addresses. The
2872 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2873 leaves it alone. Component reversal etc. now happens only for IP addresses.
2876 34. The host_aton() function is supposed to be passed a string that is known
2877 to be a valid IP address. However, in the case of IPv6 addresses, it was
2878 not checking this. This is a hostage to fortune. Exim now panics and dies
2879 if the condition is not met. A case was found where this could be provoked
2880 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2881 components; fortuitously, this particular loophole had already been fixed
2882 by change 4.50/55 or 4.44/33 above.
2884 If there are any other similar loopholes, the new check in host_aton()
2885 itself should stop them being exploited. The report I received stated that
2886 data on the command line could provoke the exploit when Exim was running as
2887 exim, but did not say which command line option was involved. All I could
2888 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2889 running as the user.
2892 35. There was a buffer overflow vulnerability in the SPA authentication code
2893 (which came originally from the Samba project). I have added a test to the
2894 spa_base64_to_bits() function which I hope fixes it.
2897 36. The daemon start-up calls getloadavg() while still root for those OS that
2898 need the first call to be done as root, but it missed one case: when
2899 deliver_queue_load_max is set with deliver_drop_privilege. This is
2900 necessary for the benefit of the queue runner, because there is no re-exec
2901 when deliver_drop_privilege is set.
2903 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2904 was in use, was not putting the data itself into the right store pool;
2905 consequently, it could be overwritten for a subsequent message in the same
2906 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2909 38. Sometimes the final signoff response after QUIT could fail to get
2910 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2911 < 0 before doing a fflush(). This bug looks as though it goes back to the
2912 introduction of TLS in release 3.20, but "sometimes" must have been rare
2913 because the tests only now provoked it.
2919 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2920 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2921 Now it returns an integer. A number of calls were still expecting a T/F
2922 return. Fortuitously, in all cases, the tests worked in OK situations,
2923 which is the norm. However, things would have gone wrong on any write
2924 failures on the smtp file descriptor. This function is used when sending
2925 messages over SMTP and also when doing verify callouts.
2927 2. When Exim is called to do synchronous delivery of a locally submitted
2928 message (the -odf or -odi options), it no longer closes stderr before doing
2931 3. Implemented the mua_wrapper option.
2933 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2935 5. Implemented the functions header_remove(), header_testname(),
2936 header_add_at_position(), and receive_remove_recipient(), and exported them
2939 6. If an ACL "warn" statement specified the addition of headers, Exim already
2940 inserted X-ACL-Warn: at the start if there was no header name. However, it
2941 was not making this test for the second and subsequent header lines if
2942 there were newlines in the string. This meant that an invalid header could
2943 be inserted if Exim was badly configured.
2945 7. Allow an ACL "warn" statement to add header lines at the start or after all
2946 the Received: headers, as well as at the end.
2948 8. Added the rcpt_4xx retry error code.
2950 9. Added postmaster_mailfrom=xxx to callout verification option.
2952 10. Added mailfrom=xxxx to the callout verification option, for verify=
2955 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2956 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2958 12. Inserted some casts to stop certain compilers warning when using pointer
2959 differences as field lengths or precisions in printf-type calls (mostly
2960 affecting debugging statements).
2962 13. Added optional readline() support for -be (dynamically loaded).
2964 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2965 same clock tick as a message's arrival, so that its received time was the
2966 same as the "first fail" time on the retry record, and that message
2967 remained on the queue past the ultimate address timeout, every queue runner
2968 would try a delivery (because it was past the ultimate address timeout) but
2969 after another failure, the ultimate address timeout, which should have then
2970 bounced the address, did not kick in. This was a "< instead of <=" error;
2971 in most cases the first failure would have been in the next clock tick
2972 after the received time, and all would be well.
2974 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2975 being recognized when the domain list was tested by the match_domain
2976 condition in an expansion string.
2978 16. Added the ${str2b64: operator.
2980 17. Exim was always calling setrlimit() to set a large limit for the number of
2981 processes, without checking whether the existing limit was already
2982 adequate. (It did check for the limit on file descriptors.) Furthermore,
2983 errors from getrlimit() and setrlimit() were being ignored. Now they are
2984 logged to the main and panic logs, but Exim does carry on, to try to do its
2985 job under whatever limits there are.
2987 18. Imported PCRE 5.0.
2989 19. Trivial typo in log message " temporarily refused connection" (the leading
2992 20. If the log selector return_path_on_delivery was set and an address was
2993 redirected to /dev/null, the delivery process crashed because it assumed
2994 that a return path would always be set for a "successful" delivery. In this
2995 case, the whole delivery is bypassed as an optimization, and therefore no
2998 21. Internal re-arrangement: the function for sending a challenge and reading
2999 a response while authentication was assuming a zero-terminated challenge
3000 string. It's now changed to take a pointer and a length, to allow for
3001 binary data in such strings.
3003 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3005 23. Exim was not respecting finduser_retries when seeking the login of the
3006 uid under which it was called; it was always trying 10 times. (The default
3007 setting of finduser_retries is zero.) Also, it was sleeping after the final
3008 failure, which is pointless.
3010 24. Implemented tls_on_connect_ports.
3012 25. Implemented acl_smtp_predata.
3014 26. If the domain in control=submission is set empty, Exim assumes that the
3015 authenticated id is a complete email address when it generates From: or
3016 Sender: header lines.
3018 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3019 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3020 chown and chgrp in /bin and hostname in /usr/bin.
3022 28. Exim was keeping the "process log" file open after each use, just as it
3023 does for the main log. This opens the possibility of it remaining open for
3024 long periods when the USR1 signal hits a daemon. Occasional processlog
3025 errors were reported, that could have been caused by this. Anyway, it seems
3026 much more sensible not to leave this file open at all, so that is what now
3029 29. The long-running daemon process does not normally write to the log once it
3030 has entered its main loop, and it closes the log before doing so. This is
3031 so that log files can straightforwardly be renamed and moved. However,
3032 there are a couple of unusual error situations where the daemon does write
3033 log entries, and I had neglected to close the log afterwards.
3035 30. The text of an SMTP error response that was received during a remote
3036 delivery was being truncated at 512 bytes. This is too short for some of
3037 the long messages that one sometimes sees. I've increased the limit to
3040 31. It is now possible to make retry rules that apply only when a message has a
3041 specific sender, in particular, an empty sender.
3043 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3044 it possible to be selective about when SMTP synchronization is enforced.
3046 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3048 32. Implemented hosts_connection_nolog.
3050 33. Added an ACL for QUIT.
3052 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3055 35. Added mailbox_size and mailbox_filecount to appendfile.
3057 36. Added control = no_multiline_responses to ACLs.
3059 37. There was a bug in the logic of the code that waits for the clock to tick
3060 in the case where the clock went backwards by a substantial amount such
3061 that the microsecond fraction of "now" was more than the microsecond
3062 fraction of "then" (but the whole seconds number was less).
3064 38. Added support for the libradius Radius client library this is found on
3065 FreeBSD (previously only the radiusclient library was supported).
3071 1. When certain lookups returned multiple values in the form name=value, the
3072 quoting of the values was not always being done properly. Specifically:
3073 (a) If the value started with a double quote, but contained no whitespace,
3075 (b) If the value contained whitespace other than a space character (i.e.
3076 tabs or newlines or carriage returns) it was not quoted.
3077 This fix has been applied to the mysql and pgsql lookups by writing a
3078 separate quoting function and calling it from the lookup code. The fix
3079 should probably also be applied to nisplus, ibase and oracle lookups, but
3080 since I cannot test any of those, I have not disturbed their existing code.
3082 2. A hit in the callout cache for a specific address caused a log line with no
3083 reason for rejecting RCPT. Now it says "Previous (cached) callout
3084 verification failure".
3086 3. There was an off-by-one bug in the queryprogram router. An over-long
3087 return line was truncated at 256 instead of 255 characters, thereby
3088 overflowing its buffer with the terminating zero. As well as fixing this, I
3089 have increased the buffer size to 1024 (and made a note to document this).
3091 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3092 when Exim is waiting for an SMTP response from a remote server, Exim
3093 restarts its select() call on the socket, thereby resetting its timeout.
3094 This is not a problem when such interrupts are rare. Somebody set up a cron
3095 job to run exiwhat every 2 minutes, which is less than the normal select()
3096 timeout (5 or 10 minutes). This meant that the select() timeout never
3097 kicked in because it was always reset. I have fixed this by comparing the
3098 time when an interrupt arrives with the time at the start of the first call
3099 to select(). If more time than the timeout has elapsed, the interrupt is
3100 treated as a timeout.
3102 5. Some internal re-factoring in preparation for the addition of Sieve
3103 extensions (by MH). In particular, the "personal" test is moved to a
3104 separate function, and given an option for scanning Cc: and Bcc: (which is
3105 not set for Exim filters).
3107 6. When Exim created an email address using the login of the caller as the
3108 local part (e.g. when creating a From: or Sender: header line), it was not
3109 quoting the local part when it contained special characters such as @.
3111 7. Installed new OpenBSD configuration files.
3113 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3114 try to make them clearer.
3116 9. Callout options, other than the timeout value, were being ignored when
3117 verifying sender addresses in header lines. For example, when using
3119 verify = header_sender/callout=no_cache
3121 the cache was (incorrectly) being used.
3123 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3124 only the Cygwin environment.
3126 11. When return_path_on_delivery was set as a log selector, if different remote
3127 addresses in the same message used different return paths and parallel
3128 remote delivery occurred, the wrong values would sometimes be logged.
3129 (Whenever a remote delivery process finished, the return path value from
3130 the most recently started remote delivery process was logged.)
3132 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3133 lines when AUTH and/or TLS are in use. This is the "received protocol"
3134 field. Exim used to use "asmtp" for authenticated SMTP, without any
3135 indication (in the protocol name) for TLS use. Now it follows the RFC and
3136 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3137 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3138 names appear in log lines as well as in Received: header lines.
3140 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3141 extensions, and comparison tests, and to fix some bugs.
3143 14. Changes to the "personal" filter test:
3145 (1) The test was buggy in that it was just doing the equivalent of
3146 "contains" tests on header lines. For example, if a user's address was
3147 anne@some.where, the "personal" test would incorrectly be true for
3149 To: susanne@some.where
3151 This test is now done by extracting each address from the header in turn,
3152 and checking the entire address. Other tests that are part of "personal"
3153 are now done using regular expressions (for example, to check local parts
3154 of addresses in From: header lines).
3156 (2) The list of non-personal local parts in From: addresses has been
3157 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3158 taken from the Sieve specification recommendations.
3160 (3) If the message contains any header line starting with "List-" it is
3161 treated as non-personal.
3163 (4) The test for "circular" in the Subject: header line has been removed
3164 because it now seems ill-conceived.
3166 15. Minor typos in src/EDITME comments corrected.
3168 16. Installed latest exipick from John Jetmore.
3170 17. If headers_add on a router specified a text string that was too long for
3171 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3172 of string_sprintf() is now avoided.
3174 18. $message_body_size was not set (it was always zero) when running the DATA
3175 ACL and the local_scan() function.
3177 19. For the "mail" command in an Exim filter, no default was being set for
3178 the once_repeat time, causing a random time value to be used if "once" was
3179 specified. (If the value happened to be <= 0, no repeat happened.) The
3180 default is now 0s, meaning "never repeat". The "vacation" command was OK
3181 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3182 (I found it when inspecting the code).
3184 20. There is now an overall timeout for performing a callout verification. It
3185 defaults to 4 times the callout timeout, which applies to individual SMTP
3186 commands during the callout. The overall timeout applies when there is more
3187 than one host that can be tried. The timeout is checked before trying the
3188 next host. This prevents very long delays if there are a large number of
3189 hosts and all are timing out (e.g. when the network connections are timing
3190 out). The value of the overall timeout can be changed by specifying an
3191 additional sub-option for "callout", called "maxwait". For example:
3193 verify = sender/callout=5s,maxwait=20s
3195 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3196 to the end before writing, but this should make it even safer).
3198 22. Exim was forgetting that it had advertised PIPELINING for the second and
3199 subsequent messages on an SMTP connection. It was also not resetting its
3200 memory on STARTTLS and an internal HELO.
3202 23. When Exim logs an SMTP synchronization error within a session, it now
3203 records whether PIPELINING has been advertised or not.
3205 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3206 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3207 rather than long int.
3209 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3211 26. Added the never_mail option to autoreply.
3217 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3218 crash if the getsockname() call failed; this can happen if a connection is
3219 closed very soon after it is established. The problem was simply in the
3220 order in which certain operations were done, causing Exim to try to write
3221 to the SMTP stream before it had set up the file descriptor. The bug has
3222 been fixed by making things happen in the correct order.
3228 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3229 before the connection was closed, thus losing the rejection response.
3231 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3232 some early Solaris releases, but causes trouble in current releases where
3233 socklen_t is defined.
3235 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3238 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3241 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3242 that releasing the top of it at the end releases what was used for sub-
3243 expansions (unless the block got too big). However, discard this block if
3244 the first thing is a variable or header, so that we can use its block when
3245 it is dynamic (useful for very large $message_headers, for example).
3247 6. Lookups now cache *every* query, not just the most recent. A new, separate
3248 store pool is used for this. It can be recovered when all lookup caches are
3249 flushed. Lookups now release memory at the end of their result strings.
3250 This has involved some general refactoring of the lookup sources.
3252 7. Some code has been added to the store_xxx() functions to reduce the amount
3253 of flapping under certain conditions.
3255 8. log_incoming_interface used to affect only the <= reception log lines. Now
3256 it causes the local interface and port to be added to several more SMTP log
3257 lines, for example "SMTP connection from", and rejection lines.
3259 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3261 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3263 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3264 could be overwritten at the end of the current message (or the start of a
3265 new message if it was set in a HELO ACL). The value is now preserved for
3266 the duration of the SMTP connection.
3268 12. If a transport had a headers_rewrite setting, and a matching header line
3269 contained an unqualified address, that address was qualified, even if it
3270 did not match any rewriting rules. The underlying bug was that the values
3271 of the flags that permit the existence of unqualified sender and recipient
3272 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3273 non-local messages, and by -bnq for local messages) were not being
3274 preserved with the message after it was received.
3276 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3277 "next input=" as part of the text comprising the host identity instead of
3278 the correct text. The code was using the same buffer for two different
3279 strings. However, depending on which order the printing function evaluated
3280 its arguments, the bug did not always show up. Under Linux, for example, my
3281 test suite worked just fine.
3283 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3284 doesn't work with some older versions of Perl. It has been changed to "my",
3285 which in any case is probably the better facility to use.
3287 15. A really picky compiler found some instances of statements for creating
3288 error messages that either had too many or two few arguments for the format
3291 16. The size of the buffer for calls to the DNS resolver has been increased
3292 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3293 for addresses that have a lot of PTR records. This alleviates a problem; it
3294 does not fully solve it.
3296 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3297 buffer now truncates the list and logs the incident, which is the same
3298 action as happens when Exim is looking up a host name and its aliases.
3299 Previously in this situation something unpredictable would happen;
3300 sometimes it was "internal error: store_reset failed".
3302 18. If a server dropped the connection unexpectedly when an Exim client was
3303 using GnuTLS and trying to read a response, the client delivery process
3304 crashed while trying to generate an error log message.
3306 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3307 string, for example:
3309 warn message = H1: something\nH2: something
3311 the text was added as a single header line from Exim's point of view
3312 though it ended up OK in the delivered message. However, searching for the
3313 second and subsequent header lines using $h_h2: did not work. This has been
3314 fixed. Similarly, if a system filter added multiple headers in this way,
3315 the routers could not see them.
3317 20. Expanded the error message when iplsearch is called with an invalid key to
3318 suggest using net-iplsearch in a host list.
3320 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3321 ACLs are no longer actually imposed (and a message to that effect is
3324 22. If a "gecos" field in a passwd entry contained escaped characters, in
3325 particular, if it contained a \" sequence, Exim got it wrong when building
3326 a From: or a Sender: header from that name. A second bug also caused
3327 incorrect handling when an unquoted " was present following a character
3328 that needed quoting.
3330 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3331 was not being matched caselessly.
3333 24. Arranged for all hyphens in the exim.8 source to be escaped with
3336 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3337 properly. Recipient callout cache records were still being keyed to include
3338 the sender, even when use_sender was set false. This led to far more
3339 callouts that were necessary. The sender is no longer included in the key
3340 when use_sender is false.
3342 26. Added "control = submission" modifier to ACLs.
3344 27. Added the ${base62d: operator to decode base 62 numbers.
3346 28. dnsdb lookups can now access SRV records.
3348 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3349 the configuration file.
3351 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3352 (-v) mode. This makes the output for a verbose queue run more intelligible.
3354 31. Added a use_postmaster feature to recipient callouts.
3356 32. Added the $body_zerocount variable, containing the number of binary zero
3357 bytes in the message body.
3359 33. The time of last modification of the "new" subdirectory is now used as the
3360 "mailbox time last read" when there is a quota error for a maildir
3363 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3365 35. Added +ignore_unknown as a special item in host lists.
3367 36. Code for decoding IPv6 addresses in host lists is now included, even if
3368 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3369 address was recognized as an IP address, but was then not correctly decoded
3370 into binary, causing unexpected and incorrect effects when compared with
3377 1. Very minor rewording of debugging text in manualroute to say "list of
3378 hosts" instead of "hostlist".
3380 2. If verify=header_syntax was set, and a header line with an unqualified
3381 address (no domain) and a large number of spaces between the end of the
3382 name and the colon was received, the reception process suffered a buffer
3383 overflow, and (when I tested it) crashed. This was caused by some obsolete
3384 code that should have been removed. The fix is to remove it!
3386 3. When running in the test harness, delay a bit after writing a bounce
3387 message to get a bit more predictability in the log output.
3389 4. Added a call to search_tidyup() just before forking a reception process. In
3390 theory, someone could use a lookup in the expansion of smtp_accept_max_
3391 per_host which, without the tidyup, could leave open a database connection.
3393 5. Added the variables $recipient_data and $sender_data which get set from a
3394 lookup success in an ACL "recipients" or "senders" condition, or a router
3395 "senders" option, similar to $domain_data and $local_part_data.
3397 6. Moved the writing of debug_print from before to after the "senders" test
3400 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3401 problems for message scanning, either using a data ACL, or using
3402 local_scan() because the Received: header was not generated till after they
3403 were called (in order to set the time as the time of reception completion).
3404 I have revised the way this works. The header is now generated after the
3405 body is received, but before the ACL or local_scan() are called. After they
3406 are run, the timestamp in the header is updated.
3412 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3413 before starting a queue runner without re-exec. This happened only when
3414 deliver_drop_privilege was set or when the Exim user was set to root. The
3415 effect of the bug was that timeouts during subsequent deliveries caused
3416 crashes instead of being properly handled. The handler is now left at its
3417 default (and expected) setting.
3419 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3420 message, again when deliver_drop_privilege is set or Exim is run as root.
3421 The bug described in (1) was not present in this case, but the tidying up
3422 of the other signals was missing. I have made the two cases consistent.
3424 3. The ignore_target_hosts setting on a manualroute router was being ignored
3425 for hosts that were looked up using the /MX notation.
3427 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3430 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3431 operated on the sender address. After changing the $sender_address to <>
3432 for the sender address verify, Exim was re-instated it as the original
3433 (before rewriting) address, but remembering that it had rewritten it, so it
3434 wasn't rewriting it again. This bug also had the effect of breaking the
3435 sender address verification caching when the sender address was rewritten.
3437 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3438 This has been changed so that if the ip literal address matches
3439 ignore_target_hosts, the router declines.
3441 7. Added expansion conditions match_domain, match_address, and match_local_
3442 part (NOT match_host).
3444 8. The placeholder for the Received: header didn't have a length field set.
3446 9. Added code to Exim itself and to exim_lock to test for a specific race
3447 condition that could lead to file corruption when using MBX delivery. The
3448 issue is with the lockfile that is created in /tmp. If this file is removed
3449 after a process has opened it but before that process has acquired a lock,
3450 there is the potential for a second process to recreate the file and also
3451 acquire a lock. This could lead to two Exim processes writing to the file
3452 at the same time. The added code performs the same test as UW imapd; it
3453 checks after acquiring the lock that its file descriptor still refers to
3454 the same named file.
3456 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3457 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3460 11. Added the smtp_active_hostname option. If used, this will typically be made
3461 to depend on the incoming interface address. Because $interface_address is
3462 not set up until the daemon has forked a reception process, error responses
3463 that can happen earlier (such as "too many connections") no longer contain
3466 12. If an expansion in a condition on a "warn" statement fails because a lookup
3467 defers, the "warn" statement is abandoned, and the next ACL statement is
3468 processed. Previously this caused the whole ACL to be aborted.
3470 13. Added the iplsearch lookup type.
3472 14. Added ident_timeout as a log selector.
3474 15. Added tls_certificate_verified as a log selector.
3476 16. Added a global option tls_require_ciphers (compare the smtp transport
3477 option of the same name). This controls incoming TLS connections.
3479 17. I finally figured out how to make tls_require_ciphers do a similar thing
3480 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3481 before starting the TLS session.
3483 18. Tabs are now shown as \t in -bP output.
3485 19. If the log selector return_path_on_delivery was set, Exim crashed when
3486 bouncing a message because it had too many Received: header lines.
3488 20. If two routers both had headers_remove settings, and the first one included
3489 a superfluous trailing colon, the final name in the first list and the
3490 first name in the second list were incorrectly joined into one item (with a
3491 colon in the middle).
3497 1. Added -C and -D options to the exinext utility, mainly to make it easier
3498 to include in the automated testing, but these could be helpful when
3499 multiple configurations are in use.
3501 2. The exinext utility was not formatting the output nicely when there was
3502 an alternate port involved in the retry record key, nor when there was a
3503 message id as well (for retries that were specific to a specific message
3504 and a specific host). It was also confused by IPv6 addresses, because of
3505 the additional colons they contain. I have fixed the IPv4 problem, and
3506 patched it up to do a reasonable job for IPv6.
3508 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3509 delivery, the log line now contains "pipelined" if PIPELINING was used.
3511 4. An SMTP transport process used to panic and die if the bind() call to set
3512 an explicit outgoing interface failed. This has been changed; it is now
3513 treated in the same way as a connect() failure.
3515 5. A reference to $sender_host_name in the part of a conditional expansion
3516 that was being skipped was still causing a DNS lookup. This no longer
3519 6. The def: expansion condition was not recognizing references to header lines
3520 that used bh_ and bheader_.
3522 7. Added the _cache feature to named lists.
3524 8. The code for checking quota_filecount in the appendfile transport was
3525 allowing one more file than it should have been.
3527 9. For compatibility with Sendmail, the command line option
3535 and sets the incoming protocol and host name (for trusted callers). The
3536 host name and its colon can be omitted when only the protocol is to be set.
3537 Note the Exim already has two private options, -pd and -ps, that refer to
3538 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3539 "s", but I don't think that's a major issue.
3541 10. A number of refactoring changes to the code, none of which should affect
3544 (a) The number of logging options was getting close to filling up the
3545 32-bit word that was used as a bit map. I have split them into two classes:
3546 those that are passed in the argument to log_write(), and those that are
3547 only ever tested independently outside of that function. These are now in
3548 separate 32-bit words, so there is plenty of room for expansion again.
3549 There is no change in the user interface or the logging behaviour.
3551 (b) When building, for example, log lines, the code previously used a
3552 macro that called string_cat() twice, in order to add two strings. This is
3553 not really sufficiently general. Furthermore, there was one instance where
3554 it was actually wrong because one of the argument was used twice, and in
3555 one call a function was used. (As it happened, calling the function twice
3556 did not affect the overall behaviour.) The macro has been replaced by a
3557 function that can join an arbitrary number of extra strings onto a growing
3560 (c) The code for expansion conditions now uses a table and a binary chop
3561 instead of a serial search (which was left over from when there were very
3562 few conditions). Also, it now recognizes conditions like "pam" even when
3563 the relevant support is not compiled in: a suitably worded error message is
3564 given if an attempt is made to use such a condition.
3566 11. Added ${time_interval:xxxxx}.
3568 12. A bug was causing one of the ddress fields not to be passed back correctly
3569 from remote delivery subprocesses. The field in question was not being
3570 subsequently used, so this caused to problems in practice.
3572 13. Added new log selectors queue_time and deliver_time.
3574 14. Might have fixed a bug in maildirsizefile handling that threw up
3575 "unexpected character" debug warnings, and recalculated the data
3576 unnecessarily. In any case, I expanded the warning message to give more
3579 15. Added the message "Restricted characters in address" to the statements in
3580 the default ACL that block characters like @ and % in local parts.
3582 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3583 Three changes have been made:
3585 (a) There was a serious bug; a negative response to MAIL caused the whole
3586 recipient domain to be cached as invalid, thereby blocking all messages
3587 to all local parts at the same domain, from all senders. This bug has
3588 been fixed. The domain is no longer cached after a negative response to
3589 MAIL if the sender used is not empty.
3591 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3594 (c) A new callout option, "use_sender" has been added for people who want
3595 the modified behaviour.
3601 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3604 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3605 indeed breaks things for older releases.
3607 3. Added additional logging to the case where there is a problem reading data
3608 from a filter that is running in a subprocess using a pipe, in order to
3609 try to track down a specific problem.
3611 4. Testing facility fudge: when running in the test harness and attempting
3612 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3613 getting "No route to host". Convert this to a timeout.
3615 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3618 6. Some OS don't have socklen_t but use size_t instead. This affects the
3619 fifth argument of getsockopt() amongst other things. This is now
3620 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3621 can be set for individual OS. I have set it for SunOS5, OSF1, and
3622 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3623 some earlier ones do not.
3625 7. Change 4.30/15 was not doing the test caselessly.
3627 8. The standard form for an IPv6 address literal was being rejected by address
3628 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3629 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3630 this, as well as the form without the "IPv6" on the front (but only when
3631 address literals are enabled, of course).
3633 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3635 10. Exim crashed if a message with an empty sender address specified by -f
3636 encountered a router with an errors_to setting. This could be provoked only
3637 by a command such as
3641 where an empty string was supplied; "<>" did not hit this bug.
3643 11. Installed PCRE release 4.5.
3645 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3646 remained set. It is now erased.
3648 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3649 times from message ids (which are base 36 rather than the normal 62).
3651 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3652 were being counted as actual protocol errors, and logged if the log
3653 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3654 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3655 and DATA following a set of rejected RCPTs do not count as protocol errors.
3656 In other words, Exim assumes they were pipelined, though this may not
3657 actually be the case. Of course, in all cases the client gets an
3658 appropriate error code.
3660 15. If a lookup fails in an ACL condition, a message about the failure may
3661 be available; it is used if testing the ACL cannot continue, because most
3662 such messages specify what the cause of the deferral is. However, some
3663 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3664 that caused an old message to be retained and used if a later statement
3665 caused a defer, replacing the real cause of the deferral.
3667 16. If an IP address had so many PTR records that the DNS lookup buffer
3668 was not large enough to hold them, Exim could crash while trying to process
3669 the truncated data. It now detects and logs this case.
3671 17. Further to 4.21/58, another change has been made: if (and only if) the
3672 first line of a message (the first header line) ends with CRLF, a bare LF
3673 in a subsequent header line has a space inserted after it, so as not to
3674 terminate the header.
3676 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3677 unnecessarily, used atoi() instead of strtol(), and didn't check the
3678 termination when getting file sizes from file names by regex.
3680 19. Completely re-implemented the support for maildirsize files, in the light
3681 of a number of problems with the previous contributed implementation
3682 (4.30/29). In particular:
3684 . If the quota is zero, the maildirsize file is maintained, but no quota is
3687 . If the maildir directory does not exist, it is created before any attempt
3688 to write a maildirsize file.
3690 . The quota value in the file is just a cache; if the quota is changed in
3691 the transport, the new value overrides.
3693 . A regular expression is available for excluding directories from the
3696 20. The autoreply transport checks the characters in options that define the
3697 message's headers; it allows continued headers, but it was checking with
3698 isspace() after an embedded newline instead of explicitly looking for a
3701 21. If all the "regular" hosts to which an address was routed had passed their
3702 expiry times, and had not reached their retry times, the address was
3703 bounced, even if fallback hosts were defined. Now Exim should go on to try
3706 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3707 equivalent code in the SMTP transport. Some hosts send humungous responses
3708 to HELO/EHLO, more than 1024 it seems.
3710 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3711 gives compiler warnings in some environments. I've now done it "properly",
3714 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3715 (because of problems with the built-in one) was declared to return uschar *
3716 instead of char *, causing compiler failure.
3718 25. Fixed a file descriptor leak when processing alias/forward files.
3720 26. Fixed a minor format string issue in dbfn.c.
3722 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3724 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3725 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3728 29. When a sender address is verified, it is cached, to save repeating the test
3729 when there is more than one recipient in a message. However, when the
3730 verification involves a callout, it is possible for different callout
3731 options to be set for different recipients. It is too complicated to keep
3732 track of this in the cache, so now Exim always runs a verification when a
3733 callout is required, relying on the callout cache for the optimization.
3734 The overhead is duplication of the address routing, but this should not be
3737 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3738 to be verified with callout=postmaster, and the main callout worked but the
3739 postmaster check failed, the verification correctly failed. However, if a
3740 subsequent RCPT command asked for sender verification *without* the
3741 postmaster check, incorrect caching caused this verification also to fail,
3744 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3745 it was not caching the DNS options (qualify_single, search_parents) that
3746 were used when the lookup failed. A subsequent lookup with different
3747 options therefore always gave the same answer, though there were cases
3748 where it should not have. (Example: a "domains = !$mx_any" option on a
3749 dnslookup router: the "domains" option is always processed without any
3750 widening, but the router might have qualify_single set.) Now Exim uses the
3751 cached value only when the same options are set.
3753 32. Added John Jetmore's "exipick" utility to the distribution.
3755 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3756 than a timeout (e.g. a certificate is required, and is not provided), an
3757 Exim server now closes the connection immediately. Previously it waited for
3758 the client to close - but if the client is SSL, it seems that they each
3759 wait for each other, leading to a delay before one of them times out.
3761 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3762 maintained 0.8.x compatibility because I don't think many are using it, and
3763 it is clearly obsolete.
3765 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3768 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3769 client certificate was expired. A simple patch fixes this, though I don't
3770 understand the full logic of why the verify callback is called multiple
3773 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3774 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3775 which causes problems with some clients (such as the Certicom SSL Plus
3776 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3777 disables the coutermeasure allowing Eudora to connect."
3779 38. Exim was not checking that a write() to a log file succeeded. This could
3780 lead to Bad Things if a log got too big, in particular if it hit a file
3781 size limit. Exim now panics and dies if it cannot write to a log file, just
3782 as it does if it cannot open a log file.
3784 39. Modified OS/Makefile-Linux so that it now contains
3786 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3788 The two -D definitions ensure that Exim is compiled with large file
3789 support, which makes it possible to handle log files that are bigger than
3792 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3793 instance) a domain was checked against a named list that involved a lookup,
3794 causing $domain_data to be set, then another domain was checked against the
3795 same list, then the first domain was re-checked, the value of $domain_data
3796 after the final check could be wrong. In particular, if the second check
3797 failed, it could be set empty. This bug probably also applied to
3800 41. The strip_trailing_dot option was not being applied to the address given
3801 with the -f command-line option.
3803 42. The code for reading a message's header from the spool was incrementing
3804 $received_count, but never initializing it. This meant that the value was
3805 incorrect (doubled) while delivering a message in the same process in which
3806 it was received. In the most common configuration of Exim, this never
3807 happens - a fresh exec is done - but it can happen when
3808 deliver_drop_privilege is set.
3810 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3811 it now includes up to 150 characters of the unexpected data in the log
3814 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3815 and building data strings. The size of both of these buffers was 10 000
3816 bytes - far larger than anybody would *ever* want, thought I. Needless to
3817 say, somebody hit the limit. I have increased the maximum line length to
3818 20 000 and the maximum data length of concatenated lines to 100 000. I have
3819 also fixed two bugs, because there was no checking on these buffers. Tsk,
3820 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3821 buffer is too small.
3823 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3824 lsearch lookups. Now it does.
3826 46. When parsing a route_list item in a manualroute router, a fixed-length
3827 buffer was used for the list of hosts. I made this 1024 bytes long,
3828 thinking that nobody would ever have a list of hosts that long. Wrong.
3829 Somebody had a whole pile of complicated expansion conditions, and the
3830 string was silently truncated, leading to an expansion error. It turns out
3831 that it is easier to change to an unlimited length (owing to other changes
3832 that have happened since this code was originally written) than to build
3833 structure for giving a limitation error. The length of the item that
3834 expands into the list of hosts is now unlimited.
3836 47. The lsearch lookup could not handle data where the length of text line was
3837 more than 4095 characters. Such lines were truncated, leading to shortened
3838 data being returned. It should now handle lines of any length.
3840 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3841 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3844 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3845 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3846 no longer happens in comment lines. A list of replacements is now placed
3847 at the head of all of the source files, except those whose only change is
3848 to replace PERL_COMMAND in the very first #! line.
3850 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3851 messages on the queue, with a bottom-up merge sort, using code contributed
3852 by Michael Haardt. This should make operations like -bp somewhat faster on
3853 large queues. It won't affect queue runners, except when queue_run_in_order
3856 51. Installed eximstats 1.31 in the distribution.
3858 52. Added support for SRV lookups to the dnslookup router.
3860 53. If an ACL referred to $message_body or $message_body_end, the value was not
3861 reset for any messages that followed in the same SMTP session.
3863 54. The store-handling optimization for building very long strings was not
3864 differentiating between the different store pools. I don't think this
3865 actually made any difference in practice, but I've tidied it.
3867 55. While running the routers to verify a sender address, $sender_address
3868 was still set to the sender address. This is wrong, because when routing to
3869 send a bounce to the sender, it would be empty. Therefore, I have changed
3870 it so that, while verifying a sender address, $sender_address is set to <>.
3871 (There is no change to what happens when verifying a recipient address.)
3873 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3874 target A or AAAA records (if not already returned) without resetting the
3875 qualify_single or search_parents options of the DNS resolver. These are
3876 inappropriate in this case because the targets of MX and SRV records must
3877 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3878 target that, when qualified, matched something in the local domain. These
3879 two options are now turned off when doing these lookups.
3881 57. It seems that at least some releases of Reiserfs (which does not have the
3882 concept of a fixed number of inodes) returns zero and not -1 for the
3883 number of available inodes. This interacted badly with check_spool_inodes,
3884 which assumed that -1 was the "no such thing" setting. What I have done is
3885 to check that the total number of inodes is greater than zero before doing
3886 the test of how many are available.
3888 58. When a "warn" ACL statement has a log_message modifier, the message is
3889 remembered, and not repeated. This is to avoid a lot of repetition when a
3890 message has many recipients that cause the same warning to be written.
3891 Howewer, Exim was preserving the list of already written lines for an
3892 entire SMTP session, which doesn't seem right. The memory is now reset if a
3893 new message is started.
3895 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3896 debugging output unless log_rewrite was also set.
3898 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3899 of (int)(handle) when we know that handle contains (void *)(-1).
3901 61. The Exim daemon panic-logs an error return when it closes the incoming
3902 connection. However "connection reset by peer" seems to be common, and
3903 isn't really an error worthy of noting specially, so that particular error
3906 62. When Exim is trying to find all the local interfaces, it used to panic and
3907 die if the ioctl to get the interface flags failed. However, it seems that
3908 on at least one OS (Solaris 9) it is possible to have an interface that is
3909 included in the list of interfaces, but for which you get a failure error
3910 for this call. This happens when the interface is not "plumbed" into a
3911 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3912 failure of the "get flags" call assumes that the interface is down.
3914 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3915 makes life easier for people who are doing arithmetic on fields extracted
3916 from dates, where you often get leading zeros that should not be
3917 interpreted as octal.
3919 64. Added qualify_domain to the redirect router, to override the global
3922 65. If a pathologically long header line contained very many addresses (the
3923 report of this problem mentioned 10 000) and each of them was rewritten,
3924 Exim could use up a very large amount of memory. (It kept on making new
3925 copies of the header line as it rewrote, and never released the old ones.)
3926 At the expense of a bit more processing, the header rewriting function has
3927 been changed so that it no longer eats memory in this way.
3929 66. The generation of the Received: header has been moved from the time that a
3930 message starts to be received, to the time that it finishes. The timestamp
3931 in the Received: header should now be very close to that of the <= log
3932 line. There are two side-effects of this change:
3934 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3935 logged header lines no longer include the local Received: line, because
3936 it has not yet been created. The same applies to a copy of the message
3937 that is returned to a non-SMTP sender when a message is rejected.
3939 (b) When a filter file is tested using -bf, no additional Received: header
3940 is added to the test message. After some thought, I decided that this
3943 This change does not affect the value of $received_for. It is still set
3944 after address rewriting, but before local_scan() is called.
3946 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3948 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3949 gave an unhelpful panic error message, and a defer error. I have managed to
3950 change this behaviour so that it now rejects any supplied certificate,
3951 which seems right, as the list of acceptable certificates is empty.
3953 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3954 gave an unhelpful defer error. I have not managed to make this reject any
3955 supplied certificates, but the error message it gives is "no certificate
3956 supplied", which is not helpful.
3958 70. exigrep's output now also includes lines that are not associated with any
3959 message, but which match the given pattern. Implemented by a patch from
3960 Martin Sluka, which also tidied up the Perl a bit.
3962 71. Recipient callout verification, like sender verification, was using <> in
3963 the MAIL FROM command. This isn't really the right thing, since the actual
3964 sender may affect whether the remote host accepts the recipient or not. I
3965 have changed it to use the actual sender in the callout; this means that
3966 the cache record is now keyed on a recipient/sender pair, not just the
3967 recipient address. There doesn't seem to be a real danger of callout loops,
3968 since a callout by the remote host to check the sender would use <>.
3969 [SEE ABOVE: changed after hitting problems.]
3971 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3972 temporary errors. However, in the case of such a code being given after
3973 the end of a data transmission (i.e. after ".") Exim was failing to write
3974 a retry record for the message. (Yes, there was some broken host that was
3975 actually sending 8xx at this point.)
3977 73. An unknown lookup type in a host list could cause Exim to panic-die when
3978 the list was checked. (An example that provoked this was putting <; in the
3979 middle of a list instead of at the start.) If this happened during a DATA
3980 ACL check, a -D file could be left lying around. This kind of configuration
3981 error no longer causes Exim to die; instead it causes a defer errror. The
3982 incident is still logged to the main and panic logs.
3984 74. Buglet left over from Exim 3 conversion. The message "too many messages
3985 in one connection" was written to the rejectlog but not the mainlog, except
3986 when address rewriting (yes!) was being logged.
3988 75. Added write_rejectlog option.
3990 76. When a system filter was run not as root (that is, when system_filter_user
3991 was set), the values of the $n variables were not being returned to the
3992 main process; thus, they were not subsequently available in the $sn
3995 77. Added +return_path_on_delivery log selector.
3997 78. A connection timeout was being treated differently from recipients deferred
3998 when testing hosts_max_try with a message that was older than the host's
3999 retry timeout. (The host should not be counted, thus allowing all hosts to
4000 be tried at least once before bouncing.) This may have been the cause of an
4001 occasionally reported bug whereby a message would remain on the queue
4002 longer than the retry timeout, but would be bounced if a delivery was
4003 forced. I say "may" because I never totally pinned down the problem;
4004 setting up timeout/retry tests is difficult. See also the next item.
4006 79. The ultimate address timeout was not being applied to errors that involved
4007 a combination of host plus message (for example, a timeout on a MAIL
4008 command). When an address resolved to a number of possible hosts, and they
4009 were not all tried for each delivery (e.g. because of hosts_max_try), a
4010 message could remain on the queue longer than the retry timeout.
4012 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4015 81. Fixed an obscure SMTP outgoing bug which required at least the following
4016 conditions: (a) there was another message waiting for the same server;
4017 (b) the server returned 5xx to all RCPT commands in the first message so
4018 that the message was not completed; (c) the server dropped the connection
4019 or gave a negative response to the RSET that Exim sends to abort the
4020 transaction. The observed case was a dropped connection after DATA that had
4021 been sent in pipelining mode. That is, the server had advertised PIPELINING
4022 but was not implementing it correctly. The effect of the bug was incorrect
4023 behaviour, such as trying another host, and this could lead to a crash.
4029 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4030 and daemon.c were passed as pointers to ints; they should have been
4031 pointers to socklen_t variables (which are typically unsigned ints).
4033 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4036 3. Fixed a really odd bug that affected only the testing scheme; patching a
4037 certain fixed string in the binary changed the value of another string that
4038 happened to be identical to the end of the original first string.
4040 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4041 name", it returns that address as the IP address. On some operating
4042 systems (e.g. Solaris), it also passes back the IP address string as the
4043 "host name". However, on others (e.g. Linux), it passes back an empty
4044 string. Exim wasn't checking for this, and was changing the host name to an
4045 empty string, assuming it had been canonicized.
4047 5. Although rare, it is permitted to have more than one PTR record for a given
4048 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4049 all the names associated with an address, because they do in Solaris.
4050 However, it seems that they do not in Linux for data that comes from the
4051 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4052 I found this out when I moved to a new Linux workstation and tried to run
4053 the Exim test suite.
4055 To get round this problem I have changed the code so that it now does its
4056 own call to the DNS to look up PTR records when searching for a host name.
4057 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4058 addresses that are only in /etc/hosts are still found.
4060 This behaviour is, however, controlled by an option called host_lookup_
4061 order, which defaults to "bydns:byaddr". If people want to use the other
4062 order, or indeed, just use one or the other means of lookup, they can
4063 specify it in this variable.
4065 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4066 some operating systems, this comes back from gethostbyaddr() as an empty
4067 string, and this is what Exim used to test for. However, it seems that in
4068 other systems, "." is yielded. Exim now tests for this case too.
4070 7. The values of check_spool_space and check_log_space are now held internally
4071 as a number of kilobytes instead of an absolute number of bytes. If a
4072 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4073 kilobyte. This means that much larger values can be stored.
4075 8. Exim monitor: an attempt to get the action menu when not actually pointing
4076 at a message produces an empty menu entitled "No message selected". This
4077 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4078 no entries in it ("Shell widget menu has zero width and/or height"). So I
4079 have added a single, blank menu entry in this case.
4081 9. Added ${quote_local_part.
4083 10. MIME decoding is now applied to the contents of Subject: header lines when
4086 11. Now that a reference to $sender_host_address automatically causes a reverse
4087 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4088 host lookup before query-style lookups in lists that might use this
4089 variable. This has therefore been abolished, and the "net-" prefix is no
4090 longer necessary for query-style lookups.
4092 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4093 have been a typo for LFLAGS, so it has been changed.
4095 13. The install script calls Exim with "-C /dev/null" in order to find the
4096 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4097 to be output. Howeve, since Exim outputs its version number before the
4098 error, it didn't break the script. It just looked ugly. I fixed this by
4099 always allowing "-C /dev/null" if the caller is root.
4101 14. Ignore overlarge ACL variable number when reading spool file - insurance
4102 against a later release with more variables having written the file.
4104 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4105 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4106 this, as well as the form without the "IPv6" on the front.
4108 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4109 OS/Makefile-Darwin file.
4111 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4112 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4114 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4115 variable. However, some OS use ints for inodes. I've added cast to long int
4116 to get rid of the compiler warning.
4118 19. I had forgotten to lock out "/../" in configuration file names when
4119 ALT_CONFIG_PREFIX was set.
4121 20. Routers used for verification do not need to specify transports. However,
4122 if such a router generated a host list, and callout was configured, Exim
4123 crashed, because it could not find a port number from the (non-existent)
4124 transport. It now assumes port 25 in this circumstance.
4126 21. Added the -t option to exigrep.
4128 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4129 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4130 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4131 them is set, LOOKUP_LSEARCH is forced.
4133 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4135 24. Added sender and host information to the "rejected by local_scan()" log
4136 line; previously there was no indication of these.
4138 25. Added .include_if_exists.
4140 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4141 Linux. It turns out that not all file systems support this. Apparently some
4142 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4143 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4144 not supported on the file descriptor, is now ignored when Exim is trying to
4145 sync a directory. This applies only to Linux.
4147 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4149 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4150 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4153 29. Added support for maildirsize files from supplied patch (modified a bit).
4155 30. The use of :fail: followed by an empty string could lead Exim to respond to
4156 sender verification failures with (e.g.):
4158 550 Verification failed for <xxx>
4159 550 Sender verify failed
4161 where the first response line was missing the '-' that indicates it is not
4162 the final line of the response.
4164 31. The loop for finding the name of the user that called Exim had a hardwired
4165 limit of 10; it now uses the value of finduser_retries, which is used for
4166 all other user lookups.
4168 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4171 33. Exim was neglecting to zero errno before one call of strtol() when
4172 expanding a string and expecting an integer value. On some systems this
4173 resulted in spurious "integer overflow" errors. Also, it was casting the
4174 result into an int without checking.
4176 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4177 did not work. The code looks as if it has *never* worked, though it appears
4178 to have been documented since at least releast 1.62. I have made it work.
4180 35. The "timeout_DNS" error in retry rules, also documented since at least
4181 1.62, also never worked. As it isn't clear exactly what this means, and
4182 clearly it isn't a major issue, I have abolished the feature by treating it
4183 as "timeout", and writing a warning to the main and panic logs.
4185 36. The display of retry rules for -brt wasn't always showing the error code
4188 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4189 timeout_connect_A, timeout_connect_MX.
4191 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4192 to the empty sender.
4194 39. The daemon was not analysing the content of -oX till after it had closed
4195 stderr and disconnected from the controlling terminal. This meant that any
4196 syntax errors were only noted on the panic log, and the return code from
4197 the command was 0. By re-arranging the code a little, I've made the
4198 decoding happen first, so such errors now appear on stderr, and the return
4199 code is 1. However, the actual setting up of the sockets still happens in
4200 the disconnected process, so errors there are still only recorded on the
4203 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4204 connections (as happens on some IP stacks) was logged at start up time as
4205 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4206 it from "IPv6 and IPv4", which means that two separate sockets are being
4209 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4210 says whether AF_INET or AF_INET6 was passed as an argument.
4212 42. Exiwhat output was messed up when time zones were included in log
4215 43. Exiwhat now gives more information about the daemon's listening ports,
4216 and whether -tls-on-connect was used.
4218 44. The "port" option of the smtp transport is now expanded.
4220 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4221 silently ignored. Now an error message is written to the main and panic
4224 46. There's a new ACL modifier called "logwrite" which writes to a log file
4225 as soon as it is encountered.
4227 47. Added $local_user_uid and $local_user_gid at routing time.
4229 48. Exim crashed when trying to verify a sender address that was being
4232 49. Exim was recognizing only a space character after ".include". It now also
4233 recognizes a tab character.
4235 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4236 extracting the relevant information from the specification. The man page no
4237 longer contains scrambled data for the -d option, and I've added a section
4238 at the front about calling Exim under different names.
4240 51. Added "extra_headers" argument to the "mail" command in filter files.
4242 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4245 53. Installed eximstats 1.29.
4247 54. Added transport_filter_timeout as a generic transport option.
4249 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4250 Cc: header lines. This was required by RFC 822, but it not required by RFC
4253 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4254 incoming messages that did not have them. Now it does so only if the
4255 message originates locally, that is, if there is no associated remote host
4256 address. When Resent- header lines are present, this applies to the Resent-
4257 lines rather than the non-Resent- lines.
4259 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4260 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4262 58. Messages for configuration errors now include the name of the main
4263 configuration file - useful now that there may be more than one file in a
4264 list (.included file names were always shown).
4266 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4267 for those rare installations that do not start the daemon as root or run it
4268 setuid root. I've cut out the call to initgroups() if the daemon is not
4271 60. The Exim user and group can now be bound into the binary as text strings
4272 that are looked up at the start of Exim's processing.
4274 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4276 62. Added $mailstore_basename variable.
4278 63. Installed patch to sieve.c from Michael Haardt.
4280 64. When Exim failed to open the panic log after failing to open the main log,
4281 the original message it was trying to log was written to stderr and debug
4282 output, but if they were not available (the usual case in production), it
4283 was lost. Now it is written to syslog before the two lines that record the
4284 failures to open the logs.
4286 65. Users' Exim filters run in subprocesses under the user's uid. It is
4287 possible for a "deliver" command or an alias in a "personal" command to
4288 provoke an address rewrite. If logging of address rewriting is configured,
4289 this fails because the process is not running as root or exim. There may be
4290 a better way of dealing with this, but for the moment (because 4.30 needs
4291 to be released), I have disabled address rewrite logging when running a
4292 filter in a non-root, non-exim process.
4298 1. The buildconfig auxiliary program wasn't quoting the value set for
4299 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4300 not defined. This bug was masked in 4.22 by the effect that was fixed in
4303 2. Some messages that were rejected after a message id was allocated were
4304 shown as "incomplete" by exigrep. It no longer does this for messages that
4305 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4307 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4308 have allow_domain_literals set, the ID did not get logged in the <= line.
4309 Domain literals are now always recognized in Message-ID: header lines.
4311 4. The first argument for a ${extract expansion item is the key name or field
4312 number. Leading and trailing spaces in this item were not being ignored,
4313 causing some misleading effects.
4315 5. When deliver_drop_privilege was set, single queue runner processes started
4316 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4317 same command in the process it spins off) were not dropping privilege.
4319 6. When the daemon running as "exim" started a queue runner, it always
4320 re-executed Exim in the spun-off process. This is a waste of effort when
4321 deliver_drop_privilege is set. The new process now just calls the
4322 queue-runner function directly.
4328 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4331 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4332 ignored. Though the use of -oP was forcing the writing of a pid file, it
4333 was always written to the default place.
4335 3. If the message "no IP address found for host xxxx" is generated during
4336 incoming verification, it is now followed by identification of the incoming
4337 connection (so you can more easily find what provoked it).
4339 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4341 5. Added some features to "harden" Exim a bit more against certain attacks:
4343 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4344 be put in Local/Makefile. This is like the never_users runtime option,
4345 but it cannot be overridden. The default setting is "root".
4347 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4348 prefix string with which any file named in a -C command line option
4351 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4352 is retained for -C and -D only if the caller of Exim is root. Without
4353 it, the exim user may also use -C and -D and retain privilege.
4355 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4356 command line option is disabled.
4358 6. Macro names set by the -D option must start with an upper case letter, just
4359 like macro names defined in the configuration file.
4361 7. Added "dereference=" facility to LDAP.
4363 8. Two instances of the typo "uknown" in the source files are fixed.
4365 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4366 the Configure-Makefile script screwed up while processing it.
4368 10. Incorporated PCRE 4.4.
4370 11. The SMTP synchronization check was not operating right at the start of an
4371 SMTP session. For example, it could not catch a HELO sent before the client
4372 waited for the greeting. There is now a check for outstanding input at the
4373 point when the greeting is written. Because of the duplex, asynchronous
4374 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4375 way, but not yet received, when the check is performed.
4377 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4378 on TCP/IP sockets, because this apparently causes some broken clients to
4381 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4382 unchanged) from the Cygwin maintainer.
4384 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4385 was supported instead of testing for mbx. Effectively a typo.
4387 15. The spa authenticator server code was not checking that the input it
4388 received was valid base64.
4390 16. The debug output line for the "set" modifier in ACLs was not showing the
4391 name of the variable that was being set.
4393 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4395 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4396 lookup on its own. Something else was needed to trigger the lookup. For
4397 example, a match in host_lookup or the need for a host name in a host list.
4398 Now, if $sender_host_name is referenced and the host name has not yet been
4399 looked up, a lookup is performed. If the lookup fails, the variable remains
4400 empty, and $host_lookup_failed is set to "1".
4402 19. Added "eqi" as a case-independent comparison operator.
4404 20. The saslauthd authentication condition could segfault if neither service
4405 nor realm was specified.
4407 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4408 error message that was logged was misleading, and incoming SMTP
4409 connections were dropped. The message is now more accurate, and temporary
4410 errors are given to SMTP connections.
4412 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4413 (except RSET) with a 421 error, until QUIT is received. However, it was
4414 failing to send a response to QUIT.
4416 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4417 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4418 ACL is now run after the helo_try_verify_hosts code.
4420 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4421 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4422 case-independent, so other case variants are also recognized). Apparently
4423 some systems use these upper case variants.
4425 25. If more than two messages were waiting for the same host, and a transport
4426 filter was specified for the transport, Exim sent two messages over the
4427 same TCP/IP connection, and then failed with "socket operation on non-
4428 socket" when it tried to send the third.
4430 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4432 27. The extern definition of crypt16() in expand.c was not being excluded when
4433 the OS had its own crypt16() function.
4435 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4436 as a preferred synonym for return_size_limit, both as an option and as an
4439 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4441 30. Changed the default configuration ACL to relax the local part checking rule
4442 for addresses that are not in any local domains. For these addresses,
4443 slashes and pipe symbols are allowed within local parts, but the sequence
4444 /../ is explicitly forbidden.
4446 31. SPA server authentication was not clearing the challenge buffer before
4449 32. log_message in a "warn" ACL statement was writing to the reject log as
4450 well as to the main log, which contradicts the documentation and doesn't
4451 seem right (because no rejection is happening). So I have stopped it.
4453 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4454 However, I am unable to do any testing of this.
4456 34. Fixed an infelicity in the appendfile transport. When checking directories
4457 for a mailbox, to see if any needed to be created, it was accidentally
4458 using path names with one or more superfluous leading slashes; tracing
4459 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4461 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4462 contents are added to the log line that is written for every discarded
4463 recipient. (Previously a log_message setting was ignored.)
4465 36. The ${quote: operator now quotes the string if it is empty.
4467 37. The install script runs exim in order to find its version number. If for
4468 some reason other than non-existence or emptiness, which it checks, it
4469 could not run './exim', it was installing it with an empty version number,
4470 i.e. as "exim-". This error state is now caught, and the installation is
4473 38. An argument was missing from the function that creates an error message
4474 when Exim fails to connect to the socket for saslauthd authentication.
4475 This could cause Exim to crash, or give a corrupted message.
4477 39. Added isip, isip4, and isip6 to ${if conditions.
4479 40. The ACL variables $acl_xx are now saved with the message, and can be
4480 accessed later in routers, transports, and filters.
4482 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4483 strings in the file are not string-expanded.
4485 42. If a MAIL command specified a SIZE value that was too large to fit into an
4486 int variable, the check against message_size_limit failed. Such values are
4487 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4488 day this will have to be increased, but I don't think I want to be around
4489 when emails are that large.
4496 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4497 iconv() is not standard in FreeBSD.
4499 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4500 IPv6 enabled. The observed symptom was a segmentation fault on return from
4501 the function os_common_find_running_interfaces() in src/os.c.
4503 3. In the check_special_case() function in daemon.c I had used "errno" as an
4504 argument name, which causes warnings on some systems. This was basically a
4505 typo, since it was named "eno" in the comments!
4507 4. The code that waits for the clock to tick (at a resolution of some fraction
4508 of a second) so as to ensure message-id uniqueness was always waiting for
4509 at least one whole tick, when it could have waited for less. [This is
4510 almost certainly not relevant at current processor speeds, where it is
4511 unlikely to ever wait at all. But we try to future-proof.]
4513 5. The function that sleeps for a time interval that includes fractions of a
4514 second contained a race. It did not block SIGALRM between setting the
4515 timer, and suspending (a couple of lines later). If the interval was short
4516 and the sigsuspend() was delayed until after it had expired, the suspension
4517 never ended. On busy systems this could lead to processes getting stuck for
4520 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4521 process, before it forks any delivery processes. The open lookup caching
4522 mechanism meant that the open file or database connection was passed into
4523 the delivery process. The problem was that delivery processes always tidy
4524 up cached lookup data. This could cause a problem for the next delivery
4525 process started by the queue runner, because the external queue runner
4526 process does not know about the closure. So the next delivery process
4527 still has data in the lookup cache. In the case of a file lookup, there was
4528 no problem because closing a file descriptor in a subprocess doesn't affect
4529 the parent. However, if the lookup was caching a connection to a database,
4530 the connection was closed, and the second delivery process was likely to
4531 see errors such as "PGSQL: query failed: server closed the connection
4532 unexpectedly". The problem has been fixed by closing all cached lookups
4533 in a queue runner before running a delivery process.
4535 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4536 seem to have the "const" qualifier which it has on other OS. I've
4539 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4540 *of the same type* (client or server) with the same public name that an
4541 error should be diagnosed.
4543 9. When Exim looked up a host name for an IP address, but failed to find the
4544 original IP address when looking up the host name (a safety check), it
4545 output the message "<ip address> does not match any IP for NULL", which was
4546 confusing, to say the least. The bug was that the host name should have
4547 appeared instead of "NULL".
4549 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4550 user that is built into the binary, and the -C or -D options is used, root
4551 privilege is dropped before the configuration file is read. In addition,
4552 logging is switched to stderr instead of the normal log files. If the
4553 configuration then re-defines the Exim user, the unprivileged environment
4554 is probably not what is expected, so Exim logs a panic warning message (but
4557 However, if deliver_drop_privilege is set, the unprivileged state may well
4558 be exactly what is intended, so the warning has been cut out in that case,
4559 and Exim is allowed to try to write to its normal log files.
4565 1. smtp_return_error_details was not giving details for temporary sender
4566 or receiver verification errors.
4568 2. Diagnose a configuration error if two authenticators have the same public
4571 3. Exim used not to create the message log file for a message until the first
4572 delivery attempt. This could be confusing when incoming messages were held
4573 for policy or load reasons. The message log file is now created at the time
4574 the message is received, and an initial "Received" line is written to it.
4576 4. The automatically generated man page for command line options had a minor
4577 bug that caused no ill effects; however, a more serious problem was that
4578 the procedure for building the man page automatically didn't always
4579 operate. Consequently, release 4.20 contains an out-of-date version. This
4580 shouldn't happen again.
4582 5. When building Exim with embedded Perl support, the script that builds the
4583 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4584 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4586 6. The freeze_tell option was not being used for messages that were frozen on
4587 arrival, either by an ACL or by local_scan().
4589 7. Added the smtp_incomplete_transaction log selector.
4591 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4592 was accepting AUTH without a new EHLO.
4594 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4595 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4598 10. Logging of TCP/IP connections (when configured) now happens in the main
4599 daemon process instead of the child process, so that the TCP/IP connection
4600 count is more accurate (but it can never be perfect).
4602 11. The use of "drop" in a nested ACL was not being handled correctly in the
4603 outer ACL. Now, if condition failure induced by the nested "drop" causes
4604 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4605 or "require"), the connection is dropped.
4607 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4608 that yield "discard" can now be used with an "accept" or a "discard" verb,
4609 but an error is generated for any others (because I can't see a useful way
4610 to define what should happen).
4612 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4613 are now processed in the same way as lines in the Exim configuration file.
4614 In particular, continuation lines are supported.
4616 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4618 15. Added -ti meaning -t -i.
4620 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4621 domains, and warn by logging if others are found.
4623 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4624 returned from GIFCONF when Exim is trying to find the list of interfaces on
4625 a host. The code in os.c has been modified to copy each ifreq to an aligned
4626 structure in all cases.
4628 Also, in some cases, the returned ifreq's were being copied to a 'struct
4629 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4630 means the last couple of bytes of an IPv6 address could be chopped if the
4631 ifreq contained only a normal sockaddr (14 bytes storage).
4633 18. Named domain lists were not supported in the hosts_treat_as_local option.
4634 An entry such as +xxxx was not recognized, and was treated as a literal
4637 19. Ensure that header lines added by a DATA ACL are included in the reject log
4638 if the ACL subsequently rejects the message.
4640 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4641 MD5 (which is deprecated).
4643 21. When testing a filter file using -bf, Exim was writing a message when it
4644 took the sender from a "From " line in the message, but it was not doing so
4645 when it took $return_path from a Return-Path: header line. It now does.
4647 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4648 with a valid header line field name (a series of printing characters
4649 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4651 23. Changed "disc" in the source to "disk" to conform to the documentation and
4652 the book and for uniformity.
4654 24. Ignore Sendmail's -Ooption=value command line item.
4656 25. When execve() failed while trying to run a command in a pipe transport,
4657 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4658 could be confused with a return value of 69 from the command itself. This
4659 has been changed to 127, the value the shell returns if it is asked to run
4660 a non-existent command. The wording for the related log line suggests a
4661 non-existent command as the problem.
4663 26. If received_header_text expands to an empty string, do not add a Received:
4664 header line to the message. (Well, it adds a token one on the spool, but
4665 marks it "old" so that it doesn't get used or transmitted.)
4667 27. Installed eximstats 1.28 (addition of -nt option).
4669 28. There was no check for failure on the call to getsockname() in the daemon
4670 code. This can fail if there is a shortage of resources on the system, with
4671 ENOMEM, for example. A temporary error is now given on failure.
4673 29. Contrary to the C standard, it seems that in some environments, the
4674 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4675 program. Exim now does this explicitly; it affects the formatting of
4676 timestamps using strftime().
4678 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4679 complaints. I've now initialized all the variables, to avoid this.
4681 32. Header lines added by a system filter were not being "seen" during
4682 transport-time rewrites.
4684 33. The info_callback() function passed to OpenSSL is set up with type void
4685 (*)(SSL *, int, int), as described somewhere. However, when calling the
4686 function (actually a macro) that sets it up, the type void(*)() is
4687 expected. I've put in a cast to prevent warnings from picky compilers.
4689 34. If a DNS black list lookup found a CNAME record, but there were no A
4690 records associated with the domain it pointed at, Exim crashed.
4692 35. If a DNS black list lookup returned more than one A record, Exim ignored
4693 all but the first. It now scans all returned addresses if a particular IP
4694 value is being sought. In this situation, the contents of the
4695 $dnslist_value variable are a list of all the addresses, separated by a
4698 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4699 to accept a host name and all its aliases if the forward lookup for any of
4700 them yielded the IP address of the incoming connection. Now it accepts only
4701 those names whose forward lookup yields the correct IP address. Any other
4702 names are discarded. This closes a loophole whereby a rogue DNS
4703 administrator could create reverse DNS records to break through a
4704 wildcarded host restriction in an ACL.
4706 37. If a user filter or a system filter that ran in a subprocess used any of
4707 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4708 the wrong values were passed to the pipe command ($thisaddress had the
4709 value of $0, $0 had the value of $1, etc). This bug was introduced by
4710 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4712 38. Improved the line breaking for long SMTP error messages from ACLs.
4713 Previously, if there was no break point between 40 and 75 characters, Exim
4714 left the rest of the message alone. Two changes have been made: (a) I've
4715 reduced the minimum length to 35 characters; (b) if it can't find a break
4716 point between 35 and 75 characters, it looks ahead and uses the first one
4717 that it finds. This may give the occasional overlong line, but at least the
4718 remaining text gets split now.
4720 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4721 file descriptors might be low, and that setting 1000 would always raise it.
4722 It turns out that in some environments, the limit is already over 1000 and
4723 that lowering it causes trouble. So now Exim takes care not to decrease it.
4725 40. When delivering a message, the value of $return_path is set to $sender_
4726 address at the start of routing (routers may change the value). By an
4727 oversight, this default was not being set up when an address was tested by
4728 -bt or -bv, which affected the outcome if any router or filter referred to
4731 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4732 log only when "message" or "log_message" are set. However, if one of the
4733 conditions was an address verification, or a call to a nested ACL, the
4734 messages generated by the underlying test were being passed through. This
4735 no longer happens. The underlying message is available in $acl_verify_
4736 message for both "message" and "log_message" expansions, so it can be
4737 passed through if needed.
4739 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4740 new expansion $bh_ to give the encoded byte string without charset
4741 translation. Translation happens only if iconv() is available; HAVE_ICONV
4742 indicates this at build time. HEADERS_CHARSET gives the charset to
4743 translate to; headers_charset can change it in the configuration, and
4744 "headers charset" can change it in an individual filter file.
4746 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4747 that creates RFC 2047 encoded "words" labels them as that charset instead
4748 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4749 expansion operator; (ii) when Exim creates a From: line for a local
4750 message; (iii) when a header line is rewritten to include a "phrase" part.
4752 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4753 buggy, causing it to skip the first lines of messages whose message ID
4754 ended in 'D'. This would not have bitten before Exim release 4.14, because
4755 message IDs were unlikely to end in 'D' before then. The effect was to have
4756 incorrect size information for certain domains.
4758 45. #include "config.h" was missing at the start of the crypt16.c module. This
4759 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4762 46. If there was a timeout during a "random" callout check, Exim treated it as
4763 a failure of the random address, and carried on sending RSET and the real
4764 address. If the delay was just some slowness somewhere, the response to the
4765 original RCPT would be taken as a response to RSET and so on, causing
4766 mayhem of various kinds.
4768 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4769 when I implemented it. It didn't allow for the fact that some option values
4770 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4771 the right test for positive values.
4773 48. Domain names in DNS records are case-independent. Exim always looks them up
4774 in lower case. Some resolvers return domain names in exactly the case they
4775 appear in the zone file, that is, they may contain uppercase letters. Not
4776 all resolvers do this - some return always lower case. Exim was treating a
4777 change of case by a resolver as a change of domain, similar to a widening
4778 of a domain abbreviation. This triggered its re-routing code and so it was
4779 trying to route what was effectively the same domain again. This normally
4780 caused routing to fail (because the router wouldn't handle the domain
4781 twice). Now Exim checks for this case specially, and just changes the
4782 casing of the domain that it ultimately uses when it transmits the message
4785 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4788 50. If a filter generated a file delivery with a non-absolute name (possible if
4789 no home directory exists for the router), the forbid_file option was not
4792 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4793 the existing equality matching.
4795 52. Exim was using ints instead of ino_t variables in some places where it was
4796 dealing with inode numbers.
4798 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4799 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4800 called TMPDIR, and if it finds it is different, it changes its value.
4802 54. The smtp_printf() function is now made available to local_scan() so
4803 additional output lines can be written before returning. There is also an
4804 smtp_fflush() function to enable the detection of a dropped connection.
4805 The variables smtp_input and smtp_batched_input are exported to
4808 55. Changed the default runtime configuration: the message "Unknown user"
4809 has been removed from the ACL, and instead placed on the localuser router,
4810 using the cannot_route_message feature. This means that any verification
4811 failures that generate their own messages won't get overridden. Similarly,
4812 the "Unrouteable address" message that was in the ACL for unverifiable
4813 relay addresses has also been removed.
4815 56. Added hosts_avoid_esmtp to the smtp transport.
4817 57. The exicyclog script was not checking for the esoteric option
4818 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4819 will work only if exicyclog is run under the appropriate euid.
4821 58. Following a discussion on the list, the rules by which Exim recognises line
4822 endings on incoming messages have been changed. The -dropcr and drop_cr
4823 options are now no-ops, retained only for backwards compatibility. The
4824 following line terminators are recognized: LF CRLF CR. However, special
4825 processing applies to CR:
4827 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4828 nor a local message in the state where . is a terminator.
4830 (ii) If a bare CR is encountered in a header line, an extra space is added
4831 after the line terminator so as not to end the header. The reasoning
4832 behind this is that bare CRs in header lines are most likely either
4833 to be mistakes, or people trying to play silly games.
4835 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4836 was being incorrectly given as 18 bytes larger than it should have been.
4837 This is a VOB (very old bug).
4839 60. This may never have affected anything current, but just in case it has:
4840 When the local host is found other than at the start of a list of hosts,
4841 the local host, those with the same MX, and any that follow, are discarded.
4842 When the list in question was part of a longer list of hosts, the following
4843 hosts (not currently being processed) were also being discarded. This no
4844 longer happens. I'm not sure if this situation could ever has previously
4847 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4850 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4851 header. This is something that is recommended in a new Internet Draft, and
4852 is something that is documented as being done by Sendmail. There are two
4853 possible values. For messages generated by the autoreply transport, Exim
4856 Auto-Submitted: auto-replied
4858 whereas for all other generated messages (e.g. bounces) it adds
4860 Auto-Submitted: auto-generated
4862 63. The "personal" condition in filters now includes a test for the
4863 Auto-Submitted: header. If it contains the string "auto-" the message it
4864 not considered personal.
4866 64. Added rcpt_include_affixes as a generic transport option.
4868 65. Added queue_only_override (default true).
4870 66. Added the syslog_duplication option.
4872 67. If what should have been the first header line of a message consisted of
4873 a space followed by a colon, Exim was mis-interpreting it as a header line.
4874 It isn't of course - it is syntactically invalid and should therefore be
4875 treated as the start of the message body. The misbehaviour could have
4876 caused a number of strange effects, including loss of data in subsequent
4877 header lines, and spool format errors.
4879 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4880 client host had authenticated. This control can now be exercised by an ACL
4881 for more flexibility.
4883 69. By default, callouts do not happen when testing with -bh. There is now a
4884 variant, -bhc, which does actually run the callout code, including
4885 consulting and updating the callout cache.
4887 70. Added support for saslauthd authentication, courtesy of Alexander
4890 71. If statvfs() failed on the spool or log directories while checking their
4891 size for availability, Exim confusingly gave the error "space shortage".
4892 Furthermore, in debugging mode it crashed with a floating point exception.
4893 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4894 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4895 serious problem, Exim now writes to the main and panic logs when this
4896 happens, with details of the failure. It then refuses to accept the
4897 incoming message, giving the message "spool directory problem" or "log
4898 directory problem" with a 421 code for SMTP messages.
4900 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4901 0, 1, and 2 exist, because some OS complain for execs without them (see
4902 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4903 descriptors. However, the code omitted to check that the open succeeded,
4904 causing mysterious errors if for some reason the permissions on /dev/null
4905 got screwed. Now Exim writes a message to the main and panic logs, and
4906 bombs out if it can't open /dev/null.
4908 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4909 interact so that it is all more flexible. It is supposed to remain
4910 backwards compatible. Also added extra_local_interfaces.
4912 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4913 to bomb out with an assertion failure - to the client this appears as a
4914 connection drop. This problem occurs in the part of the code that was taken
4915 from the Samba project. Fortunately, the assertion is in a very simple
4916 function, so I have fixed this by reproducing the function inline in the
4917 one place where it is called, and arranging for authentication to fail
4918 instead of killing the process with assert().
4920 75. The SPA client code was not working when the server requested OEM rather
4921 than Unicode encoding.
4923 76. Added code to make require_files with a specific uid setting more usable in
4924 the case where statting the file as root fails - usually a non-root-mounted
4925 NFS file system. When this happens and the failure is EACCES, Exim now
4926 forks a subprocess and does the per-uid checking as the relevant uid.
4928 77. Added process_log_path.
4930 78. If log_file_path was not explicitly set, a setting of check_log_space or
4931 check_log_inodes was ignored.
4933 79. If a space check for the spool or log partitions fails, the incident is now
4934 logged. Of course, in the latter case the data may get lost...
4936 80. Added the %p formatting code to string_format() so that it can be used to
4937 print addresses in debug_print(). Adjusted all the address printing in the
4938 debugging in store.c to use %p rather than %d.
4940 81. There was a concern that a line of code in smtp_in.c could overflow a
4941 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4942 initially expressed, the concern was not well-founded, because trailing
4943 spaces are removed early. However, if the trailing spaces were followed by
4944 a NULL, they did not get removed, so the overflow was possible. Two fixes
4947 (a) I re-wrote the offending code in a cleaner fashion.
4948 (b) If an incoming SMTP command contains a NULL character, it is rejected
4951 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4952 runs initgroups(), so that if the Exim user is in any additional groups,
4953 they will be used during message reception.
4959 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob