1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.581 2009/10/26 13:14:23 nm4 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TK/01 Added patch by Johannes Berg that expands the main option
10 "spamd_servers" if it starts with a dollar sign.
12 TK/02 Write list of recipients to X-Envelope-Sender header when building
13 the mbox-format spool file for content scanning (suggested by Jakob
16 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
17 (http://www.dcc-servers.net/) support via dccifd. Activated by
18 setting EXPERIMENTAL_DCC=yes in Local/Makefile. Check out
19 experimental_spec.txt for more documentation.
21 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
22 by Mark Daniel Reidel <mr@df.eu>.
24 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
25 When building exim an external PCRE library is now needed -
26 PCRE is a system library on the majority of modern systems.
27 See entry on PCRE_LIBS in EDITME file.
29 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
30 conversation. Added nologin parameter to request.
31 Patch contributed by Kirill Miazine
33 TF/01 Do not log submission mode rewrites if they do not change the address.
35 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
37 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
38 log files in place. Contributed by Roberto Lima
40 NM/04 Bugzilla 667: close socket used by dovecot authenticator
42 TF/03 Bugzilla 615: When checking the local_parts router precondition
43 after a local_part_suffix or local_part_prefix option, Exim now
44 does not use the address's named list lookup cache, since this
45 contains cached lookups for the whole local part.
47 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
48 Robert Millan. Documentation is in experimental-spec.txt
50 TF/04 Bugzilla 668: Fix parallel build (make -j).
52 NM/05 Bugzilla 437: Prevent Maildix aux files being created with mode 000
54 NM/05 Bugzilla 598: Improvement to Dovecot authenticator handling.
55 Patch provided by Jan Srzednicki
57 TF/05 Leading white space used to be stripped from $spam_report which
58 wrecked the formatting. Now it is preserved.
60 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
61 that they are available at delivery time.
63 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
65 TF/08 TLS error reporting now respects the incoming_interface and
66 incoming_port log selectors.
68 TF/09 Produce a more useful error message if an SMTP transport's hosts
69 setting expands to an empty string.
71 NM/06 Bugzilla 744: EXPN did not work under TLS.
72 Patch provided by Phil Pennock
74 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
75 Patch provided by Richard Godbee
77 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
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 622: 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).
99 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
100 Patch provided by Graeme Fowler
102 NM/16 Bugzilla 851: Documentation example syntax fix.
104 NM/17 Changed NOTICE file to remove references to embedded PCRE.
106 NM/18 Bugzilla 894: Fix issue with very long lines including comments in lsearch
108 NM/19 Bugzilla 745: TLS version reporting
109 Patch provided by Phil Pennock
111 NM/20 Bugzilla 167: bool: condition support
112 Patch provided by Phil Pennock
114 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken clients
115 Patch provided by Phil Pennock
117 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date
118 Patch provided by Brad "anomie" Jorsch
120 NM/23 Bugzilla 687: Fix misparses in eximstats
121 Patch provided by Heiko Schlittermann
123 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid
124 Patch provided by Heiko Schlittermann
126 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file
128 NM/26 Bugzilla 799: Documentation correction for ratelimit
130 NM/27 Bugzilla 802: Improvements to local interface IP addr detection
131 Patch provided by David Brownlee
133 NM/28 Bugzilla 807: Improvements to LMTP delivery logging
135 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes
137 NM/30 Bugzilla 888: TLS documentation bugfixes
143 TK/01 Add preliminary DKIM support. Currently requires a forked version of
144 ALT-N's libdkim that I have put here:
145 http://duncanthrax.net/exim-experimental/
147 Note to Michael Haardt: I had to rename some vars in sieve.c. They
148 were called 'true' and it seems that C99 defines that as a reserved
149 keyword to be used with 'bool' variable types. That means you could
150 not include C99-style headers which use bools without triggering
151 build errors in sieve.c.
153 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
154 as mailq or other aliases. Changed the --help handling significantly
155 to do whats expected. exim_usage() emits usage/help information.
157 SC/01 Added the -bylocaldomain option to eximstats.
159 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
161 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
163 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
169 PH/01 Another patch from the Sieve maintainer.
171 PH/02 When an IPv6 address is converted to a string for single-key lookup
172 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
173 dots are used instead of colons so that keys in lsearch files need not
174 contain colons. This was done some time before quoting was made available
175 in lsearch files. However, iplsearch files do require colons in IPv6 keys
176 (notated using the quote facility) so as to distinguish them from IPv4
177 keys. This meant that lookups for IP addresses in host lists did not work
178 for iplsearch lookups.
180 This has been fixed by arranging for IPv6 addresses to be expressed with
181 colons if the lookup type is iplsearch. This is not incompatible, because
182 previously such lookups could never work.
184 The situation is now rather anomolous, since one *can* have colons in
185 ordinary lsearch keys. However, making the change in all cases is
186 incompatible and would probably break a number of configurations.
188 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
191 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
192 conversion specification without a maximum field width, thereby enabling
193 a rogue spamd server to cause a buffer overflow. While nobody in their
194 right mind would setup Exim to query an untrusted spamd server, an
195 attacker that gains access to a server running spamd could potentially
196 exploit this vulnerability to run arbitrary code as the Exim user.
198 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
199 $primary_hostname instead of what libspf2 thinks the hosts name is.
201 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
202 a directory entry by the name of the lookup key. Previously, if a
203 symlink pointed to a non-existing file or a file in a directory that
204 Exim lacked permissions to read, a lookup for a key matching that
205 symlink would fail. Now it is enough that a matching directory entry
206 exists, symlink or not. (Bugzilla 503.)
208 PH/03 The body_linecount and body_zerocount variables are now exported in the
211 PH/04 Added the $dnslist_matched variable.
213 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
214 This means they are set thereafter only if the connection becomes
217 PH/06 Added the client_condition to authenticators so that some can be skipped
218 by clients under certain conditions.
220 PH/07 The error message for a badly-placed control=no_multiline_responses left
221 "_responses" off the end of the name.
223 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
225 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
226 (without spaces) instead of just copying the configuration text.
228 PH/10 Added the /noupdate option to the ratelimit ACL condition.
230 PH/11 Added $max_received_linelength.
232 PH/12 Added +ignore_defer and +include_defer to host lists.
234 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
235 way in which PCRE > 7.0 is built.
237 PH/14 Implemented queue_only_load_latch.
239 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
240 MAIL command. The effect was to mangle the value on 64-bit systems.
242 PH/16 Another patch from the Sieve maintainer.
244 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
246 PH/18 If a system quota error occurred while trying to create the file for
247 a maildir delivery, the message "Mailbox is full" was not appended to the
248 bounce if the delivery eventually timed out. Change 4.67/27 below applied
249 only to a quota excession during the actual writing of the file.
251 PH/19 It seems that peer DN values may contain newlines (and other non-printing
252 characters?) which causes problems in log lines. The DN values are now
253 passed through string_printing() before being added to log lines.
255 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
256 and InterBase are left for another time.)
258 PH/21 Added message_body_newlines option.
260 PH/22 Guard against possible overflow in moan_check_errorcopy().
262 PH/23 POSIX allows open() to be a macro; guard against that.
264 PH/24 If the recipient of an error message contained an @ in the local part
265 (suitably quoted, of course), incorrect values were put in $domain and
266 $local_part during the evaluation of errors_copy.
272 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
273 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
276 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
277 issue a MAIL command.
279 PH/02 In an ACL statement such as
281 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
283 if a client was not listed at all, or was listed with a value other than
284 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
285 the condition was not true (as it should be), so access was not denied.
286 The bug was that the ! inversion was incorrectly passed on to the second
287 item. This has been fixed.
289 PH/03 Added additional dnslists conditions == and =& which are different from
290 = and & when the dns lookup returns more than one IP address.
292 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
293 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
295 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
296 FSYNC, which compiles an option called disable_fsync that allows for
297 bypassing fsync(). The documentation is heavily laced with warnings.
299 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
301 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
302 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
303 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
304 including adding "make clean"; (3) Added -fPIC when compiling the test
305 dynamically loaded module, to get rid of a warning.
307 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
308 message fails, move_frozen_messages = true and ignore_bounce_errors_after
309 = 0s. The bug is otherwise harmless.
311 PH/07 There was a bug in the dovecot authenticator such that the value of
312 $auth1 could be overwritten, and so not correctly preserved, after a
313 successful authentication. This usually meant that the value preserved by
314 the server_setid option was incorrect.
316 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
318 PH/09 Installed PCRE release 7.0.
320 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
321 run for batched SMTP input. It is now run at the start of every message
322 in the batch. While fixing this I discovered that the process information
323 (output by running exiwhat) was not always getting set for -bs and -bS
324 input. This is fixed, and it now also says "batched" for BSMTP.
326 PH/11 Added control=no_pipelining.
328 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
329 patch, slightly modified), and move the expansion of helo_data till after
330 the connection is made in the smtp transport (so it can use these
333 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
335 PH/14 Added log_selector = +pid.
337 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
339 PH/16 Add ${if forany and ${if forall.
341 PH/17 Added dsn_from option to vary the From: line in DSNs.
343 PH/18 Flush SMTP output before performing a callout, unless control =
344 no_callout_flush is set.
346 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
347 was true (the default) a successful delivery failed to delete the retry
348 item, thus causing premature timeout of the address. The bug is now
351 PH/20 Added hosts_avoid_pipelining to the smtp transport.
353 PH/21 Long custom messages for fakedefer and fakereject are now split up
354 into multiline reponses in the same way that messages for "deny" and
355 other ACL rejections are.
357 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
358 with slight modification.
360 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
361 draft, changing the syntax and factoring some duplicate code".
363 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
364 for deliveries of the second and subsequent messages over the same SMTP
367 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
368 ${reduce, with only minor "tidies".
370 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
372 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
373 expansion side effects.
375 PH/27 When a message times out after an over-quota error from an Exim-imposed
376 quota, the bounce message says "mailbox is full". This message was not
377 being given when it was a system quota that was exceeded. It now should
380 MH/03 Made $recipients available in local_scan(). local_scan() already has
381 better access to the recipient list through recipients_list[], but
382 $recipients can be useful in postmaster-provided expansion strings.
384 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
385 in the case of a MAIL command with additional options following the
386 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
387 were accidentally chopped off.
389 PH/29 SMTP synchronization checks are implemented when a command is read -
390 there is a check that no more input is waiting when there shouldn't be
391 any. However, for some commands, a delay in an ACL can mean that it is
392 some time before the response is written. In this time, more input might
393 arrive, invalidly. So now there are extra checks after an ACL has run for
394 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
395 pipelining has not been advertised.
397 PH/30 MH's patch to allow iscntrl() characters to be list separators.
399 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
400 returned in the SMTP response when smtp_return_error_details was false.
403 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
404 instead of the C I/O that was originally supplied, because problems were
407 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
408 Exim which did not show up earlier: it was assuming that a call to
409 SSL_CTX_set_info_callback() might give an error value. In fact, there is
410 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
411 was a macro that became an assignment, so it seemed to work. This has
412 changed to a proper function call with a void return, hence the compile
413 error. Exim's code has been fixed.
415 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
418 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
420 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
422 PH/37 If a message is not accepted after it has had an id assigned (e.g.
423 because it turns out to be too big or there is a timeout) there is no
424 "Completed" line in the log. When some messages of this type were
425 selected by exigrep, they were listed as "not completed". Others were
426 picked up by some special patterns. I have improved the selection
427 criteria to be more general.
429 PH/38 The host_find_failed option in the manualroute router can now be set
430 to "ignore", to completely ignore a host whose IP address cannot be
431 found. If all hosts are ignored, the behaviour is controlled by the new
432 host_all_ignored option.
434 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
435 homing or because of multiple MX records with /mx) generated more than
436 one IP address, and the following item turned out to be the local host,
437 all the secondary addresses of the first item were incorrectly removed
438 from the list, along with the local host and any following hosts (which
439 is what is supposed to happen).
441 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
442 whoever called Exim into the -H file. In the case of the daemon it was
443 behaving confusingly. When first started, it used values for whoever
444 started the daemon, but after a SIGHUP it used the Exim user (because it
445 calls itself on a restart). I have changed the code so that it now always
448 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
449 message are rejected with the same error (e.g. no authentication or bad
450 sender address), and a DATA command is nevertheless sent (as can happen
451 with PIPELINING or a stupid MUA), the error message that was given to the
452 RCPT commands is included in the rejection of the DATA command. This is
453 intended to be helpful for MUAs that show only the final error to their
456 PH/42 Another patch from the Sieve maintainer.
458 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
459 Eximstats - Fixed some broken HTML links and added missing column headers
461 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
462 columns for Rejects, Temp Rejects, Ham, and Spam rows.
464 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
466 PH/43 Yet another patch from the Sieve maintainer.
468 PH/44 I found a way to check for a TCP/IP connection going away before sending
469 the response to the final '.' that terminates a message, but only in the
470 case where the client has not sent further data following the '.'
471 (unfortunately, this is allowed). However, in many cases there won't be
472 any further data because there won't be any more messages to send. A call
473 to select() can be used: if it shows that the input is "ready", there is
474 either input waiting, or the socket has been closed. An attempt to read
475 the next input character can distinguish the two cases. Previously, Exim
476 would have sent an OK response which the client would never have see.
477 This could lead to message repetition. This fix should cure that, at
478 least in a lot of common cases.
480 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
481 advertised in response to EHLO.
487 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
488 fixed by 4.65/MH/01 (is this a record?) are fixed:
490 (i) An empty string was always treated as zero by the numeric comparison
491 operators. This behaviour has been restored.
493 (ii) It is documented that the numeric comparison operators always treat
494 their arguments as decimal numbers. This was broken in that numbers
495 starting with 0 were being interpreted as octal.
497 While fixing these problems I realized that there was another issue that
498 hadn't been noticed. Values of message_size_limit (both the global option
499 and the transport option) were treated as octal if they started with 0.
500 The documentation was vague. These values are now always treated as
501 decimal, and I will make that clear in the documentation.
507 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
508 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
511 MH/01 Don't check that the operands of numeric comparison operators are
512 integers when their expansion is in "skipping" mode (fixes bug
513 introduced by 4.64-PH/07).
515 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
516 child addresses, Exim now panics and dies. Previously, because the count
517 is held in a short int, deliveries were likely to be lost. As such a
518 large number of recipients for a single message is ridiculous
519 (performance will be very, very poor), I have chosen to impose a limit
520 rather than extend the field.
526 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
527 leftover -K file (the existence of which was triggered by #402).
528 While we were at it, introduced process PID as part of the -K
529 filename. This should rule out race conditions when creating
532 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
533 processing considerably. Previous code took too long for large mails,
534 triggering a timeout which in turn triggers #401.
536 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
537 in the DK code in transports.c. sendfile() is not really portable,
538 hence the _LINUX specificness.
540 TF/01 In the add_headers option to the mail command in an Exim filter,
541 there was a bug that Exim would claim a syntax error in any
542 header after the first one which had an odd number of characters
545 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
546 callout verification, Exim cached a "reject" for the entire domain. This
547 is correct for most verifications, but it is not correct for a recipient
548 verification with use_sender or use_postmaster set, because in that case
549 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
550 case of MAIL FROM:<> rejection from other early rejections (e.g.
551 rejection of HELO). When verifying a recipient using a non-null MAIL
552 address, the cache is ignored if it shows MAIL FROM:<> rejection.
553 Whatever the result of the callout, the value of the domain cache is
554 left unchanged (for any other kind of callout, getting as far as trying
555 RCPT means that the domain itself is ok).
557 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
560 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
561 manifest itself as EPIPE rather than ECONNECT. When tidying away a
562 session, the daemon ignores ECONNECT errors and logs others; it now
563 ignores EPIPE as well.
565 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
566 (quoted-printable decoding).
568 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
569 later the small subsequent patch to fix an introduced bug.
571 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
573 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
575 PH/08 An error is now given if message_size_limit is specified negative.
577 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
578 to be given (somewhat) arbitrary names.
580 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
583 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
584 miscellaneous code fixes
586 PH/10 Added the log_reject_target ACL modifier to specify where to log
589 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
590 hostname. This is wrong, because it relates to the incoming message (and
591 probably the interface on which it is arriving) and not to the outgoing
592 callout (which could be using a different interface). This has been
593 changed to use the value of the helo_data option from the smtp transport
594 instead - this is what is used when a message is actually being sent. If
595 there is no remote transport (possible with a router that sets up host
596 addresses), $smtp_active_hostname is used.
598 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
599 tweaks were necessary in order to get it to work (see also 21 below):
600 (a) The code assumed that strncpy() returns a negative number on buffer
601 overflow, which isn't the case. Replaced with Exim's string_format()
603 (b) There were several signed/unsigned issues. I just did the minimum
604 hacking in of casts. There is scope for a larger refactoring.
605 (c) The code used strcasecmp() which is not a standard C function.
606 Replaced with Exim's strcmpic() function.
607 (d) The code set only $1; it now sets $auth1 as well.
608 (e) A simple test gave the error "authentication client didn't specify
609 service in request". It would seem that Dovecot has changed its
610 interface. Fortunately there's a specification; I followed it and
611 changed what the client sends and it appears to be working now.
613 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
616 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
617 address A is aliased to B and C, where B exists and C does not. Without
618 -v the output is "A verified" because verification stops after a
619 successful redirection if more than one address is generated. However,
620 with -v the child addresses are also verified. Exim was outputting "A
621 failed to verify" and then showing the successful verification for C,
622 with its parentage. It now outputs "B failed to verify", showing B's
623 parentage before showing the successful verification of C.
625 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
626 look up a TXT record in a specific list after matching in a combined
629 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
630 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
631 they consult the DNS. I had assumed they would set it the way they
632 wanted; and indeed my experiments on Linux seem to show that in some
633 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
634 To be on the safe side, however, I have now made the interface to
635 host_find_byname() similar to host_find_bydns(), with an argument
636 containing the DNS resolver options. The host_find_byname() function now
637 sets these options at its start, just as host_find_bydns() does. The smtp
638 transport options dns_qualify_single and dns_search_parents are passed to
639 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
640 of host_find_byname() use the default settings of RES_DEFNAMES
641 (qualify_single) but not RES_DNSRCH (search_parents).
643 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
644 spool_read_header() do less string testing, by means of a preliminary
645 switch on the second character of optional "-foo" lines. (This is
646 overdue, caused by the large number of possibilities that now exist.
647 Originally there were few.) While I was there, I also converted the
648 str(n)cmp tests so they don't re-test the leading "-" and the first
649 character, in the hope this might squeeze out yet more improvement.
651 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
652 flag allowing group syntax was set by the header_syntax check but not
653 turned off, possible causing trouble later; (2) The flag was not being
654 set at all for the header_verify test, causing "group"-style headers to
655 be rejected. I have now set it in this case, and also caused header_
656 verify to ignore an empty address taken from a group. While doing this, I
657 came across some other cases where the code for allowing group syntax
658 while scanning a header line wasn't quite right (mostly, not resetting
659 the flag correctly in the right place). These bugs could have caused
660 trouble for malformed header lines. I hope it is now all correct.
662 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
663 with the "reply" argument non-NULL. The code, however (which originally
664 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
665 but it didn't always do it. This confused somebody who was copying the
666 code for some other use. I have removed all the tests.
668 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
669 feature that was used to support insecure browsers during the U.S. crypto
670 embargo. It requires special client support, and Exim is probably the
671 only MTA that supported it -- and would never use it because real RSA is
672 always available. This code has been removed, because it had the bad
673 effect of slowing Exim down by computing (never used) parameters for the
674 RSA_EXPORT functionality.
676 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
677 authenticator to fail if there's a tab character in the incoming data
678 (there should never be unless someone is messing about, as it's supposed
679 to be base64-encoded). Also added, on Timo's advice, the "secured" option
680 if the connection is using TLS or if the remote IP is the same as the
681 local IP, and the "valid-client-cert option" if a client certificate has
684 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
685 authenticators. This can be used for authorization after authentication
686 succeeds. (In the case of plaintext, it servers for both authentication
689 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
690 if any retry times were supplied.
692 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
693 connection, where there is no client IP address to check. In this
694 situation, the verify now always succeeds.
696 PH/25 Applied John Jetmore's -Mset patch.
698 PH/26 Added -bem to be like -Mset, but loading a message from a file.
700 PH/27 In a string expansion for a processed (not raw) header when multiple
701 headers of the same name were present, leading whitespace was being
702 removed from all of them, but trailing whitespace was being removed only
703 from the last one. Now trailing whitespace is removed from each header
704 before concatenation. Completely empty headers in a concatenation (as
707 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
708 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
710 PH/29 [Removed. This was a change that I later backed out, and forgot to
711 correct the ChangeLog entry (that I had efficiently created) before
712 committing the later change.]
714 PH/30 Exim was sometimes attempting to deliver messages that had suffered
715 address errors (4xx response to RCPT) over the same connection as other
716 messages routed to the same hosts. Such deliveries are always "forced",
717 so retry times are not inspected. This resulted in far too many retries
718 for the affected addresses. The effect occurred only when there were more
719 hosts than the hosts_max_try setting in the smtp transport when it had
720 the 4xx errors. Those hosts that it had tried were not added to the list
721 of hosts for which the message was waiting, so if all were tried, there
722 was no problem. Two fixes have been applied:
724 (i) If there are any address or message errors in an SMTP delivery, none
725 of the hosts (tried or untried) are now added to the list of hosts
726 for which the message is waiting, so the message should not be a
727 candidate for sending over the same connection that was used for a
728 successful delivery of some other message. This seems entirely
729 reasonable: after all the message is NOT "waiting for some host".
730 This is so "obvious" that I'm not sure why it wasn't done
731 previously. Hope I haven't missed anything, but it can't do any
732 harm, as the worst effect is to miss an optimization.
734 (ii) If, despite (i), such a delivery is accidentally attempted, the
735 routing retry time is respected, so at least it doesn't keep
736 hammering the server.
738 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
739 in ${readsocket because some servers need this prod.
741 PH/32 Added some extra debug output when updating a wait-xxx database.
743 PH/33 The hint "could be header name not terminated by colon", which has been
744 given for certain expansion errors for a long time, was not being given
745 for the ${if def:h_colon_omitted{... case.
747 PH/34 The spec says: "With one important exception, whenever a domain list is
748 being scanned, $domain contains the subject domain." There was at least
749 one case where this was not true.
751 PH/35 The error "getsockname() failed: connection reset by peer" was being
752 written to the panic log as well as the main log, but it isn't really
753 panic-worthy as it just means the connection died rather early on. I have
754 removed the panic log writing for the ECONNRESET error when getsockname()
757 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
758 runs only) independently of the message's sender address. This meant
759 that, if the 4xx error was in fact related to the sender, a different
760 message to the same recipient with a different sender could confuse
761 things. In particualar, this can happen when sending to a greylisting
762 server, but other circumstances could also provoke similar problems.
763 I have changed the default so that the retry time for these errors is now
764 based a combination of the sender and recipient addresses. This change
765 can be overridden by setting address_retry_include_sender=false in the
768 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
769 remote server are returned as part of bounce messages. This was not
770 happening for LMTP over a pipe (the lmtp transport), but now it is the
771 same for both kinds of LMTP.
773 PH/38 Despite being documented as not happening, Exim was rewriting addresses
774 in header lines that were in fact CNAMEs. This is no longer the case.
776 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
777 and queue runs started by the daemon processed all messages. This has
778 been fixed so that -R and -S can now usefully be given with -q<time>.
780 PH/40 Import PCRE release 6.7 (fixes some bugs).
782 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
784 PH/42 Give an error if -q is specified more than once.
786 PH/43 Renamed the variables $interface_address and $interface_port as
787 $received_ip_address and $received_port, to make it clear that these
788 values apply to message reception, and not to the outgoing interface when
789 a message is delivered. (The old names remain recognized, of course.)
791 PH/44 There was no timeout on the connect() call when using a Unix domain
792 socket in the ${readsocket expansion. There now is.
794 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
795 be meaningful with "accept".
797 SC/01 Eximstats V1.43
798 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
800 SC/02 Eximstats V1.44
801 Use a glob alias rather than an array ref in the generated
802 parser. This improves both readability and performance.
804 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
805 Collect SpamAssassin and rejection statistics.
806 Don't display local sender or destination tables unless
807 there is data to show.
808 Added average volumes into the top table text output.
810 SC/04 Eximstats V1.46
811 Collect data on the number of addresses (recipients)
812 as well as the number of messages.
814 SC/05 Eximstats V1.47
815 Added 'Message too big' to the list of mail rejection
816 reasons (thanks to Marco Gaiarin).
818 SC/06 Eximstats V1.48
819 Mainlog lines which have GMT offsets and are too short to
820 have a flag are now skipped.
822 SC/07 Eximstats V1.49 (Alain Williams)
823 Added the -emptyok flag.
825 SC/08 Eximstats V1.50
826 Fixes for obtaining the IP address from reject messages.
828 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
829 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
830 whitesspace changes from 4.64-PH/27
832 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
835 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
836 are found, allow negative numbers in numeric criteria)
838 JJ/06 exipick.20061117.2, added new $message_body_missing variable
840 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
841 to match changes made in 4.64-PH/43
843 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
845 PH/47 Put in an explicit test for a DNS lookup of an address record where the
846 "domain" is actually an IP address, and force a failure. This locks out
847 those revolvers/nameservers that support "A-for-A" lookups, in
848 contravention of the specifications.
850 PH/48 When a host name was looked up from an IP address, and the subsequent
851 forward lookup of the name timed out, the host name was left in
852 $sender_host_name, contrary to the specification.
854 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
855 restricted to single-key lookups, Exim was not diagnosing an error if
856 * or *@ was used with a query-style lookup.
858 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
860 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
861 long ago, but noone interested enough thought of it. Let's just say that
862 the "1.1" means that there are some new functions that weren't there at
863 some point in the past.
865 PH/51 Error processing for expansion failure of helo_data from an smtp
866 transport during callout processing was broken.
868 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
869 tested/used via the -bh/-bhc/-bs options.
871 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
872 bug, fixed in subsequent PCRE releases).
874 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
875 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
877 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
883 SC/01 Use a glob alias rather than an array ref in eximstats generated
884 parser. This improves both readability and performance.
886 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
887 Don't display local sender or destination tables in eximstats unless
888 there is data to show.
889 Added average volumes into the eximstats top table text output.
891 SC/03 Collect data on the number of addresses (recipients) as well
892 as the number of messages in eximstats.
894 TF/01 Correct an error in the documentation for the redirect router. Exim
895 does (usually) call initgroups() when daemonizing.
897 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
898 with consistent privilege compared to when running as a daemon.
900 TF/03 Note in the spec that $authenticated_id is not set for local
901 submissions from trusted users.
903 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
904 Thanks to Dean Brooks <dean@iglou.com> for the patch.
906 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
907 by adding some example configuration directives to the default
908 configuration file. A little bit of work is required to uncomment the
909 directives and define how usernames and passwords are checked, but
910 there is now a framework to start from.
912 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
913 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
914 without this. I don't know how relevant this is to other LDAP libraries.
916 PH/02 Add the verb name to the "unknown ACL verb" error.
918 PH/03 Magnus Holmgren's patch for filter_prepend_home.
920 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
922 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
923 directory not expanded when it should be if an expanded home directory
924 was set for the address (which is overridden by the transport).
926 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
929 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
930 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
931 because it is too late at that time, and has no effect.
933 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
934 security issue with \' (bugzilla #107). I could not use the
935 PQescapeStringConn() function, because it needs a PGconn value as one of
938 PH/08 When testing addresses using -bt, indicate those final addresses that
939 are duplicates that would not cause an additional delivery. At least one
940 person was confused, thinking that -bt output corresponded to deliveries.
941 (Suppressing duplicates isn't a good idea as you lose the information
942 about possibly different redirections that led to the duplicates.)
944 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
945 systems where poll() doesn't work, in particular OS X.
947 PH/10 Added more information to debugging output for retry time not reached.
949 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
950 operations in malware.c.
952 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
955 PH/13 If write_rejectlog was set false when logging was sent to syslog with
956 syslog_duplication set false, log lines that would normally be written
957 both the the main log and to the reject log were not written to syslog at
960 PH/14 In the default configuration, change the use of "message" in ACL warn
961 statements to "add_header".
963 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
964 not followed by a command (e.g. "seen endif").
966 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
967 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
970 PH/17 Added extra conditions to the default value of delay_warning_condition
974 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
975 { match{$h_precedence:}{(?i)bulk|list|junk} } \
976 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
979 The Auto-Submitted: and various List- headers are standardised, whereas I
980 don't think Precedence: ever was.
982 PH/18 Refactored debugging code in route_finduser() to show more information,
983 in particular, the error code if getpwnam() issues one.
985 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
986 This is apparently needed in addition to the PH/07 change above to avoid
987 any possible encoding problems.
989 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
990 but not after initializing Perl.
992 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
993 output them only if debugging. By default they are written stderr,
994 apparently, which is not desirable.
996 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
999 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1002 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1004 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1005 authenticated or an ident call has been made. Suppress the default
1006 values for $authenticated_id and $authenticated_sender (but permit -oMai
1007 and -oMas) when testing with -bh.
1009 PH/24 Re-jigged the order of the tests in the default configuration so that the
1010 tests for valid domains and recipients precede the DNS black list and CSA
1011 tests, on the grounds that those ones are more expensive.
1013 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1014 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1015 This bug exists in every version of Exim that I still have, right back to
1018 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1019 However, an attempt to turn on case-sensitivity in a regex key by
1020 including (?-i) didn't work because the subject string was already
1021 lowercased, and the effects were non-intuitive. It turns out that a
1022 one-line patch can be used to allow (?-i) to work as expected.
1028 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1029 other effects) broke the use of negated acl sub-conditions.
1031 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1034 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1035 "Deny" causes Exim to reject the incoming connection with a 554 error.
1036 Unfortunately, if there is a major crisis, such as a disk failure,
1037 tcp-wrappers gives "deny", whereas what one would like would be some
1038 kind of temporary error. A kludge has been added to help with this.
1039 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1040 554 error is used if errno is still zero or contains ENOENT (which occurs
1041 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1044 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1046 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1047 errors. Otherwise a message that provokes a temporary error (when other
1048 messages do not) can cause a whole host to time out.
1050 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1051 addresses were routed directly to files or pipes from a redirect router.
1052 File deliveries just didn't batch; pipe deliveries might have suffered
1055 PH/06 A failure to get a lock for a hints database would erroneously always say
1056 "Failed to get write lock", even when it was really a read lock.
1058 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1059 of 0600. This has been changed to use the value of the lockfile_mode
1060 option (which defaults to 0600).
1062 PH/08 Applied small patch from the Sieve maintainer.
1064 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1065 folder from quota calculations, a direct delivery into this folder messed
1066 up the contents of the maildirsize file. This was because the regex was
1067 used only to exclude .Trash (or whatever) when the size of the mailbox
1068 was calculated. There was no check that a delivery was happening into an
1069 excluded directory. This bug has been fixed by ignoring all quota
1070 processing for deliveries into excluded directories.
1072 PH/10 Added the maildirfolder_create_regex option to appendfile.
1078 PH/01 The code for finding all the local interface addresses on a FreeBSD
1079 system running IPv6 was broken. This may well have applied to all BSD
1080 systems, as well as to others that have similar system calls. The broken
1081 code found IPv4 interfaces correctly, but gave incorrect values for the
1082 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1083 that it would not match correctly against @[] and not recognize the IPv6
1086 PH/02 The ipliteral router was not recognizing addresses of the form user@
1087 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1089 PH/03 Added disable_ipv6.
1091 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1092 decoded form, because it is most often used to construct To: headers
1093 lines in autoreplies, and the decoded form may well be syntactically
1094 invalid. However, $reply_address has leading white space removed, and all
1095 newlines turned into spaces so that the autoreply transport does not
1098 PH/05 If group was specified without a user on a router, and no group or user
1099 was specified on a transport, the group from the router was ignored.
1101 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1102 for visible compile-time settings that can be used to change these
1103 numbers, for those that want even more. Backwards compatibility with old
1104 spool files has been maintained. However, going back to a previous Exim
1105 release will lost any variables that are in spool files.
1107 PH/07 Two small changes when running in the test harness: increase delay when
1108 passing a TCP/IP connection to a new process, in case the original
1109 process has to generate a bounce, and remove special handling of
1110 127.0.0.2 (sic), which is no longer necessary.
1112 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1113 be the same on different OS.
1115 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1118 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1119 whether --show-vars was specified or not
1121 JJ/02 exipick: Added support for new ACL variable spool format introduced
1124 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1125 syntactically invalid From: or Reply-to: line, and a filter used this to
1126 generate an autoreply, and therefore failed to obtain an address for the
1127 autoreply, Exim could try to deliver to a non-existent relative file
1128 name, causing unrelated and misleading errors. What now happens is that
1129 it logs this as a hard delivery error, but does not attempt to create a
1132 PH/11 The exinext utility has a -C option for testing purposes, but although
1133 the given file was scanned by exinext itself; it wasn't being passed on
1134 when Exim was called.
1136 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1137 an end-of-file indication when reading a command response.
1139 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1140 compiled. In many other places in Exim, IPv6 addresses are always
1141 recognized, so I have changed this. It also means that IPv4 domain
1142 literals of the form [IPV4:n.n.n.n] are now always recognized.
1144 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1145 used if the router is not running as root, for example, when verifying at
1146 ACL time, or when using -bh. The debugging output from this situation was
1147 non-existent - all you got was a failure to exec. I have made two
1150 (a) Failures to set uid/gid, the current directory, or a process leader
1151 in a subprocess such as that created by queryprogram now generate
1152 suitable debugging ouput when -d is set.
1154 (b) The queryprogram router detects when it is not running as root,
1155 outputs suitable debugging information if -d is set, and then runs
1156 the subprocess without attempting to change uid/gid.
1158 PH/15 Minor change to Makefile for building test_host (undocumented testing
1161 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1162 additional section of a DNS packet that returns MX or SRV records.
1163 Instead, it always explicitly searches for A/AAAA records. This avoids
1164 major problems that occur when a DNS server includes only records of one
1165 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1166 fixed another bug: if SRV records were looked up and the corresponding
1167 address records were *not* found in the additional section, the port
1168 values from the SRV records were lost.
1170 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1171 using the correct key (the original address) when searching the retry
1172 rules in order to find which one to use for generating the retry hint.
1174 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1175 adding the default one. Similarly, if it contains a Reply-To: header, the
1176 errors_reply_to option, if set, is not used.
1178 PH/19 When calculating a retry time, Exim used to measure the "time since
1179 failure" by looking at the "first failed" field in the retry record. Now
1180 it does not use this if it is later than than the arrival time of the
1181 message. Instead it uses the arrival time. This makes for better
1182 behaviour in cases where some deliveries succeed, thus re-setting the
1183 "first failed" field. An example is a quota failure for a huge message
1184 when small messages continue to be delivered. Without this change, the
1185 "time since failure" will always be short, possible causing more frequent
1186 delivery attempts for the huge message than are intended.
1187 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1189 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1190 $1, $2, $3) because the numerical variables can be reset during some
1191 expansion items (e.g. "match"), thereby losing the authentication data.
1193 PH/21 Make -bV show the size of off_t variables so that the test suite can
1194 decide whether to run tests for quotas > 2G.
1196 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1197 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1198 filecount value is greater than 2G or if a quota value is greater than 2G
1199 on a system where the size of off_t is not greater than 4, a panic error
1202 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1203 never match. The debug and -bh output now contains an explicit error
1204 message indicating a malformed IPv4 address or mask.
1206 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1207 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1208 PH/23 above applies.
1210 PH/25 Do not write to syslog when running in the test harness. The only
1211 occasion when this arises is a failure to open the main or panic logs
1212 (for which there is an explicit test).
1214 PH/26 Added the /no_tell option to "control=freeze".
1216 PH/27 If a host name lookup failed very early in a connection, for example, if
1217 the IP address matched host_lookup and the reverse lookup yielded a name
1218 that did not have a forward lookup, an error message of the form "no IP
1219 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1220 could be logged. Now it outputs the IP address instead of "NULL".
1222 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1223 allows the sender and the authenticated sender to be set when
1224 submitting a message from within Exim. Since child_open_exim() is
1225 documented for local_scan(), the new function should be too.
1227 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1228 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1229 results in an empty string is now treated as unset.
1231 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1233 PH/31 Added sender_verify_fail logging option.
1235 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1236 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1237 tidied the source and removed it altogether.
1239 PH/33 When a queue run was abandoned because the load average was too high, a
1240 log line was always written; now it is written only if the queue_run log
1241 selector is set. In addition, the log line for abandonment now contains
1242 information about the queue run such as the pid. This is always present
1243 in "start" and "stop" lines but was omitted from the "abandon" line.
1245 PH/34 Omit spaces between a header name and the colon in the error message that
1246 is given when verify = headers_syntax fails (if there are lots of them,
1247 the message gets confusing).
1249 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1250 names, as there are now some PTR records that contain slashes. This check
1251 is only to protect against broken name servers that fall over on strange
1252 characters, so the fact that it applies to all lookups doesn't matter.
1254 PH/36 Now that the new test suite is complete, we can remove some of the
1255 special code in Exim that was needed for the old test suite. For example,
1256 sorting DNS records because real resolvers return them in an arbitrary
1257 order. The new test suite's fake resolver always returns records in the
1260 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1261 bounces) except when queue_only is set, to avoid logging races between
1262 the different processes.
1264 PH/38 Panic-die if .include specifies a non-absolute path.
1266 PH/39 A tweak to the "H" retry rule from its user.
1268 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1269 a label. They prevented compilation on older perls.
1271 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1272 a warning to be raised on newish perls.
1274 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1275 on queue. Changes to match documented behaviour of showing count of
1276 messages matching specified criteria.
1278 PH/40 Changed the default ident timeout from 30s to 5s.
1280 PH/41 Added support for the use of login_cap features, on those BSD systems
1281 that have them, for controlling the resources used by pipe deliveries.
1283 PH/42 The content-scanning code uses fopen() to create files in which to put
1284 message data. Previously it was not paying any attention to the mode of
1285 the files. Exim runs with umask(0) because the rest of the code creates
1286 files with open(), and sets the required mode explicitly. Thus, these
1287 files were ending up world-writeable. This was not a big issue, because,
1288 being within the spool directory, they were not world-accessible. I have
1289 created a function called modefopen, which takes an additional mode
1290 argument. It sets umask(777), creates the file, chmods it to the required
1291 mode, then resets the umask. All the relevant calls to fopen() in the
1292 content scanning code have been changed to use this function.
1294 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1295 to 24 hours. This avoids potential overflow problems when processing G
1296 and H retry rules. I suspect nobody ever tinkers with this value.
1298 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1300 PH/45 When the plaintext authenticator is running as a client, the server's
1301 challenges are checked to ensure they are valid base64 strings. By
1302 default, the authentication attempt is cancelled if an invalid string is
1303 received. Setting client_ignore_invalid_base64 true ignores these errors.
1304 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1305 they are received. Thus, the responses can be made to depend on the
1306 challenges. If an invalid string is ignored, an empty string is placed in
1309 PH/46 Messages that are created by the autoreply transport now contains a
1310 References: header, in accordance with RFCs 2822 and 3834.
1312 PH/47 Added authenticated_sender_force to the smtp transport.
1314 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1316 PH/49 Installed latest patch from the Sieve maintainer.
1318 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1319 was also set, the appendfile transport was unnecessarily scanning a
1320 directory of message files (e.g. for maildir delivery) to find the count
1321 of files (along with the size), even though it did not need this
1322 information. It now does the scan only if it needs to find either the
1323 size of the count of files.
1325 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1327 PH/52 Two bugs concerned with error handling when the smtp transport is
1330 (i) Exim was not creating retry information for temporary errors given
1331 for individual recipients after the DATA command when the smtp transport
1332 was used in LMTP mode. This meant that they could be retried too
1333 frequently, and not timed out correctly.
1335 (ii) Exim was setting the flag that allows error details to be returned
1336 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1337 recipients that were returned after the DATA command.
1339 PH/53 This is related to PH/52, but is more general: for any failing address,
1340 when detailed error information was permitted to be returned to the
1341 sender, but the error was temporary, then after the final timeout, only
1342 "retry timeout exceeded" was returned. Now it returns the full error as
1343 well as "retry timeout exceeded".
1345 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1346 do this, and (what is worse) MTAs that accept it.
1348 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1349 will now be deprecated.
1351 PH/56 New os.c-cygwin from the Cygwin maintainer.
1353 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1354 formats (previously only available in exim formats via -bpr, -bpru,
1355 and -bpra. Now also available in native and exiqgrep formats)
1357 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1358 with very large, slow to parse queues
1360 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1362 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1364 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1365 responses to authentication challenges, though it was showing the
1366 challenges; (ii) I've removed the CR characters from the debug output for
1369 PH/58 Allow for the insertion of a newline as well as a space when a string
1370 is turned into more than one encoded-word during RFC 2047 encoding. The
1371 Sieve code now uses this.
1373 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1374 data_4xx, lost_connection, tls_required.
1376 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1377 message was being sent as an SMTP response.
1379 PH/61 Add -l and -k options to exicyclog.
1381 PH/62 When verifying, if an address was redirected to one new address, so that
1382 verification continued, and the new address failed or deferred after
1383 having set something in $address_data, the value of $address_data was not
1384 passed back to the ACL. This was different to the case when no
1385 redirection occurred. The value is now passed back in both cases.
1387 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1388 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1389 use login_cap.h, so on its own it isn't the distinguishing feature. The
1390 new name refers directly to the setclassresources() function.
1392 PH/65 Added configuration files for NetBSD3.
1394 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1396 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1397 is preferred over IPv4.
1399 PH/68 The bounce_return_message and bounce_return_body options were not being
1400 honoured for bounces generated during the reception of non-SMTP messages.
1401 In particular, this applied to messages rejected by the ACL. This bug has
1402 been fixed. However, if bounce_return_message is true and bounce_return_
1403 body is false, the headers that are returned for a non-SMTP message
1404 include only those that have been read before the error was detected.
1405 (In the case of an ACL rejection, they have all been read.)
1407 PH/69 The HTML version of the specification is now built in a directory called
1408 spec_html instead of spec.html, because the latter looks like a path with
1409 a MIME-type, and this confuses some software.
1411 PH/70 Catch two compiler warnings in sieve.c.
1413 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1414 function verify_get_ident() calls ip_connect() to connect a socket, but
1415 if the "connect()" function timed out, ip_connect() used to close the
1416 socket. However, verify_get_ident() also closes the socket later, and in
1417 between Exim writes to the log, which may get opened at this point. When
1418 the socket was closed in ip_connect(), the log could get the same file
1419 descriptor number as the socket. This naturally causes chaos. The fix is
1420 not to close the socket in ip_connect(); the socket should be closed by
1421 the function that creates it. There was only one place in the code where
1422 this was missing, in the iplookup router, which I don't think anybody now
1423 uses, but I've fixed it anyway.
1425 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1426 well as to direct DNS lookups. Otherwise the handling of names in host
1427 lists is inconsistent and therefore confusing.
1433 PH/01 Two changes to the default runtime configuration:
1435 (1) Move the checks for relay_from_hosts and authenticated clients from
1436 after to before the (commented out) DNS black list checks.
1438 (2) Add control=submission to the relay_from_hosts and authenticated
1439 clients checks, on the grounds that messages accepted by these
1440 statements are most likely to be submissions.
1442 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1444 (1) Generate an error if the third argument for the ${prvs expansion is
1447 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1450 (3) Reset the variables that are obtained from the first argument of
1451 ${prvscheck and used in the second argument before leaving the code,
1452 because their memory is reclaimed, so using them afterwards may do
1455 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1456 one (it's much easier than Tom thought :-).
1458 (5) Because of (4), we can now allow for the use of $prvscheck_result
1459 inside the third argument.
1461 PH/03 For some reason, the default setting of PATH when running a command from
1462 a pipe transport was just "/usr/bin". I have changed it to
1465 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1466 anything to be listed in the output from -bV.
1468 PH/05 When a filter generated an autoreply, the entire To: header line was
1469 quoted in the delivery log line, like this:
1471 => >A.N.Other <ano@some.domain> <original@ddress> ...
1473 This has been changed so that it extracts the operative address. There
1474 may be more than one such address. If so, they are comma-separated, like
1477 => >ano@some.domain,ona@other.domain <original@ddress> ...
1479 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1480 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1481 not being looked up in the rDNS to get a host name, Exim was showing the
1482 IP address twice in Received: lines, even though the IP addresses were
1483 identical. For example:
1485 Received: from [1.2.3.4] (helo=[1.2.3.4])
1487 However, if the real host name was known, it was omitting the HELO data
1488 if it matched the actual IP address. This has been tidied up so that it
1489 doesn't show the same IP address twice.
1491 PH/07 When both +timestamp and +memory debugging was on, the value given by
1492 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1493 called by the debug printing, thereby overwriting the timestamp buffer.
1494 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1496 PH/08 When the original message was included in an autoreply transport, it
1497 always said "this is a copy of the message, including all the headers",
1498 even if body_only or headers_only was set. It now gives an appropriate
1501 PH/09 Applied a patch from the Sieve maintainer which:
1503 o fixes some comments
1504 o adds the (disabled) notify extension core
1505 o adds some debug output for the result of if/elsif tests
1506 o points to the current vacation draft in the documentation
1507 and documents the missing references header update
1511 o fixes a bug in processing the envelope test (when testing
1512 multiple envelope elements, the last element determinted the
1515 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1516 Electronic Mail") by including:
1518 Auto-submitted: auto-generated
1520 in the messages that it generates (bounce messages and others, such as
1521 warnings). In the case of bounce messages for non-SMTP mesages, there was
1522 also a typo: it was using "Auto_submitted" (underscore instead of
1523 hyphen). Since every message generated by Exim is necessarily in response
1524 to another message, thes have all been changed to:
1526 Auto-Submitted: auto-replied
1528 in accordance with these statements in the RFC:
1530 The auto-replied keyword:
1532 - SHOULD be used on messages sent in direct response to another
1533 message by an automatic process,
1535 - MUST NOT be used on manually-generated messages,
1537 - MAY be used on Delivery Status Notifications (DSNs) and Message
1538 Disposition Notifications (MDNs),
1540 - MUST NOT be used on messages generated by automatic or periodic
1541 processes, except for messages which are automatic responses to
1544 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1545 to the default Received: header definition.
1547 PH/12 Added log selector acl_warn_skipped (default on).
1549 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1550 variables because (a) they are in the wrong storage pool and (b) even if
1551 they were copied, it wouldn't work properly because of the caching.
1553 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1554 checking when decoding. Apparently there are clients that generate
1555 overlong encoded strings. Why am I not surprised?
1557 PH/15 If the first argument of "${if match_address" was not empty, but did not
1558 contain an "@" character, Exim crashed. Now it writes a panic log message
1559 and treats the condition as false.
1561 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1563 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1564 extension "envelope-auth". The code is finished and in agreement with
1565 other implementations, but there is no documentation so far and in fact,
1566 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1567 not changing the active code.
1569 Print executed "if" and "elsif" statements when debugging is used. This
1570 helps a great deal to understand what a filter does.
1572 Document more things not specified clearly in RFC3028. I had all this
1573 sorted out, when out of a sudden new issues came to my mind. Oops."
1575 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1578 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1579 canonical form (as documented). However, after a host name lookup from
1580 the IP address, check_host() was doing a simple string comparison with
1581 addresses acquired from the DNS when checking that the found name did
1582 have the original IP as one of its addresses. Since any found IPv6
1583 addresses are likely to be in abbreviated form, the comparison could
1584 fail. Luckily, there already exists a function for doing the comparison
1585 by converting both addresses to binary, so now that is used instead of
1586 the text comparison.
1588 PH/20 There was another similar case to PH/19, when a complete host name was
1589 given in a host list; looking up its IP address could give an abbreviated
1590 form, whereas the current host's name might or might not be abbreviated.
1591 The same fix has been applied.
1597 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1598 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1601 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1602 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1604 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1606 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1607 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1608 Also, for "Auto-Submitted", treat anything other than "no" as
1609 non-personal, in accordance with RFC 3834. (Previously it treated
1610 anything starting "auto-" as non-personal.)
1612 TF/01 The control=submission/name=... option had a problem with syntax
1613 errors if the name included a slash character. The /name= option
1614 now slurps the rest of the string, so it can include any characters
1615 but it must come last in the list of options (after /sender_retain
1618 PH/05 Some modifications to the interface to the fake nameserver for the new
1626 TK/01 Added the "success_on_redirect" address verification option. See
1627 NewStuff for rationale and an example.
1629 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1631 PH/02 Patch to exigrep to allow it to work on syslog lines.
1633 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1634 fread() to skip over the body file's header line, because in Cygwin the
1635 header line is locked and is inaccessible.
1637 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1638 co-exist for some time) to make it clear that it is the Exim ID that is
1639 referenced, not the Message-ID: header line.
1641 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1642 string_format() function, because snprintf() does not exist on all
1645 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1646 ${stat: expansion item.
1648 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1649 protocol synchronization error", to keep the pedants happy.
1651 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1652 well as for IRIX systems, when gcc is being used. See the host.c source
1655 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1657 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1660 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1661 transport and to the smtp transport in LMTP mode.
1663 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1665 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1666 run a filter in a subprocess. This could lead to confusion in subsequent
1667 lookups in the parent process. There should also be a search_tidyup() at
1668 the end of the subprocess.
1670 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1671 only if the host matched helo_try_verify_hosts, which caused the
1672 verification to occur when the EHLO/HELO command was issued. The ACL just
1673 tested the remembered result. Now, if a previous verification attempt has
1674 not happened, "verify = helo" does it there and then.
1676 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1678 TK/03 Fix log output including CR from clamd.
1680 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1681 exist provoked a memory error which could cause a segfault.
1683 PH/15 Installed PCRE 6.2
1685 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1687 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1688 of the problem. Specifically, suggested +O2 rather than +O1 for the
1691 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1693 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1694 record, the port was not correctly logged when the outgoing_port log
1695 selector was set (it logged the transort's default port).
1697 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1698 fallback_hosts, and "hosts" in the smtp transport.
1700 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1701 host errors such as "Connection refused".
1703 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1704 authentication with radiusclient 0.4.9:
1706 - Error returned from rc_read_config was caught wrongly
1707 - Username/password not passed on to radius server due to wrong length.
1709 The presumption is that some radiusclient API changes for 4.51/PH/17
1710 were not taken care of correctly. The code is still untested by me (my
1711 Linux distribution still has 0.3.2 of radiusclient), but it was
1712 contributed by a Radius user.
1714 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1715 expanding the "port" option of the smtp transport.
1717 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1718 while reading a MIME header. Thanks to Tom Hughes for a patch.
1720 PH/24 Include config.h inside local_scan.h so that configuration settings are
1723 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1724 that in an ACL for RCPT (for example), you can examine exactly what was
1727 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1728 commands, but it was not correctly comparing the address with the actual
1729 client host address. Thus, it would show the EHLO address in Received:
1730 header lines when this was not necessary.
1732 PH/27 Added the % operator to ${eval:}.
1734 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1735 it should be ignoring failures (because with -C, for example, it has lost
1736 privilege). It wasn't ignoring creation failures other than "already
1739 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1742 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1743 because an input line was too long, either on its own, or by virtue of
1744 too many continuations, the temporary file was not being removed, and the
1745 return code was incorrect.
1747 PH/31 Missing "BOOL" in function definition in filtertest.c.
1749 PH/32 Applied Sieve patches from the maintainer.
1751 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1753 PH/33 Added "verify = not_blind".
1755 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1756 Local/Makefile (with some defaults set). These are used in built scripts
1757 such as exicyclog, but they have never been used in the exim_install
1758 script (though there are many overriding facilities there). I have
1759 arranged that the exim_install script now takes note of these two
1762 PH/35 Installed configuration files for Dragonfly.
1764 PH/36 When a locally submitted message by a trusted user did not contain a
1765 From: header, and the sender address was obtained from -f or from an SMTP
1766 MAIL command, and the trusted user did not use -F to supply a sender
1767 name, $originator_name was incorrectly used when constructing a From:
1768 header. Furthermore, $originator_name was used for submission mode
1769 messages from external hosts without From: headers in a similar way,
1770 which is clearly wrong.
1772 PH/37 Added control=suppress_local_fixups.
1774 PH/38 When log_selector = +received_sender was set, and the addition of the
1775 sender made the log line's construction buffer exactly full, or one byte
1776 less than full, an overflow happened when the terminating "\n" was
1779 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1780 when the result of a list match is failure because a DNS lookup failed.
1782 PH/40 RM_COMMAND is now used in the building process.
1784 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1785 the "build-* directories that it finds.
1787 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1788 address in a domain literal was a prefix of an interface address.
1790 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1791 when verifying a sender address, unless rewrite_headers is false.
1793 PH/44 Wrote a long comment about why errors_to addresses are verified as
1794 recipients, not senders.
1796 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1797 the ratelimit ACL was added.
1799 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1801 PH/46 Added extra information about PostgreSQL errors to the error string.
1803 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1804 avoiding the need to install special zones in a real server. This is
1805 backwards compatible; if it can't find the fake resolver, it drops back.
1806 Thus, both old and new test suites can be run.
1808 TF/02 Added util/ratelimit.pl
1810 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1811 clock is set back in time.
1813 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1814 Candler <B.Candler@pobox.com>.
1816 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1817 OK for addresses that are the result of redirections.
1819 PH/48 A number of further additions for the benefit of the new test suite,
1820 including a fake gethostbyname() that interfaces to the fake DNS resolver
1823 TF/06 The fix for widen_domains has also been applied to qualify_single and
1824 search_parents which are the other dnslookup options that can cause
1827 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1830 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1832 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1833 service" if it encountered an MX record with an empty target hostname.
1834 The message is now "an MX or SRV record indicated no SMTP service".
1836 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1837 if the DNS of the sending site is misconfigured. This is quite a
1838 common situation. This change restores the behaviour of treating a
1839 helo verification defer as a failure.
1841 PH/51 If self=fail was set on a router, the bounce message did not include the
1842 actual error message.
1848 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1850 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1851 command itself ended in error, the underlying message about the transport
1852 filter timeout was being overwritten with the pipe command error. Now the
1853 underlying error message should be appended to the second error message.
1855 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1857 PH/02 Reduce the amount of output that "make" produces by default. Full output
1858 can still be requested.
1860 PH/03 The warning log line about a condition test deferring for a "warn" verb
1861 was being output only once per connection, rather than after each
1862 occurrence (because it was using the same function as for successful
1863 "warn" verbs). This seems wrong, so I have changed it.
1865 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1866 it should not have, which might have caused a crash in the right
1867 circumstances, but probably never did.
1869 PH/04 Installed a modified version of Tony Finch's patch to make submission
1870 mode fix the return path as well as the Sender: header line, and to
1871 add a /name= option so that you can make the user's friendly name appear
1874 TF/03 Added the control = fakedefer ACL modifier.
1876 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1877 Mark Lowes for thorough testing.
1879 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1881 TK/03 Merged latest SRS patch from Miles Wilton.
1883 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1884 with the definition in sysexits.h (which is #included earlier).
1885 Fortunately, Exim does not actually use EX_OK. The code used to try to
1886 preserve the sysexits.h value, by assumimg that macro definitions were
1887 scanned for macro replacements. I have been disabused of this notion,
1888 so now the code just undefines EX_OK before #including unistd.h.
1890 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1891 in the smtp transport. When a block could not be written in a single
1892 write() function, the timeout was being re-applied to each part-write.
1893 This seems wrong - if the receiver was accepting one byte at a time it
1894 would take for ever. The timeout is now adjusted when this happens. It
1895 doesn't have to be particularly precise.
1897 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1898 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1900 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1901 without a domain if the check to <postmaster@domain> fails.
1903 SC/01 Eximstats: added -xls and the ability to specify output files
1904 (patch written by Frank Heydlauf).
1906 SC/02 Eximstats: use FileHandles for outputing results.
1908 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1910 SC/04 Eximstats: fixed display of large numbers with -nvr option
1912 SC/05 Eximstats: fixed merging of reports with empty tables.
1914 SC/06 Eximstats: added the -include_original_destination flag
1916 SC/07 Eximstats: removed tabs and trailing whitespace.
1918 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1920 TK/06 MBOX spool code: Add real "From " MBOX separator line
1921 so the .eml file is really in mbox format (even though
1922 most programs do not really care). Patch from Alex Miller.
1924 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1925 The latter is generated from $received_to and is only set if the
1926 message has one envelope recipient. SA can use these headers,
1927 obviously out-of-the-box. Patch from Alex Miller.
1929 PH/08 The ${def test on a variable was returning false if the variable's
1930 value was "0", contrary to what the specification has always said!
1931 The result should be true unless the variable is empty.
1933 PH/09 The syntax error of a character other than { following "${if
1934 def:variable_name" (after optional whitespace) was not being diagnosed.
1935 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1936 accidental colon was present, for example, could give incorrect results.
1938 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1939 result is used (not including appendfile, where other changes are about
1942 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1943 This involved changing a lot of size variables from int to off_t. It
1944 should work with maildirs and everything.
1946 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1947 spamd dying while we are connected to it.
1949 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1952 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1953 function for BDB 4.3.
1955 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1957 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1958 the sources for PCRE have been re-arranged and more files are now
1961 PH/15 The code I had for printing potentially long long variables in PH/11
1962 above was not the best (it lost precision). The length of off_t variables
1963 is now inspected at build time, and an appropriate printing format (%ld
1964 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1965 to be "long long int" or "long int". This is needed for the internal
1966 formatting function string_vformat().
1968 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1969 the configuration file to be ":syslog", then the script "guesses" where
1970 the logs files are, rather than using the compiled in default. In our
1971 case the guess is not the same as the compiled default, so the script
1972 suddenly stopped working when I started to use syslog. The patch checks
1973 to see if log_file_path is "". If so, it attempts to read it from exim
1974 with no configuration file to get the compiled in version, before it
1975 falls back to the previous guessing code."
1977 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1978 implementing BATV in an Exim configuration. See NewStuff for the gory
1981 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1982 Makefile that are specific to HP-UX.
1984 PH/18 If the "use_postmaster" option was set for a recipient callout together
1985 with the "random" option, the postmaster address was used as the MAIL
1986 FROM address for the random test, but not for the subsequent recipient
1987 test. It is now used for both.
1989 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1990 patch removes a few documentation additions to RFC 3028, because the
1991 latest draft now contains them. It adds the new en;ascii-case comparator
1992 and a new error check for 8bit text in MIME parts. Comparator and
1993 require names are now matched exactly. I enabled the subaddress
1994 extension, but it is not well tested yet (read: it works for me)."
1996 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1997 rework some of the code of TK/09 above to avoid the hardwired use of
1998 "%lld" and "long long". Replaced the call to snprintf() with a call to
2001 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2002 records point to non-existent hosts", "retry timeout exceeded", and
2003 "retry time not reached for any host after a long failure period".
2005 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2006 experimental DomainKeys support:
2008 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2009 (2) On an error such as an illegally used "control", the wrong name for
2010 the control was given.
2012 These problems did NOT occur unless DomainKeys support was compiled.
2014 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2016 PH/24 Added ${if match_ip condition.
2018 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2019 databases so that it will be absolutely obvious if a crash occurs in the
2020 DB library. This is a regular occurrence (often caused by mis-matched
2023 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2024 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2025 picked up on a user's system that detects such things. There doesn't seem
2026 to be a gcc warning option for this - only an attribute that has to be
2027 put on the function's prototype. It seems that in Fedora Core 4 they have
2028 set this on a number of new functions. No doubt there will be more in due
2031 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2032 specify a transport. However, if an address that was verified by such a
2033 router was the subject of a callout, Exim crashed because it tried to
2034 read the rcpt_include_affixes from the non-existent transport. Now it
2035 just assumes that the setting of that option is false. This bug was
2036 introduced by 4.51/PH/31.
2038 PH/28 Changed -d+all to exclude +memory, because that information is very
2039 rarely of interest, but it makes the output a lot bigger. People tend to
2040 do -d+all out of habit.
2042 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2043 code in os-type was giving problems when libc.so lives in lib64, like on
2046 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2047 aren't the modern standard, and it seems that some systems' include files
2048 don't always have them. Exim was already checking for some of the newer
2049 ones like T_AAAA, and defining it itself. I've added checks for all the
2050 record types that Exim uses.
2052 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2053 not automatically generating a new one, as it is supposed to. This
2054 prevented TLS from working. If the file did exist, but contained invalid
2055 data, a new version was generated, as expected. It was only the case of a
2056 non-existent file that was broken.
2058 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2059 with a change in libdomainkeys > 0.64.
2061 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2062 from DNS. If the selector record carries the flag, it now has
2063 precedence over the domain-wide flag.
2065 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2067 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2068 the use of a port name that isn't defined in /etc/services) occurred, the
2069 message was deferred as in a normal delivery, and thus remained on the
2070 spool, instead of being failed because of the mua_wrapper setting. This
2071 is now fixed, and I tidied up some of the mua_wrapper messages at the
2074 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2075 the messages in a hash of arrays rather than using individual hashes.
2076 This is a bit cleaner and results in dramatic memory savings, albeit
2077 at a slight CPU cost.
2079 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2080 as requested by Marc Sherman.
2082 SC/10 Eximstats: added histograms for user specified patterns as requested
2085 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2087 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2088 fopen() in the content-scanning modules that did not already have it.
2094 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2095 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2097 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2099 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2101 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2102 to test Sieve filters that use "vacation".
2104 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2105 that changes the way the GnuTLS parameters are stored in the cache file.
2106 The new format can be generated externally. For backward compatibility,
2107 if the data in the cache doesn't make sense, Exim assumes it has read an
2108 old-format file, and it generates new data and writes a new file. This
2109 means that you can't go back to an older release without removing the
2112 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2113 work if there are any delivery delays because "one_time" forces the
2114 parent to be marked "delivered", so its unseen clone is never tried
2115 again. For this reason, Exim now forbids the simultaneous setting of
2118 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2119 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2120 entry if you want to know the details. The fix, however, neglected to
2121 consider the case when local delivery batching is involved. The test for
2122 "previously delivered" was not happening when checking to see if an
2123 address could be batched with a previous (undelivered) one; under
2124 certain circumstances this could lead to multiple deliveries to the same
2127 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2128 in its include files, and this causes problems building Exim.
2130 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2131 header_syntax) but Exim was just ignoring anything given after a slash.
2132 In particular, this caused confusion with an attempt to use "verify =
2133 reverse_host_lookup/defer_ok". An error is now given when options are
2134 supplied for verify items that do not have them. (Maybe reverse_host_
2135 lookup should have a defer_ok option, but that's a different point.)
2137 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2138 defined by RFC 821) to 2048, because there were problems with some AUTH
2139 commands, and RFC 1869 says the size should be increased for extended
2140 SMTP commands that take arguments.
2142 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2145 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2146 "unknown" error; now it says that the functionality isn't in the binary.
2148 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2149 an address' error message when a string expansion fails (syntax or
2150 whatever). Otherwise the password may appear in the log. Following change
2151 PH/42 below, there is no longer a chance of it appearing in a bounce
2154 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2156 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2157 of its addresses was ever tried. (Bugzilla bug #2.)
2159 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2160 the result incorrectly in the debug output. (It correctly added a newline
2161 to what was transported.)
2163 TF/01 Added $received_time.
2165 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2166 commented out examples of how to interface to a virus scanner and to
2167 SpamAssassin. Also added commented examples of av_scanner and
2168 spamd_address settings.
2170 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2171 and controls are allowed in which ACLs. There were a couple of minor
2172 errors. Some of the entries in the conditions table (which is a table of
2173 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2174 negation of where the condition IS allowed.
2176 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2178 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2179 header file does not have a version number, so I've had to invent a new
2180 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2181 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2182 radiusclient), but it was contributed by a Radius user.
2184 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2185 files or directories, for OpenSSL.
2187 PH/19 When an Exim process that is running as root has to create an Exim log
2188 file, it does so in a subprocess that runs as exim:exim so as to get the
2189 ownership right at creation (otherwise, other Exim processes might see
2190 the file with the wrong ownership). There was no test for failure of this
2191 fork() call, which would lead to the process getting stuck as it waited
2192 for a non-existent subprocess. Forks do occasionally fail when resources
2193 run out. I reviewed all the other calls to fork(); they all seem to check
2196 PH/20 When checking for unexpected SMTP input at connect time (before writing
2197 the banner), Exim was not dealing correctly with a non-positive return
2198 from the read() function. If the client had disconnected by this time,
2199 the result was a log entry for a synchronization error with an empty
2200 string after "input=" when read() returned zero. If read() returned -1
2201 (an event I could not check), uninitialized data bytes were printed.
2202 There were reports of junk text (parts of files, etc) appearing after
2205 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2207 PH/22 Added support for macro redefinition, and (re)definition in between
2208 driver and ACL definitions.
2210 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2211 forgetting to use the resulting value; it was using the unexpanded value.
2213 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2214 hadn't been configured. The fix is from Juergen Kreileder, who
2215 understands it better than I do:
2217 "Here's what I see happening with three configured cyrus_sasl
2218 authenticators configured (plain, login, cram-md5):
2220 On startup auth_cyrus_sasl_init() gets called for each of these.
2221 This means three calls to sasl_listmech() without a specified mech_list.
2222 => SASL tests which mechs of all available mechs actually work
2223 => three warnings about OTP not working
2224 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2226 With the patch, sasl_listmech() also gets called three times. But now
2227 SASL's mech_list option is set to the server_mech specified in the the
2228 authenticator. Or in other words, the answer from sasl_listmech()
2229 gets limited to just the mech you're testing for (which is different
2231 => the return list contains just 'plain' or 'login', 'cram-md5' or
2232 nothing depending on the value of ob->server_mech.
2234 I've just tested the patch: Authentication still works fine,
2235 unavailable mechs specified in the exim configuration are still
2236 caught, and the auth.log warnings about OTP are gone."
2238 PH/25 When debugging is enabled, the contents of the command line are added
2239 to the debugging output, even when log_selector=+arguments is not
2242 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2243 answer is "GNU", and only if the return is "GNU/something" is the answer
2246 PH/27 $acl_verify_message is now set immediately after the failure of a
2247 verification in an ACL, and so is available in subsequent modifiers. In
2248 particular, the message can be preserved by coding like this:
2250 warn !verify = sender
2251 set acl_m0 = $acl_verify_message
2253 Previously, $acl_verify_message was set only while expanding "message"
2254 and "log_message" when a very denied access.
2256 PH/28 Modified OS/os.c-Linux with
2258 -#ifndef OS_LOAD_AVERAGE
2259 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2261 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2262 nomenclature these days.)
2264 PH/29 Installed patch from the Sieve maintainer that adds the options
2265 sieve_useraddress and sieve_subaddress to the redirect router.
2267 PH/30 In these circumstances:
2268 . Two addresses routed to the same list of hosts;
2269 . First host does not offer TLS;
2270 . First host accepts first address;
2271 . First host gives temporary error to second address;
2272 . Second host offers TLS and a TLS session is established;
2273 . Second host accepts second address.
2274 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2275 and peerdn, if requested) that were in fact used only for the second
2278 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2279 attention to any local part prefix or suffix that was matched by the
2280 router that accepted the address. It now behaves in the same way as it
2281 does for delivery: the affixes are removed from the local part unless
2282 rcpt_include_affixes is set on the transport.
2284 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2285 timeout during the DATA phase of an incoming message.
2287 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2288 applied a patch sanctioned by the Sieve maintainer.
2290 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2291 the uid or gid is negative. A case of a negative gid caused this to be
2292 noticed. The fix allows for either to be negative.
2294 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2295 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2298 PH/36 The change PH/12 above was broken. Fixed it.
2300 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2301 the grounds that routing the same address twice would always produce the
2302 same answer. This might have been true once, but it is certainly no
2303 longer true now. Routing a child address may depend on the previous
2304 routing that produced that child. Some complicated redirection strategies
2305 went wrong when messages had multiple recipients, and made Exim's
2306 behaviour dependent on the order in which the addresses were given.
2308 I have moved the duplicate checking until after the routing is complete.
2309 Exim scans the addresses that are assigned to local and remote
2310 transports, and removes any duplicates. This means that more work will be
2311 done, as duplicates will always all be routed, but duplicates are
2312 presumably rare, so I don't expect this is of any significance.
2314 For deliveries to pipes, files, and autoreplies, the duplicate checking
2315 still happens during the routing process, since they are not going to be
2318 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2319 It corrects a timeout issue with spamd. This is Ian's comment: "The
2320 background is that sometimes spamd either never reads data from a
2321 connection it has accepted, or it never writes response data. The exiscan
2322 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2323 blindly assumes that writes won't block so it may never time out."
2325 PH/39 Allow G after quota size as well as K and M.
2327 PH/40 The value set for $authenticated_id in an authenticator may not contain
2328 binary zeroes or newlines because the value is written to log lines and
2329 to spool files. There was no check on this. Now the value is run through
2330 the string_printing() function so that such characters are converted to
2331 printable escape sequences.
2333 PH/41 $message_linecount is a new variable that contains the total number of
2334 lines in the message. Compare $body_linecount, which is the count for the
2337 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2338 bounce and delay warning messages, except in certain special cases, which
2341 (a) An SMTP error message from a remote host;
2342 (b) A message specified in a :fail: redirection;
2343 (c) A message specified in a "fail" command in a system filter;
2344 (d) A message specified in a FAIL return from the queryprogram router;
2345 (e) A message specified by the cannot_route_message router option.
2347 In these cases only, Exim does include the error details in bounce and
2348 warning messages. There are also a few cases where bland messages such
2349 as "unrouteable address" or "local delivery error" are given.
2351 PH/43 $value is now also set for the "else" part of a ${run expansion.
2353 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2354 being worked on, but at least Exim now implements the latest version to
2357 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2358 process to complete was treated as a delivery failure, a timeout while
2359 writing the message to the pipe was logged, but erroneously treated as a
2360 successful delivery. Such timeouts include transport filter timeouts. For
2361 consistency with the overall process timeout, these timeouts are now
2362 treated as errors, giving rise to delivery failures by default. However,
2363 there is now a new Boolean option for the pipe transport called
2364 timeout_defer, which, if set TRUE, converts the failures into defers for
2365 both kinds of timeout. A transport filter timeout is now identified in
2368 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2369 systems where "make" and "gmake" are different, calling "gmake" at top
2370 level broke things. I've arranged for the value of $(MAKE) to be passed
2371 from the Makefile to this script so that it can call the same version of
2375 A note about Exim versions 4.44 and 4.50
2376 ----------------------------------------
2378 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2379 changes of various kinds. As a consequence, a big documentation update was
2380 needed. This delayed the release for rather longer than seemed good, especially
2381 in the light of a couple of (minor) security issues. Therefore, the changes
2382 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2383 release. So 4.44 and 4.50 are in effect two different branches that both start
2386 I have left the 4.50 change log unchanged; it contains all the changes since
2387 4.43. The change log for 4.44 is below; many of its items are identical to
2388 those for 4.50. This seems to be the most sensible way to preserve the
2389 historical information.
2395 1. Minor wording change to the doc/README.SIEVE file.
2397 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2398 computation of the current number of files was incorrect.
2400 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2401 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2404 4. Give more explanation in the error message when the command for a transport
2405 filter fails to execute.
2407 5. There are several places where Exim runs a non-Exim command in a
2408 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2409 was being done only for the command run by the queryprogram router. It is
2410 now done for all such subprocesses. The other cases are: ${run, transport
2411 filters, and the commands run by the lmtp and pipe transports.
2413 6. Added CONFIGURE_GROUP build-time option.
2415 7. Some older OS have a limit of 256 on the maximum number of file
2416 descriptors. Exim was using setrlimit() to set 1000 as a large value
2417 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2418 systems. I've change it so that if it can't get 1000, it tries for 256.
2420 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2421 was an oversight, and furthermore, ever since the addition of extra
2422 controls (e.g. 4.43/32), the checks on when to allow different forms of
2423 "control" were broken. There should now be diagnostics for all cases when a
2424 control that does not make sense is encountered.
2426 9. Added the /retain_sender option to "control=submission".
2428 10. $recipients is now available in the predata ACL (oversight).
2430 11. Tidy the search cache before the fork to do a delivery from a message
2431 received from the command line. Otherwise the child will trigger a lookup
2432 failure and thereby defer the delivery if it tries to use (for example) a
2433 cached ldap connection that the parent has called unbind on.
2435 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2436 of $address_data from the recipient verification was clobbered by the
2437 sender verification.
2439 13. The value of address_data from a sender verification is now available in
2440 $sender_address_data in subsequent conditions in the ACL statement.
2442 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2444 15. Added a new option "connect=<time>" to callout options, to set a different
2447 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2448 was its contents. (It was OK if the option was not defined at all.)
2450 17. A "Completed" log line is now written for messages that are removed from
2451 the spool by the -Mrm option.
2453 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2454 information about exactly what failed.
2456 19. Added -dd to debug only the daemon process.
2458 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2459 handles timeouts, both on the server side and network timeouts. Renamed the
2460 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2462 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2463 It is now set to "smtps".
2465 22. $host_address is now set to the target address during the checking of
2466 ignore_target_hosts.
2468 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2469 being passed; this would have caused $sender_host_name to have been used if
2470 matching the list had actually called for a host name (not very likely,
2471 since this list is usually IP addresses). A host name is now passed as
2474 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2475 code that specifies a non-restarting handler (typically sigaction() in
2476 modern systems) in an attempt to fix a rare and obscure crash bug.
2478 25. Narrowed the window for a race in the daemon that could cause it to ignore
2479 SIGCHLD signals. This is not a major problem, because they are used only to
2480 wake it up if nothing else does.
2482 26. A malformed maildirsize file could cause Exim to calculate negative values
2483 for the mailbox size or file count. Odd effects could occur as a result.
2484 The maildirsize information is now recalculated if the size or filecount
2487 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2488 support for a long time. Removed HAVE_SYS_VFS_H.
2490 28. Installed the latest version of exipick from John Jetmore.
2492 29. In an address list, if the pattern was not a regular expression, an empty
2493 subject address (from a bounce message) matched only if the pattern was an
2494 empty string. Non-empty patterns were not even tested. This was the wrong
2495 because it is perfectly reasonable to use an empty address as part of a
2496 database query. An empty address is now tested by patterns that are
2497 lookups. However, all the other forms of pattern expect the subject to
2498 contain a local part and a domain, and therefore, for them, an empty
2499 address still always fails if the pattern is not itself empty.
2501 30. Exim went into a mad DNS loop when attempting to do a callout where the
2502 host was specified on an smtp transport, and looking it up yielded more
2503 than one IP address.
2505 31. Re-factored the code for checking spool and log partition space into a
2506 function that finds that data and another that does the check. The former
2507 is then used to implement four new variables: $spool_space, $log_space,
2508 $spool_inodes, and $log_inodes.
2510 32. The RFC2047 encoding function was originally intended for short strings
2511 such as real names; it was not keeping to the 75-character limit for
2512 encoded words that the RFC imposes. It now respects the limit, and
2513 generates multiple encoded words if necessary. To be on the safe side, I
2514 have increased the buffer size for the ${rfc2047: expansion operator from
2517 33. It is now permitted to omit both strings after an "if" condition; if the
2518 condition is true, the result is "true". As before, when the second string
2519 is omitted, a false condition yields an empty string. This makes it less
2520 cumbersome to write custom ACL and router conditions.
2522 34. Failure to deliver a bounce message always caused it to be frozen, even if
2523 there was an errors_to setting on the router. The errors_to setting is now
2526 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2527 canonical form (fully expanded) before being placed in
2528 $sender_host_address.
2530 36. The table in the code that translates DNS record types into text (T_A to
2531 "A" for instance) was missing entries for NS and CNAME. It is just possible
2532 that this could have caused confusion if both these types were looked up
2533 for the same domain, because the text type is used as part of Exim's
2534 per-process caching. But the chance of anyone hitting this buglet seems
2537 37. The dnsdb lookup has been extended in a number of ways.
2539 (1) There is a new type, "zns", which walks up the domain tree until it
2540 finds some nameserver records. It should be used with care.
2542 (2) There is a new type, "mxh", which is like "mx" except that it returns
2543 just the host names, not the priorities.
2545 (3) It is now possible to give a list of domains (or IP addresses) to be
2546 looked up. The behaviour when one of the lookups defers can be
2547 controlled by a keyword.
2549 (4) It is now possible to specify the separator character for use when
2550 multiple records are returned.
2552 38. The dnslists ACL condition has been extended: it is now possible to supply
2553 a list of IP addresses and/or domains to be looked up in a particular DNS
2556 39. Added log_selector=+queue_time_overall.
2558 40. When running the queue in the test harness, wait just a tad after forking a
2559 delivery process, to get repeatability of debugging output.
2561 41. Include certificate and key file names in error message when GnuTLS fails
2562 to set them up, because the GnuTLS error message doesn't include the name
2563 of the failing file when there is a problem reading it.
2565 42. Allow both -bf and -bF in the same test run.
2567 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2569 44. The "Exiscan patch" is now merged into the mainline Exim source.
2571 45. Sometimes the final signoff response after QUIT could fail to get
2572 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2573 < 0 before doing a fflush(). This bug looks as though it goes back to the
2574 introduction of TLS in release 3.20, but "sometimes" must have been rare
2575 because the tests only now provoked it.
2577 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2578 (this can affect the format of dates).
2580 47. exim_tidydb, when checking for the continued existence of a message for
2581 which it has found a message-specific retry record, was not finding
2582 messages that were in split spool directories. Consequently, it was
2583 deleting retry records that should have stayed in existence.
2585 48. Steve fixed some bugs in eximstats.
2587 49. The SPA authentication driver was not abandoning authentication and moving
2588 on to the next authenticator when an expansion was forced to fail,
2589 contradicting the general specification for all authenticators. Instead it
2590 was generating a temporary error. It now behaves as specified.
2592 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2593 (the order specifies the preference for clients). The order is now AES256,
2594 AES128, 3DES, ARCFOUR128.
2596 51. Small patch to Sieve code - explicitly set From: when generating an
2599 52. Exim crashed if a remote delivery caused a very long error message to be
2600 recorded - for instance if somebody sent an entire SpamAssassin report back
2601 as a large number of 550 error lines. This bug was coincidentally fixed by
2602 increasing the size of one of Exim's internal buffers (big_buffer) that
2603 happened as part of the Exiscan merge. However, to be on the safe side, I
2604 have made the code more robust (and fixed the comments that describe what
2607 53. Now that there can be additional text after "Completed" in log lines (if
2608 the queue_time_overall log selector is set), a one-byte patch to exigrep
2609 was needed to allow it to recognize "Completed" as not the last thing in
2612 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2613 patch that reportedly fixes this has been added. I am not expert enough to
2614 create a test for it. This is what the patch creator wrote:
2616 "I found a little strange behaviour of ldap code when working with
2617 Windows 2003 AD Domain, where users was placed in more than one
2618 Organization Units. When I tried to give exim partial DN, the exit code
2619 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2620 But simultaneously result of request was absolutely normal ldap result,
2621 so I produce this patch..."
2623 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2624 REFERENCE, so I have modified the code to exclude the patch when that macro
2627 55. Some experimental protocols are using DNS PTR records for new purposes. The
2628 keys for these records are domain names, not reversed IP addresses. The
2629 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2630 leaves it alone. Component reversal etc. now happens only for IP addresses.
2633 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2635 57. Double the size of the debug message buffer (to 2048) so that more of very
2636 long debug lines gets shown.
2638 58. The exicyclog utility now does better if the number of log files to keep
2639 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2641 59. Two changes related to the smtp_active_hostname option:
2643 (1) $smtp_active_hostname is now available as a variable.
2644 (2) The default for smtp_banner uses $smtp_active_hostname instead
2645 of $primary_hostname.
2647 60. The host_aton() function is supposed to be passed a string that is known
2648 to be a valid IP address. However, in the case of IPv6 addresses, it was
2649 not checking this. This is a hostage to fortune. Exim now panics and dies
2650 if the condition is not met. A case was found where this could be provoked
2651 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2652 components; fortuitously, this particular loophole had already been fixed
2653 by change 4.50/55 above.
2655 If there are any other similar loopholes, the new check in host_aton()
2656 itself should stop them being exploited. The report I received stated that
2657 data on the command line could provoke the exploit when Exim was running as
2658 exim, but did not say which command line option was involved. All I could
2659 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2660 running as the user.
2663 61. There was a buffer overflow vulnerability in the SPA authentication code
2664 (which came originally from the Samba project). I have added a test to the
2665 spa_base64_to_bits() function which I hope fixes it.
2668 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2669 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2671 63. The daemon start-up calls getloadavg() while still root for those OS that
2672 need the first call to be done as root, but it missed one case: when
2673 deliver_queue_load_max is set with deliver_drop_privilege. This is
2674 necessary for the benefit of the queue runner, because there is no re-exec
2675 when deliver_drop_privilege is set.
2677 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2678 This has been fixed.
2680 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2681 was in use, was not putting the data itself into the right store pool;
2682 consequently, it could be overwritten for a subsequent message in the same
2683 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2686 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2688 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2689 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2690 the function were treating the return as a boolean value, which happened to
2691 work because 0=false and not-0=true, but is not correct code.
2693 68. The host_aton() function was not handling scoped IPv6 addresses (those
2694 with, for example, "%eth0" on the end) correctly.
2696 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2697 negated items (that is, ~something) in unsigned ints. Some compilers
2698 apparently mutter when there is no cast.
2700 70. If an address verification called from an ACL failed, and did not produce a
2701 user-specific message (i.e. there was only a "system" message), nothing was
2702 put in $acl_verify_message. In this situation, it now puts the system
2705 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2706 session; change 4.31/43 added the unwanted input to the log line - except
2707 that it did not do this in the start of session case. It now does.
2709 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2710 This is wrong and can cause the other end to generate a synchronization
2711 error if it is another Exim or anything else that does the synchronization
2712 check. A QUIT command is no longer sent after a timeout.
2714 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2715 during host lookups.
2717 74. The defer_ok option of callout verification was not working if it was used
2718 when verifying addresses in header lines, that is, for this case:
2720 verify = header_sender/callout=defer_ok
2722 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2723 those file descriptors could be used for SMTP connections. If anything
2724 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2725 could be sent to the SMTP client, causing chaos. The daemon now opens
2726 stdin, stdout, and stderr to /dev/null when it puts itself into the
2729 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2730 log by default. The user can override this with suitable Perl magic.
2732 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2733 the log message when discard triggers, was not working for the DATA ACL or
2734 for the non-SMTP ACL.
2736 78. Error message wording change in sieve.c.
2738 79. If smtp_accept_max_per_host was set, the number of connections could be
2739 restricted to fewer than expected, because the daemon was trying to set up
2740 a new connection before checking whether the processes handling previous
2741 connections had finished. The check for completed processes is now done
2742 earlier. On busy systems, this bug wouldn't be noticed because something
2743 else would have woken the daemon, and it would have reaped the completed
2746 80. If a message was submitted locally by a user whose login name contained one
2747 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2748 It caused a spool format error. I have fixed the spool reading code. A
2749 related problem was that the "from" clause in the Received: line became
2750 illegal because of the space(s). It is now covered by ${quote_local_part.
2752 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2755 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2757 83. Patch from Sieve maintainer to fix unterminated string problem in
2758 "vacation" handling.
2760 84. Some minor changes to the Linux configuration files to help with other
2761 OS variants using glibc.
2763 85. One more patch for Sieve to update vacation handling to latest spec.
2766 ----------------------------------------------------
2767 See the note above about the 4.44 and 4.50 releases.
2768 ----------------------------------------------------
2774 1. Change 4.43/35 introduced a bug that caused file counts to be
2775 incorrectly computed when quota_filecount was set in an appendfile
2778 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2779 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2782 3. Give more explanation in the error message when the command for a transport
2783 filter fails to execute.
2785 4. There are several places where Exim runs a non-Exim command in a
2786 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2787 was being done only for the command run by the queryprogram router. It is
2788 now done for all such subprocesses. The other cases are: ${run, transport
2789 filters, and the commands run by the lmtp and pipe transports.
2791 5. Some older OS have a limit of 256 on the maximum number of file
2792 descriptors. Exim was using setrlimit() to set 1000 as a large value
2793 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2794 systems. I've change it so that if it can't get 1000, it tries for 256.
2796 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2797 was an oversight, and furthermore, ever since the addition of extra
2798 controls (e.g. 4.43/32), the checks on when to allow different forms of
2799 "control" were broken. There should now be diagnostics for all cases when a
2800 control that does not make sense is encountered.
2802 7. $recipients is now available in the predata ACL (oversight).
2804 8. Tidy the search cache before the fork to do a delivery from a message
2805 received from the command line. Otherwise the child will trigger a lookup
2806 failure and thereby defer the delivery if it tries to use (for example) a
2807 cached ldap connection that the parent has called unbind on.
2809 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2810 of $address_data from the recipient verification was clobbered by the
2811 sender verification.
2813 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2814 was its contents. (It was OK if the option was not defined at all.)
2816 11. A "Completed" log line is now written for messages that are removed from
2817 the spool by the -Mrm option.
2819 12. $host_address is now set to the target address during the checking of
2820 ignore_target_hosts.
2822 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2823 being passed; this would have caused $sender_host_name to have been used if
2824 matching the list had actually called for a host name (not very likely,
2825 since this list is usually IP addresses). A host name is now passed as
2828 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2829 code that specifies a non-restarting handler (typically sigaction() in
2830 modern systems) in an attempt to fix a rare and obscure crash bug.
2832 15. Narrowed the window for a race in the daemon that could cause it to ignore
2833 SIGCHLD signals. This is not a major problem, because they are used only to
2834 wake it up if nothing else does.
2836 16. A malformed maildirsize file could cause Exim to calculate negative values
2837 for the mailbox size or file count. Odd effects could occur as a result.
2838 The maildirsize information is now recalculated if the size or filecount
2841 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2842 support for a long time. Removed HAVE_SYS_VFS_H.
2844 18. Updated exipick to current release from John Jetmore.
2846 19. Allow an empty sender to be matched against a lookup in an address list.
2847 Previously the only cases considered were a regular expression, or an
2850 20. Exim went into a mad DNS lookup loop when doing a callout where the
2851 host was specified on the transport, if the DNS lookup yielded more than
2854 21. The RFC2047 encoding function was originally intended for short strings
2855 such as real names; it was not keeping to the 75-character limit for
2856 encoded words that the RFC imposes. It now respects the limit, and
2857 generates multiple encoded words if necessary. To be on the safe side, I
2858 have increased the buffer size for the ${rfc2047: expansion operator from
2861 22. Failure to deliver a bounce message always caused it to be frozen, even if
2862 there was an errors_to setting on the router. The errors_to setting is now
2865 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2866 canonical form (fully expanded) before being placed in
2867 $sender_host_address.
2869 24. Updated eximstats to version 1.33
2871 25. Include certificate and key file names in error message when GnuTLS fails
2872 to set them up, because the GnuTLS error message doesn't include the name
2873 of the failing file when there is a problem reading it.
2875 26. Expand error message when OpenSSL has problems setting up cert/key files.
2878 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2879 (this can affect the format of dates).
2881 28. exim_tidydb, when checking for the continued existence of a message for
2882 which it has found a message-specific retry record, was not finding
2883 messages that were in split spool directories. Consequently, it was
2884 deleting retry records that should have stayed in existence.
2886 29. eximstats updated to version 1.35
2887 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2888 1.35 - bugfix such that pie charts by volume are generated correctly
2890 30. The SPA authentication driver was not abandoning authentication and moving
2891 on to the next authenticator when an expansion was forced to fail,
2892 contradicting the general specification for all authenticators. Instead it
2893 was generating a temporary error. It now behaves as specified.
2895 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2896 (the order specifies the preference for clients). The order is now AES256,
2897 AES128, 3DES, ARCFOUR128.
2899 31. Small patch to Sieve code - explicitly set From: when generating an
2902 32. Exim crashed if a remote delivery caused a very long error message to be
2903 recorded - for instance if somebody sent an entire SpamAssassin report back
2904 as a large number of 550 error lines. This bug was coincidentally fixed by
2905 increasing the size of one of Exim's internal buffers (big_buffer) that
2906 happened as part of the Exiscan merge. However, to be on the safe side, I
2907 have made the code more robust (and fixed the comments that describe what
2910 33. Some experimental protocols are using DNS PTR records for new purposes. The
2911 keys for these records are domain names, not reversed IP addresses. The
2912 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2913 leaves it alone. Component reversal etc. now happens only for IP addresses.
2916 34. The host_aton() function is supposed to be passed a string that is known
2917 to be a valid IP address. However, in the case of IPv6 addresses, it was
2918 not checking this. This is a hostage to fortune. Exim now panics and dies
2919 if the condition is not met. A case was found where this could be provoked
2920 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2921 components; fortuitously, this particular loophole had already been fixed
2922 by change 4.50/55 or 4.44/33 above.
2924 If there are any other similar loopholes, the new check in host_aton()
2925 itself should stop them being exploited. The report I received stated that
2926 data on the command line could provoke the exploit when Exim was running as
2927 exim, but did not say which command line option was involved. All I could
2928 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2929 running as the user.
2932 35. There was a buffer overflow vulnerability in the SPA authentication code
2933 (which came originally from the Samba project). I have added a test to the
2934 spa_base64_to_bits() function which I hope fixes it.
2937 36. The daemon start-up calls getloadavg() while still root for those OS that
2938 need the first call to be done as root, but it missed one case: when
2939 deliver_queue_load_max is set with deliver_drop_privilege. This is
2940 necessary for the benefit of the queue runner, because there is no re-exec
2941 when deliver_drop_privilege is set.
2943 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2944 was in use, was not putting the data itself into the right store pool;
2945 consequently, it could be overwritten for a subsequent message in the same
2946 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2949 38. Sometimes the final signoff response after QUIT could fail to get
2950 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2951 < 0 before doing a fflush(). This bug looks as though it goes back to the
2952 introduction of TLS in release 3.20, but "sometimes" must have been rare
2953 because the tests only now provoked it.
2959 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2960 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2961 Now it returns an integer. A number of calls were still expecting a T/F
2962 return. Fortuitously, in all cases, the tests worked in OK situations,
2963 which is the norm. However, things would have gone wrong on any write
2964 failures on the smtp file descriptor. This function is used when sending
2965 messages over SMTP and also when doing verify callouts.
2967 2. When Exim is called to do synchronous delivery of a locally submitted
2968 message (the -odf or -odi options), it no longer closes stderr before doing
2971 3. Implemented the mua_wrapper option.
2973 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2975 5. Implemented the functions header_remove(), header_testname(),
2976 header_add_at_position(), and receive_remove_recipient(), and exported them
2979 6. If an ACL "warn" statement specified the addition of headers, Exim already
2980 inserted X-ACL-Warn: at the start if there was no header name. However, it
2981 was not making this test for the second and subsequent header lines if
2982 there were newlines in the string. This meant that an invalid header could
2983 be inserted if Exim was badly configured.
2985 7. Allow an ACL "warn" statement to add header lines at the start or after all
2986 the Received: headers, as well as at the end.
2988 8. Added the rcpt_4xx retry error code.
2990 9. Added postmaster_mailfrom=xxx to callout verification option.
2992 10. Added mailfrom=xxxx to the callout verification option, for verify=
2995 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2996 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2998 12. Inserted some casts to stop certain compilers warning when using pointer
2999 differences as field lengths or precisions in printf-type calls (mostly
3000 affecting debugging statements).
3002 13. Added optional readline() support for -be (dynamically loaded).
3004 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3005 same clock tick as a message's arrival, so that its received time was the
3006 same as the "first fail" time on the retry record, and that message
3007 remained on the queue past the ultimate address timeout, every queue runner
3008 would try a delivery (because it was past the ultimate address timeout) but
3009 after another failure, the ultimate address timeout, which should have then
3010 bounced the address, did not kick in. This was a "< instead of <=" error;
3011 in most cases the first failure would have been in the next clock tick
3012 after the received time, and all would be well.
3014 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3015 being recognized when the domain list was tested by the match_domain
3016 condition in an expansion string.
3018 16. Added the ${str2b64: operator.
3020 17. Exim was always calling setrlimit() to set a large limit for the number of
3021 processes, without checking whether the existing limit was already
3022 adequate. (It did check for the limit on file descriptors.) Furthermore,
3023 errors from getrlimit() and setrlimit() were being ignored. Now they are
3024 logged to the main and panic logs, but Exim does carry on, to try to do its
3025 job under whatever limits there are.
3027 18. Imported PCRE 5.0.
3029 19. Trivial typo in log message " temporarily refused connection" (the leading
3032 20. If the log selector return_path_on_delivery was set and an address was
3033 redirected to /dev/null, the delivery process crashed because it assumed
3034 that a return path would always be set for a "successful" delivery. In this
3035 case, the whole delivery is bypassed as an optimization, and therefore no
3038 21. Internal re-arrangement: the function for sending a challenge and reading
3039 a response while authentication was assuming a zero-terminated challenge
3040 string. It's now changed to take a pointer and a length, to allow for
3041 binary data in such strings.
3043 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3045 23. Exim was not respecting finduser_retries when seeking the login of the
3046 uid under which it was called; it was always trying 10 times. (The default
3047 setting of finduser_retries is zero.) Also, it was sleeping after the final
3048 failure, which is pointless.
3050 24. Implemented tls_on_connect_ports.
3052 25. Implemented acl_smtp_predata.
3054 26. If the domain in control=submission is set empty, Exim assumes that the
3055 authenticated id is a complete email address when it generates From: or
3056 Sender: header lines.
3058 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3059 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3060 chown and chgrp in /bin and hostname in /usr/bin.
3062 28. Exim was keeping the "process log" file open after each use, just as it
3063 does for the main log. This opens the possibility of it remaining open for
3064 long periods when the USR1 signal hits a daemon. Occasional processlog
3065 errors were reported, that could have been caused by this. Anyway, it seems
3066 much more sensible not to leave this file open at all, so that is what now
3069 29. The long-running daemon process does not normally write to the log once it
3070 has entered its main loop, and it closes the log before doing so. This is
3071 so that log files can straightforwardly be renamed and moved. However,
3072 there are a couple of unusual error situations where the daemon does write
3073 log entries, and I had neglected to close the log afterwards.
3075 30. The text of an SMTP error response that was received during a remote
3076 delivery was being truncated at 512 bytes. This is too short for some of
3077 the long messages that one sometimes sees. I've increased the limit to
3080 31. It is now possible to make retry rules that apply only when a message has a
3081 specific sender, in particular, an empty sender.
3083 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3084 it possible to be selective about when SMTP synchronization is enforced.
3086 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3088 32. Implemented hosts_connection_nolog.
3090 33. Added an ACL for QUIT.
3092 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3095 35. Added mailbox_size and mailbox_filecount to appendfile.
3097 36. Added control = no_multiline_responses to ACLs.
3099 37. There was a bug in the logic of the code that waits for the clock to tick
3100 in the case where the clock went backwards by a substantial amount such
3101 that the microsecond fraction of "now" was more than the microsecond
3102 fraction of "then" (but the whole seconds number was less).
3104 38. Added support for the libradius Radius client library this is found on
3105 FreeBSD (previously only the radiusclient library was supported).
3111 1. When certain lookups returned multiple values in the form name=value, the
3112 quoting of the values was not always being done properly. Specifically:
3113 (a) If the value started with a double quote, but contained no whitespace,
3115 (b) If the value contained whitespace other than a space character (i.e.
3116 tabs or newlines or carriage returns) it was not quoted.
3117 This fix has been applied to the mysql and pgsql lookups by writing a
3118 separate quoting function and calling it from the lookup code. The fix
3119 should probably also be applied to nisplus, ibase and oracle lookups, but
3120 since I cannot test any of those, I have not disturbed their existing code.
3122 2. A hit in the callout cache for a specific address caused a log line with no
3123 reason for rejecting RCPT. Now it says "Previous (cached) callout
3124 verification failure".
3126 3. There was an off-by-one bug in the queryprogram router. An over-long
3127 return line was truncated at 256 instead of 255 characters, thereby
3128 overflowing its buffer with the terminating zero. As well as fixing this, I
3129 have increased the buffer size to 1024 (and made a note to document this).
3131 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3132 when Exim is waiting for an SMTP response from a remote server, Exim
3133 restarts its select() call on the socket, thereby resetting its timeout.
3134 This is not a problem when such interrupts are rare. Somebody set up a cron
3135 job to run exiwhat every 2 minutes, which is less than the normal select()
3136 timeout (5 or 10 minutes). This meant that the select() timeout never
3137 kicked in because it was always reset. I have fixed this by comparing the
3138 time when an interrupt arrives with the time at the start of the first call
3139 to select(). If more time than the timeout has elapsed, the interrupt is
3140 treated as a timeout.
3142 5. Some internal re-factoring in preparation for the addition of Sieve
3143 extensions (by MH). In particular, the "personal" test is moved to a
3144 separate function, and given an option for scanning Cc: and Bcc: (which is
3145 not set for Exim filters).
3147 6. When Exim created an email address using the login of the caller as the
3148 local part (e.g. when creating a From: or Sender: header line), it was not
3149 quoting the local part when it contained special characters such as @.
3151 7. Installed new OpenBSD configuration files.
3153 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3154 try to make them clearer.
3156 9. Callout options, other than the timeout value, were being ignored when
3157 verifying sender addresses in header lines. For example, when using
3159 verify = header_sender/callout=no_cache
3161 the cache was (incorrectly) being used.
3163 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3164 only the Cygwin environment.
3166 11. When return_path_on_delivery was set as a log selector, if different remote
3167 addresses in the same message used different return paths and parallel
3168 remote delivery occurred, the wrong values would sometimes be logged.
3169 (Whenever a remote delivery process finished, the return path value from
3170 the most recently started remote delivery process was logged.)
3172 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3173 lines when AUTH and/or TLS are in use. This is the "received protocol"
3174 field. Exim used to use "asmtp" for authenticated SMTP, without any
3175 indication (in the protocol name) for TLS use. Now it follows the RFC and
3176 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3177 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3178 names appear in log lines as well as in Received: header lines.
3180 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3181 extensions, and comparison tests, and to fix some bugs.
3183 14. Changes to the "personal" filter test:
3185 (1) The test was buggy in that it was just doing the equivalent of
3186 "contains" tests on header lines. For example, if a user's address was
3187 anne@some.where, the "personal" test would incorrectly be true for
3189 To: susanne@some.where
3191 This test is now done by extracting each address from the header in turn,
3192 and checking the entire address. Other tests that are part of "personal"
3193 are now done using regular expressions (for example, to check local parts
3194 of addresses in From: header lines).
3196 (2) The list of non-personal local parts in From: addresses has been
3197 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3198 taken from the Sieve specification recommendations.
3200 (3) If the message contains any header line starting with "List-" it is
3201 treated as non-personal.
3203 (4) The test for "circular" in the Subject: header line has been removed
3204 because it now seems ill-conceived.
3206 15. Minor typos in src/EDITME comments corrected.
3208 16. Installed latest exipick from John Jetmore.
3210 17. If headers_add on a router specified a text string that was too long for
3211 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3212 of string_sprintf() is now avoided.
3214 18. $message_body_size was not set (it was always zero) when running the DATA
3215 ACL and the local_scan() function.
3217 19. For the "mail" command in an Exim filter, no default was being set for
3218 the once_repeat time, causing a random time value to be used if "once" was
3219 specified. (If the value happened to be <= 0, no repeat happened.) The
3220 default is now 0s, meaning "never repeat". The "vacation" command was OK
3221 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3222 (I found it when inspecting the code).
3224 20. There is now an overall timeout for performing a callout verification. It
3225 defaults to 4 times the callout timeout, which applies to individual SMTP
3226 commands during the callout. The overall timeout applies when there is more
3227 than one host that can be tried. The timeout is checked before trying the
3228 next host. This prevents very long delays if there are a large number of
3229 hosts and all are timing out (e.g. when the network connections are timing
3230 out). The value of the overall timeout can be changed by specifying an
3231 additional sub-option for "callout", called "maxwait". For example:
3233 verify = sender/callout=5s,maxwait=20s
3235 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3236 to the end before writing, but this should make it even safer).
3238 22. Exim was forgetting that it had advertised PIPELINING for the second and
3239 subsequent messages on an SMTP connection. It was also not resetting its
3240 memory on STARTTLS and an internal HELO.
3242 23. When Exim logs an SMTP synchronization error within a session, it now
3243 records whether PIPELINING has been advertised or not.
3245 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3246 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3247 rather than long int.
3249 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3251 26. Added the never_mail option to autoreply.
3257 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3258 crash if the getsockname() call failed; this can happen if a connection is
3259 closed very soon after it is established. The problem was simply in the
3260 order in which certain operations were done, causing Exim to try to write
3261 to the SMTP stream before it had set up the file descriptor. The bug has
3262 been fixed by making things happen in the correct order.
3268 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3269 before the connection was closed, thus losing the rejection response.
3271 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3272 some early Solaris releases, but causes trouble in current releases where
3273 socklen_t is defined.
3275 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3278 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3281 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3282 that releasing the top of it at the end releases what was used for sub-
3283 expansions (unless the block got too big). However, discard this block if
3284 the first thing is a variable or header, so that we can use its block when
3285 it is dynamic (useful for very large $message_headers, for example).
3287 6. Lookups now cache *every* query, not just the most recent. A new, separate
3288 store pool is used for this. It can be recovered when all lookup caches are
3289 flushed. Lookups now release memory at the end of their result strings.
3290 This has involved some general refactoring of the lookup sources.
3292 7. Some code has been added to the store_xxx() functions to reduce the amount
3293 of flapping under certain conditions.
3295 8. log_incoming_interface used to affect only the <= reception log lines. Now
3296 it causes the local interface and port to be added to several more SMTP log
3297 lines, for example "SMTP connection from", and rejection lines.
3299 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3301 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3303 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3304 could be overwritten at the end of the current message (or the start of a
3305 new message if it was set in a HELO ACL). The value is now preserved for
3306 the duration of the SMTP connection.
3308 12. If a transport had a headers_rewrite setting, and a matching header line
3309 contained an unqualified address, that address was qualified, even if it
3310 did not match any rewriting rules. The underlying bug was that the values
3311 of the flags that permit the existence of unqualified sender and recipient
3312 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3313 non-local messages, and by -bnq for local messages) were not being
3314 preserved with the message after it was received.
3316 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3317 "next input=" as part of the text comprising the host identity instead of
3318 the correct text. The code was using the same buffer for two different
3319 strings. However, depending on which order the printing function evaluated
3320 its arguments, the bug did not always show up. Under Linux, for example, my
3321 test suite worked just fine.
3323 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3324 doesn't work with some older versions of Perl. It has been changed to "my",
3325 which in any case is probably the better facility to use.
3327 15. A really picky compiler found some instances of statements for creating
3328 error messages that either had too many or two few arguments for the format
3331 16. The size of the buffer for calls to the DNS resolver has been increased
3332 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3333 for addresses that have a lot of PTR records. This alleviates a problem; it
3334 does not fully solve it.
3336 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3337 buffer now truncates the list and logs the incident, which is the same
3338 action as happens when Exim is looking up a host name and its aliases.
3339 Previously in this situation something unpredictable would happen;
3340 sometimes it was "internal error: store_reset failed".
3342 18. If a server dropped the connection unexpectedly when an Exim client was
3343 using GnuTLS and trying to read a response, the client delivery process
3344 crashed while trying to generate an error log message.
3346 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3347 string, for example:
3349 warn message = H1: something\nH2: something
3351 the text was added as a single header line from Exim's point of view
3352 though it ended up OK in the delivered message. However, searching for the
3353 second and subsequent header lines using $h_h2: did not work. This has been
3354 fixed. Similarly, if a system filter added multiple headers in this way,
3355 the routers could not see them.
3357 20. Expanded the error message when iplsearch is called with an invalid key to
3358 suggest using net-iplsearch in a host list.
3360 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3361 ACLs are no longer actually imposed (and a message to that effect is
3364 22. If a "gecos" field in a passwd entry contained escaped characters, in
3365 particular, if it contained a \" sequence, Exim got it wrong when building
3366 a From: or a Sender: header from that name. A second bug also caused
3367 incorrect handling when an unquoted " was present following a character
3368 that needed quoting.
3370 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3371 was not being matched caselessly.
3373 24. Arranged for all hyphens in the exim.8 source to be escaped with
3376 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3377 properly. Recipient callout cache records were still being keyed to include
3378 the sender, even when use_sender was set false. This led to far more
3379 callouts that were necessary. The sender is no longer included in the key
3380 when use_sender is false.
3382 26. Added "control = submission" modifier to ACLs.
3384 27. Added the ${base62d: operator to decode base 62 numbers.
3386 28. dnsdb lookups can now access SRV records.
3388 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3389 the configuration file.
3391 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3392 (-v) mode. This makes the output for a verbose queue run more intelligible.
3394 31. Added a use_postmaster feature to recipient callouts.
3396 32. Added the $body_zerocount variable, containing the number of binary zero
3397 bytes in the message body.
3399 33. The time of last modification of the "new" subdirectory is now used as the
3400 "mailbox time last read" when there is a quota error for a maildir
3403 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3405 35. Added +ignore_unknown as a special item in host lists.
3407 36. Code for decoding IPv6 addresses in host lists is now included, even if
3408 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3409 address was recognized as an IP address, but was then not correctly decoded
3410 into binary, causing unexpected and incorrect effects when compared with
3417 1. Very minor rewording of debugging text in manualroute to say "list of
3418 hosts" instead of "hostlist".
3420 2. If verify=header_syntax was set, and a header line with an unqualified
3421 address (no domain) and a large number of spaces between the end of the
3422 name and the colon was received, the reception process suffered a buffer
3423 overflow, and (when I tested it) crashed. This was caused by some obsolete
3424 code that should have been removed. The fix is to remove it!
3426 3. When running in the test harness, delay a bit after writing a bounce
3427 message to get a bit more predictability in the log output.
3429 4. Added a call to search_tidyup() just before forking a reception process. In
3430 theory, someone could use a lookup in the expansion of smtp_accept_max_
3431 per_host which, without the tidyup, could leave open a database connection.
3433 5. Added the variables $recipient_data and $sender_data which get set from a
3434 lookup success in an ACL "recipients" or "senders" condition, or a router
3435 "senders" option, similar to $domain_data and $local_part_data.
3437 6. Moved the writing of debug_print from before to after the "senders" test
3440 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3441 problems for message scanning, either using a data ACL, or using
3442 local_scan() because the Received: header was not generated till after they
3443 were called (in order to set the time as the time of reception completion).
3444 I have revised the way this works. The header is now generated after the
3445 body is received, but before the ACL or local_scan() are called. After they
3446 are run, the timestamp in the header is updated.
3452 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3453 before starting a queue runner without re-exec. This happened only when
3454 deliver_drop_privilege was set or when the Exim user was set to root. The
3455 effect of the bug was that timeouts during subsequent deliveries caused
3456 crashes instead of being properly handled. The handler is now left at its
3457 default (and expected) setting.
3459 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3460 message, again when deliver_drop_privilege is set or Exim is run as root.
3461 The bug described in (1) was not present in this case, but the tidying up
3462 of the other signals was missing. I have made the two cases consistent.
3464 3. The ignore_target_hosts setting on a manualroute router was being ignored
3465 for hosts that were looked up using the /MX notation.
3467 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3470 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3471 operated on the sender address. After changing the $sender_address to <>
3472 for the sender address verify, Exim was re-instated it as the original
3473 (before rewriting) address, but remembering that it had rewritten it, so it
3474 wasn't rewriting it again. This bug also had the effect of breaking the
3475 sender address verification caching when the sender address was rewritten.
3477 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3478 This has been changed so that if the ip literal address matches
3479 ignore_target_hosts, the router declines.
3481 7. Added expansion conditions match_domain, match_address, and match_local_
3482 part (NOT match_host).
3484 8. The placeholder for the Received: header didn't have a length field set.
3486 9. Added code to Exim itself and to exim_lock to test for a specific race
3487 condition that could lead to file corruption when using MBX delivery. The
3488 issue is with the lockfile that is created in /tmp. If this file is removed
3489 after a process has opened it but before that process has acquired a lock,
3490 there is the potential for a second process to recreate the file and also
3491 acquire a lock. This could lead to two Exim processes writing to the file
3492 at the same time. The added code performs the same test as UW imapd; it
3493 checks after acquiring the lock that its file descriptor still refers to
3494 the same named file.
3496 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3497 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3500 11. Added the smtp_active_hostname option. If used, this will typically be made
3501 to depend on the incoming interface address. Because $interface_address is
3502 not set up until the daemon has forked a reception process, error responses
3503 that can happen earlier (such as "too many connections") no longer contain
3506 12. If an expansion in a condition on a "warn" statement fails because a lookup
3507 defers, the "warn" statement is abandoned, and the next ACL statement is
3508 processed. Previously this caused the whole ACL to be aborted.
3510 13. Added the iplsearch lookup type.
3512 14. Added ident_timeout as a log selector.
3514 15. Added tls_certificate_verified as a log selector.
3516 16. Added a global option tls_require_ciphers (compare the smtp transport
3517 option of the same name). This controls incoming TLS connections.
3519 17. I finally figured out how to make tls_require_ciphers do a similar thing
3520 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3521 before starting the TLS session.
3523 18. Tabs are now shown as \t in -bP output.
3525 19. If the log selector return_path_on_delivery was set, Exim crashed when
3526 bouncing a message because it had too many Received: header lines.
3528 20. If two routers both had headers_remove settings, and the first one included
3529 a superfluous trailing colon, the final name in the first list and the
3530 first name in the second list were incorrectly joined into one item (with a
3531 colon in the middle).
3537 1. Added -C and -D options to the exinext utility, mainly to make it easier
3538 to include in the automated testing, but these could be helpful when
3539 multiple configurations are in use.
3541 2. The exinext utility was not formatting the output nicely when there was
3542 an alternate port involved in the retry record key, nor when there was a
3543 message id as well (for retries that were specific to a specific message
3544 and a specific host). It was also confused by IPv6 addresses, because of
3545 the additional colons they contain. I have fixed the IPv4 problem, and
3546 patched it up to do a reasonable job for IPv6.
3548 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3549 delivery, the log line now contains "pipelined" if PIPELINING was used.
3551 4. An SMTP transport process used to panic and die if the bind() call to set
3552 an explicit outgoing interface failed. This has been changed; it is now
3553 treated in the same way as a connect() failure.
3555 5. A reference to $sender_host_name in the part of a conditional expansion
3556 that was being skipped was still causing a DNS lookup. This no longer
3559 6. The def: expansion condition was not recognizing references to header lines
3560 that used bh_ and bheader_.
3562 7. Added the _cache feature to named lists.
3564 8. The code for checking quota_filecount in the appendfile transport was
3565 allowing one more file than it should have been.
3567 9. For compatibility with Sendmail, the command line option
3575 and sets the incoming protocol and host name (for trusted callers). The
3576 host name and its colon can be omitted when only the protocol is to be set.
3577 Note the Exim already has two private options, -pd and -ps, that refer to
3578 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3579 "s", but I don't think that's a major issue.
3581 10. A number of refactoring changes to the code, none of which should affect
3584 (a) The number of logging options was getting close to filling up the
3585 32-bit word that was used as a bit map. I have split them into two classes:
3586 those that are passed in the argument to log_write(), and those that are
3587 only ever tested independently outside of that function. These are now in
3588 separate 32-bit words, so there is plenty of room for expansion again.
3589 There is no change in the user interface or the logging behaviour.
3591 (b) When building, for example, log lines, the code previously used a
3592 macro that called string_cat() twice, in order to add two strings. This is
3593 not really sufficiently general. Furthermore, there was one instance where
3594 it was actually wrong because one of the argument was used twice, and in
3595 one call a function was used. (As it happened, calling the function twice
3596 did not affect the overall behaviour.) The macro has been replaced by a
3597 function that can join an arbitrary number of extra strings onto a growing
3600 (c) The code for expansion conditions now uses a table and a binary chop
3601 instead of a serial search (which was left over from when there were very
3602 few conditions). Also, it now recognizes conditions like "pam" even when
3603 the relevant support is not compiled in: a suitably worded error message is
3604 given if an attempt is made to use such a condition.
3606 11. Added ${time_interval:xxxxx}.
3608 12. A bug was causing one of the ddress fields not to be passed back correctly
3609 from remote delivery subprocesses. The field in question was not being
3610 subsequently used, so this caused to problems in practice.
3612 13. Added new log selectors queue_time and deliver_time.
3614 14. Might have fixed a bug in maildirsizefile handling that threw up
3615 "unexpected character" debug warnings, and recalculated the data
3616 unnecessarily. In any case, I expanded the warning message to give more
3619 15. Added the message "Restricted characters in address" to the statements in
3620 the default ACL that block characters like @ and % in local parts.
3622 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3623 Three changes have been made:
3625 (a) There was a serious bug; a negative response to MAIL caused the whole
3626 recipient domain to be cached as invalid, thereby blocking all messages
3627 to all local parts at the same domain, from all senders. This bug has
3628 been fixed. The domain is no longer cached after a negative response to
3629 MAIL if the sender used is not empty.
3631 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3634 (c) A new callout option, "use_sender" has been added for people who want
3635 the modified behaviour.
3641 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3644 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3645 indeed breaks things for older releases.
3647 3. Added additional logging to the case where there is a problem reading data
3648 from a filter that is running in a subprocess using a pipe, in order to
3649 try to track down a specific problem.
3651 4. Testing facility fudge: when running in the test harness and attempting
3652 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3653 getting "No route to host". Convert this to a timeout.
3655 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3658 6. Some OS don't have socklen_t but use size_t instead. This affects the
3659 fifth argument of getsockopt() amongst other things. This is now
3660 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3661 can be set for individual OS. I have set it for SunOS5, OSF1, and
3662 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3663 some earlier ones do not.
3665 7. Change 4.30/15 was not doing the test caselessly.
3667 8. The standard form for an IPv6 address literal was being rejected by address
3668 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3669 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3670 this, as well as the form without the "IPv6" on the front (but only when
3671 address literals are enabled, of course).
3673 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3675 10. Exim crashed if a message with an empty sender address specified by -f
3676 encountered a router with an errors_to setting. This could be provoked only
3677 by a command such as
3681 where an empty string was supplied; "<>" did not hit this bug.
3683 11. Installed PCRE release 4.5.
3685 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3686 remained set. It is now erased.
3688 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3689 times from message ids (which are base 36 rather than the normal 62).
3691 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3692 were being counted as actual protocol errors, and logged if the log
3693 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3694 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3695 and DATA following a set of rejected RCPTs do not count as protocol errors.
3696 In other words, Exim assumes they were pipelined, though this may not
3697 actually be the case. Of course, in all cases the client gets an
3698 appropriate error code.
3700 15. If a lookup fails in an ACL condition, a message about the failure may
3701 be available; it is used if testing the ACL cannot continue, because most
3702 such messages specify what the cause of the deferral is. However, some
3703 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3704 that caused an old message to be retained and used if a later statement
3705 caused a defer, replacing the real cause of the deferral.
3707 16. If an IP address had so many PTR records that the DNS lookup buffer
3708 was not large enough to hold them, Exim could crash while trying to process
3709 the truncated data. It now detects and logs this case.
3711 17. Further to 4.21/58, another change has been made: if (and only if) the
3712 first line of a message (the first header line) ends with CRLF, a bare LF
3713 in a subsequent header line has a space inserted after it, so as not to
3714 terminate the header.
3716 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3717 unnecessarily, used atoi() instead of strtol(), and didn't check the
3718 termination when getting file sizes from file names by regex.
3720 19. Completely re-implemented the support for maildirsize files, in the light
3721 of a number of problems with the previous contributed implementation
3722 (4.30/29). In particular:
3724 . If the quota is zero, the maildirsize file is maintained, but no quota is
3727 . If the maildir directory does not exist, it is created before any attempt
3728 to write a maildirsize file.
3730 . The quota value in the file is just a cache; if the quota is changed in
3731 the transport, the new value overrides.
3733 . A regular expression is available for excluding directories from the
3736 20. The autoreply transport checks the characters in options that define the
3737 message's headers; it allows continued headers, but it was checking with
3738 isspace() after an embedded newline instead of explicitly looking for a
3741 21. If all the "regular" hosts to which an address was routed had passed their
3742 expiry times, and had not reached their retry times, the address was
3743 bounced, even if fallback hosts were defined. Now Exim should go on to try
3746 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3747 equivalent code in the SMTP transport. Some hosts send humungous responses
3748 to HELO/EHLO, more than 1024 it seems.
3750 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3751 gives compiler warnings in some environments. I've now done it "properly",
3754 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3755 (because of problems with the built-in one) was declared to return uschar *
3756 instead of char *, causing compiler failure.
3758 25. Fixed a file descriptor leak when processing alias/forward files.
3760 26. Fixed a minor format string issue in dbfn.c.
3762 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3764 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3765 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3768 29. When a sender address is verified, it is cached, to save repeating the test
3769 when there is more than one recipient in a message. However, when the
3770 verification involves a callout, it is possible for different callout
3771 options to be set for different recipients. It is too complicated to keep
3772 track of this in the cache, so now Exim always runs a verification when a
3773 callout is required, relying on the callout cache for the optimization.
3774 The overhead is duplication of the address routing, but this should not be
3777 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3778 to be verified with callout=postmaster, and the main callout worked but the
3779 postmaster check failed, the verification correctly failed. However, if a
3780 subsequent RCPT command asked for sender verification *without* the
3781 postmaster check, incorrect caching caused this verification also to fail,
3784 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3785 it was not caching the DNS options (qualify_single, search_parents) that
3786 were used when the lookup failed. A subsequent lookup with different
3787 options therefore always gave the same answer, though there were cases
3788 where it should not have. (Example: a "domains = !$mx_any" option on a
3789 dnslookup router: the "domains" option is always processed without any
3790 widening, but the router might have qualify_single set.) Now Exim uses the
3791 cached value only when the same options are set.
3793 32. Added John Jetmore's "exipick" utility to the distribution.
3795 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3796 than a timeout (e.g. a certificate is required, and is not provided), an
3797 Exim server now closes the connection immediately. Previously it waited for
3798 the client to close - but if the client is SSL, it seems that they each
3799 wait for each other, leading to a delay before one of them times out.
3801 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3802 maintained 0.8.x compatibility because I don't think many are using it, and
3803 it is clearly obsolete.
3805 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3808 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3809 client certificate was expired. A simple patch fixes this, though I don't
3810 understand the full logic of why the verify callback is called multiple
3813 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3814 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3815 which causes problems with some clients (such as the Certicom SSL Plus
3816 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3817 disables the coutermeasure allowing Eudora to connect."
3819 38. Exim was not checking that a write() to a log file succeeded. This could
3820 lead to Bad Things if a log got too big, in particular if it hit a file
3821 size limit. Exim now panics and dies if it cannot write to a log file, just
3822 as it does if it cannot open a log file.
3824 39. Modified OS/Makefile-Linux so that it now contains
3826 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3828 The two -D definitions ensure that Exim is compiled with large file
3829 support, which makes it possible to handle log files that are bigger than
3832 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3833 instance) a domain was checked against a named list that involved a lookup,
3834 causing $domain_data to be set, then another domain was checked against the
3835 same list, then the first domain was re-checked, the value of $domain_data
3836 after the final check could be wrong. In particular, if the second check
3837 failed, it could be set empty. This bug probably also applied to
3840 41. The strip_trailing_dot option was not being applied to the address given
3841 with the -f command-line option.
3843 42. The code for reading a message's header from the spool was incrementing
3844 $received_count, but never initializing it. This meant that the value was
3845 incorrect (doubled) while delivering a message in the same process in which
3846 it was received. In the most common configuration of Exim, this never
3847 happens - a fresh exec is done - but it can happen when
3848 deliver_drop_privilege is set.
3850 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3851 it now includes up to 150 characters of the unexpected data in the log
3854 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3855 and building data strings. The size of both of these buffers was 10 000
3856 bytes - far larger than anybody would *ever* want, thought I. Needless to
3857 say, somebody hit the limit. I have increased the maximum line length to
3858 20 000 and the maximum data length of concatenated lines to 100 000. I have
3859 also fixed two bugs, because there was no checking on these buffers. Tsk,
3860 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3861 buffer is too small.
3863 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3864 lsearch lookups. Now it does.
3866 46. When parsing a route_list item in a manualroute router, a fixed-length
3867 buffer was used for the list of hosts. I made this 1024 bytes long,
3868 thinking that nobody would ever have a list of hosts that long. Wrong.
3869 Somebody had a whole pile of complicated expansion conditions, and the
3870 string was silently truncated, leading to an expansion error. It turns out
3871 that it is easier to change to an unlimited length (owing to other changes
3872 that have happened since this code was originally written) than to build
3873 structure for giving a limitation error. The length of the item that
3874 expands into the list of hosts is now unlimited.
3876 47. The lsearch lookup could not handle data where the length of text line was
3877 more than 4095 characters. Such lines were truncated, leading to shortened
3878 data being returned. It should now handle lines of any length.
3880 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3881 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3884 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3885 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3886 no longer happens in comment lines. A list of replacements is now placed
3887 at the head of all of the source files, except those whose only change is
3888 to replace PERL_COMMAND in the very first #! line.
3890 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3891 messages on the queue, with a bottom-up merge sort, using code contributed
3892 by Michael Haardt. This should make operations like -bp somewhat faster on
3893 large queues. It won't affect queue runners, except when queue_run_in_order
3896 51. Installed eximstats 1.31 in the distribution.
3898 52. Added support for SRV lookups to the dnslookup router.
3900 53. If an ACL referred to $message_body or $message_body_end, the value was not
3901 reset for any messages that followed in the same SMTP session.
3903 54. The store-handling optimization for building very long strings was not
3904 differentiating between the different store pools. I don't think this
3905 actually made any difference in practice, but I've tidied it.
3907 55. While running the routers to verify a sender address, $sender_address
3908 was still set to the sender address. This is wrong, because when routing to
3909 send a bounce to the sender, it would be empty. Therefore, I have changed
3910 it so that, while verifying a sender address, $sender_address is set to <>.
3911 (There is no change to what happens when verifying a recipient address.)
3913 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3914 target A or AAAA records (if not already returned) without resetting the
3915 qualify_single or search_parents options of the DNS resolver. These are
3916 inappropriate in this case because the targets of MX and SRV records must
3917 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3918 target that, when qualified, matched something in the local domain. These
3919 two options are now turned off when doing these lookups.
3921 57. It seems that at least some releases of Reiserfs (which does not have the
3922 concept of a fixed number of inodes) returns zero and not -1 for the
3923 number of available inodes. This interacted badly with check_spool_inodes,
3924 which assumed that -1 was the "no such thing" setting. What I have done is
3925 to check that the total number of inodes is greater than zero before doing
3926 the test of how many are available.
3928 58. When a "warn" ACL statement has a log_message modifier, the message is
3929 remembered, and not repeated. This is to avoid a lot of repetition when a
3930 message has many recipients that cause the same warning to be written.
3931 Howewer, Exim was preserving the list of already written lines for an
3932 entire SMTP session, which doesn't seem right. The memory is now reset if a
3933 new message is started.
3935 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3936 debugging output unless log_rewrite was also set.
3938 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3939 of (int)(handle) when we know that handle contains (void *)(-1).
3941 61. The Exim daemon panic-logs an error return when it closes the incoming
3942 connection. However "connection reset by peer" seems to be common, and
3943 isn't really an error worthy of noting specially, so that particular error
3946 62. When Exim is trying to find all the local interfaces, it used to panic and
3947 die if the ioctl to get the interface flags failed. However, it seems that
3948 on at least one OS (Solaris 9) it is possible to have an interface that is
3949 included in the list of interfaces, but for which you get a failure error
3950 for this call. This happens when the interface is not "plumbed" into a
3951 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3952 failure of the "get flags" call assumes that the interface is down.
3954 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3955 makes life easier for people who are doing arithmetic on fields extracted
3956 from dates, where you often get leading zeros that should not be
3957 interpreted as octal.
3959 64. Added qualify_domain to the redirect router, to override the global
3962 65. If a pathologically long header line contained very many addresses (the
3963 report of this problem mentioned 10 000) and each of them was rewritten,
3964 Exim could use up a very large amount of memory. (It kept on making new
3965 copies of the header line as it rewrote, and never released the old ones.)
3966 At the expense of a bit more processing, the header rewriting function has
3967 been changed so that it no longer eats memory in this way.
3969 66. The generation of the Received: header has been moved from the time that a
3970 message starts to be received, to the time that it finishes. The timestamp
3971 in the Received: header should now be very close to that of the <= log
3972 line. There are two side-effects of this change:
3974 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3975 logged header lines no longer include the local Received: line, because
3976 it has not yet been created. The same applies to a copy of the message
3977 that is returned to a non-SMTP sender when a message is rejected.
3979 (b) When a filter file is tested using -bf, no additional Received: header
3980 is added to the test message. After some thought, I decided that this
3983 This change does not affect the value of $received_for. It is still set
3984 after address rewriting, but before local_scan() is called.
3986 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3988 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3989 gave an unhelpful panic error message, and a defer error. I have managed to
3990 change this behaviour so that it now rejects any supplied certificate,
3991 which seems right, as the list of acceptable certificates is empty.
3993 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3994 gave an unhelpful defer error. I have not managed to make this reject any
3995 supplied certificates, but the error message it gives is "no certificate
3996 supplied", which is not helpful.
3998 70. exigrep's output now also includes lines that are not associated with any
3999 message, but which match the given pattern. Implemented by a patch from
4000 Martin Sluka, which also tidied up the Perl a bit.
4002 71. Recipient callout verification, like sender verification, was using <> in
4003 the MAIL FROM command. This isn't really the right thing, since the actual
4004 sender may affect whether the remote host accepts the recipient or not. I
4005 have changed it to use the actual sender in the callout; this means that
4006 the cache record is now keyed on a recipient/sender pair, not just the
4007 recipient address. There doesn't seem to be a real danger of callout loops,
4008 since a callout by the remote host to check the sender would use <>.
4009 [SEE ABOVE: changed after hitting problems.]
4011 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4012 temporary errors. However, in the case of such a code being given after
4013 the end of a data transmission (i.e. after ".") Exim was failing to write
4014 a retry record for the message. (Yes, there was some broken host that was
4015 actually sending 8xx at this point.)
4017 73. An unknown lookup type in a host list could cause Exim to panic-die when
4018 the list was checked. (An example that provoked this was putting <; in the
4019 middle of a list instead of at the start.) If this happened during a DATA
4020 ACL check, a -D file could be left lying around. This kind of configuration
4021 error no longer causes Exim to die; instead it causes a defer errror. The
4022 incident is still logged to the main and panic logs.
4024 74. Buglet left over from Exim 3 conversion. The message "too many messages
4025 in one connection" was written to the rejectlog but not the mainlog, except
4026 when address rewriting (yes!) was being logged.
4028 75. Added write_rejectlog option.
4030 76. When a system filter was run not as root (that is, when system_filter_user
4031 was set), the values of the $n variables were not being returned to the
4032 main process; thus, they were not subsequently available in the $sn
4035 77. Added +return_path_on_delivery log selector.
4037 78. A connection timeout was being treated differently from recipients deferred
4038 when testing hosts_max_try with a message that was older than the host's
4039 retry timeout. (The host should not be counted, thus allowing all hosts to
4040 be tried at least once before bouncing.) This may have been the cause of an
4041 occasionally reported bug whereby a message would remain on the queue
4042 longer than the retry timeout, but would be bounced if a delivery was
4043 forced. I say "may" because I never totally pinned down the problem;
4044 setting up timeout/retry tests is difficult. See also the next item.
4046 79. The ultimate address timeout was not being applied to errors that involved
4047 a combination of host plus message (for example, a timeout on a MAIL
4048 command). When an address resolved to a number of possible hosts, and they
4049 were not all tried for each delivery (e.g. because of hosts_max_try), a
4050 message could remain on the queue longer than the retry timeout.
4052 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4055 81. Fixed an obscure SMTP outgoing bug which required at least the following
4056 conditions: (a) there was another message waiting for the same server;
4057 (b) the server returned 5xx to all RCPT commands in the first message so
4058 that the message was not completed; (c) the server dropped the connection
4059 or gave a negative response to the RSET that Exim sends to abort the
4060 transaction. The observed case was a dropped connection after DATA that had
4061 been sent in pipelining mode. That is, the server had advertised PIPELINING
4062 but was not implementing it correctly. The effect of the bug was incorrect
4063 behaviour, such as trying another host, and this could lead to a crash.
4069 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4070 and daemon.c were passed as pointers to ints; they should have been
4071 pointers to socklen_t variables (which are typically unsigned ints).
4073 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4076 3. Fixed a really odd bug that affected only the testing scheme; patching a
4077 certain fixed string in the binary changed the value of another string that
4078 happened to be identical to the end of the original first string.
4080 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4081 name", it returns that address as the IP address. On some operating
4082 systems (e.g. Solaris), it also passes back the IP address string as the
4083 "host name". However, on others (e.g. Linux), it passes back an empty
4084 string. Exim wasn't checking for this, and was changing the host name to an
4085 empty string, assuming it had been canonicized.
4087 5. Although rare, it is permitted to have more than one PTR record for a given
4088 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4089 all the names associated with an address, because they do in Solaris.
4090 However, it seems that they do not in Linux for data that comes from the
4091 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4092 I found this out when I moved to a new Linux workstation and tried to run
4093 the Exim test suite.
4095 To get round this problem I have changed the code so that it now does its
4096 own call to the DNS to look up PTR records when searching for a host name.
4097 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4098 addresses that are only in /etc/hosts are still found.
4100 This behaviour is, however, controlled by an option called host_lookup_
4101 order, which defaults to "bydns:byaddr". If people want to use the other
4102 order, or indeed, just use one or the other means of lookup, they can
4103 specify it in this variable.
4105 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4106 some operating systems, this comes back from gethostbyaddr() as an empty
4107 string, and this is what Exim used to test for. However, it seems that in
4108 other systems, "." is yielded. Exim now tests for this case too.
4110 7. The values of check_spool_space and check_log_space are now held internally
4111 as a number of kilobytes instead of an absolute number of bytes. If a
4112 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4113 kilobyte. This means that much larger values can be stored.
4115 8. Exim monitor: an attempt to get the action menu when not actually pointing
4116 at a message produces an empty menu entitled "No message selected". This
4117 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4118 no entries in it ("Shell widget menu has zero width and/or height"). So I
4119 have added a single, blank menu entry in this case.
4121 9. Added ${quote_local_part.
4123 10. MIME decoding is now applied to the contents of Subject: header lines when
4126 11. Now that a reference to $sender_host_address automatically causes a reverse
4127 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4128 host lookup before query-style lookups in lists that might use this
4129 variable. This has therefore been abolished, and the "net-" prefix is no
4130 longer necessary for query-style lookups.
4132 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4133 have been a typo for LFLAGS, so it has been changed.
4135 13. The install script calls Exim with "-C /dev/null" in order to find the
4136 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4137 to be output. Howeve, since Exim outputs its version number before the
4138 error, it didn't break the script. It just looked ugly. I fixed this by
4139 always allowing "-C /dev/null" if the caller is root.
4141 14. Ignore overlarge ACL variable number when reading spool file - insurance
4142 against a later release with more variables having written the file.
4144 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4145 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4146 this, as well as the form without the "IPv6" on the front.
4148 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4149 OS/Makefile-Darwin file.
4151 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4152 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4154 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4155 variable. However, some OS use ints for inodes. I've added cast to long int
4156 to get rid of the compiler warning.
4158 19. I had forgotten to lock out "/../" in configuration file names when
4159 ALT_CONFIG_PREFIX was set.
4161 20. Routers used for verification do not need to specify transports. However,
4162 if such a router generated a host list, and callout was configured, Exim
4163 crashed, because it could not find a port number from the (non-existent)
4164 transport. It now assumes port 25 in this circumstance.
4166 21. Added the -t option to exigrep.
4168 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4169 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4170 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4171 them is set, LOOKUP_LSEARCH is forced.
4173 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4175 24. Added sender and host information to the "rejected by local_scan()" log
4176 line; previously there was no indication of these.
4178 25. Added .include_if_exists.
4180 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4181 Linux. It turns out that not all file systems support this. Apparently some
4182 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4183 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4184 not supported on the file descriptor, is now ignored when Exim is trying to
4185 sync a directory. This applies only to Linux.
4187 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4189 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4190 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4193 29. Added support for maildirsize files from supplied patch (modified a bit).
4195 30. The use of :fail: followed by an empty string could lead Exim to respond to
4196 sender verification failures with (e.g.):
4198 550 Verification failed for <xxx>
4199 550 Sender verify failed
4201 where the first response line was missing the '-' that indicates it is not
4202 the final line of the response.
4204 31. The loop for finding the name of the user that called Exim had a hardwired
4205 limit of 10; it now uses the value of finduser_retries, which is used for
4206 all other user lookups.
4208 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4211 33. Exim was neglecting to zero errno before one call of strtol() when
4212 expanding a string and expecting an integer value. On some systems this
4213 resulted in spurious "integer overflow" errors. Also, it was casting the
4214 result into an int without checking.
4216 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4217 did not work. The code looks as if it has *never* worked, though it appears
4218 to have been documented since at least releast 1.62. I have made it work.
4220 35. The "timeout_DNS" error in retry rules, also documented since at least
4221 1.62, also never worked. As it isn't clear exactly what this means, and
4222 clearly it isn't a major issue, I have abolished the feature by treating it
4223 as "timeout", and writing a warning to the main and panic logs.
4225 36. The display of retry rules for -brt wasn't always showing the error code
4228 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4229 timeout_connect_A, timeout_connect_MX.
4231 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4232 to the empty sender.
4234 39. The daemon was not analysing the content of -oX till after it had closed
4235 stderr and disconnected from the controlling terminal. This meant that any
4236 syntax errors were only noted on the panic log, and the return code from
4237 the command was 0. By re-arranging the code a little, I've made the
4238 decoding happen first, so such errors now appear on stderr, and the return
4239 code is 1. However, the actual setting up of the sockets still happens in
4240 the disconnected process, so errors there are still only recorded on the
4243 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4244 connections (as happens on some IP stacks) was logged at start up time as
4245 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4246 it from "IPv6 and IPv4", which means that two separate sockets are being
4249 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4250 says whether AF_INET or AF_INET6 was passed as an argument.
4252 42. Exiwhat output was messed up when time zones were included in log
4255 43. Exiwhat now gives more information about the daemon's listening ports,
4256 and whether -tls-on-connect was used.
4258 44. The "port" option of the smtp transport is now expanded.
4260 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4261 silently ignored. Now an error message is written to the main and panic
4264 46. There's a new ACL modifier called "logwrite" which writes to a log file
4265 as soon as it is encountered.
4267 47. Added $local_user_uid and $local_user_gid at routing time.
4269 48. Exim crashed when trying to verify a sender address that was being
4272 49. Exim was recognizing only a space character after ".include". It now also
4273 recognizes a tab character.
4275 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4276 extracting the relevant information from the specification. The man page no
4277 longer contains scrambled data for the -d option, and I've added a section
4278 at the front about calling Exim under different names.
4280 51. Added "extra_headers" argument to the "mail" command in filter files.
4282 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4285 53. Installed eximstats 1.29.
4287 54. Added transport_filter_timeout as a generic transport option.
4289 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4290 Cc: header lines. This was required by RFC 822, but it not required by RFC
4293 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4294 incoming messages that did not have them. Now it does so only if the
4295 message originates locally, that is, if there is no associated remote host
4296 address. When Resent- header lines are present, this applies to the Resent-
4297 lines rather than the non-Resent- lines.
4299 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4300 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4302 58. Messages for configuration errors now include the name of the main
4303 configuration file - useful now that there may be more than one file in a
4304 list (.included file names were always shown).
4306 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4307 for those rare installations that do not start the daemon as root or run it
4308 setuid root. I've cut out the call to initgroups() if the daemon is not
4311 60. The Exim user and group can now be bound into the binary as text strings
4312 that are looked up at the start of Exim's processing.
4314 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4316 62. Added $mailstore_basename variable.
4318 63. Installed patch to sieve.c from Michael Haardt.
4320 64. When Exim failed to open the panic log after failing to open the main log,
4321 the original message it was trying to log was written to stderr and debug
4322 output, but if they were not available (the usual case in production), it
4323 was lost. Now it is written to syslog before the two lines that record the
4324 failures to open the logs.
4326 65. Users' Exim filters run in subprocesses under the user's uid. It is
4327 possible for a "deliver" command or an alias in a "personal" command to
4328 provoke an address rewrite. If logging of address rewriting is configured,
4329 this fails because the process is not running as root or exim. There may be
4330 a better way of dealing with this, but for the moment (because 4.30 needs
4331 to be released), I have disabled address rewrite logging when running a
4332 filter in a non-root, non-exim process.
4338 1. The buildconfig auxiliary program wasn't quoting the value set for
4339 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4340 not defined. This bug was masked in 4.22 by the effect that was fixed in
4343 2. Some messages that were rejected after a message id was allocated were
4344 shown as "incomplete" by exigrep. It no longer does this for messages that
4345 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4347 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4348 have allow_domain_literals set, the ID did not get logged in the <= line.
4349 Domain literals are now always recognized in Message-ID: header lines.
4351 4. The first argument for a ${extract expansion item is the key name or field
4352 number. Leading and trailing spaces in this item were not being ignored,
4353 causing some misleading effects.
4355 5. When deliver_drop_privilege was set, single queue runner processes started
4356 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4357 same command in the process it spins off) were not dropping privilege.
4359 6. When the daemon running as "exim" started a queue runner, it always
4360 re-executed Exim in the spun-off process. This is a waste of effort when
4361 deliver_drop_privilege is set. The new process now just calls the
4362 queue-runner function directly.
4368 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4371 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4372 ignored. Though the use of -oP was forcing the writing of a pid file, it
4373 was always written to the default place.
4375 3. If the message "no IP address found for host xxxx" is generated during
4376 incoming verification, it is now followed by identification of the incoming
4377 connection (so you can more easily find what provoked it).
4379 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4381 5. Added some features to "harden" Exim a bit more against certain attacks:
4383 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4384 be put in Local/Makefile. This is like the never_users runtime option,
4385 but it cannot be overridden. The default setting is "root".
4387 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4388 prefix string with which any file named in a -C command line option
4391 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4392 is retained for -C and -D only if the caller of Exim is root. Without
4393 it, the exim user may also use -C and -D and retain privilege.
4395 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4396 command line option is disabled.
4398 6. Macro names set by the -D option must start with an upper case letter, just
4399 like macro names defined in the configuration file.
4401 7. Added "dereference=" facility to LDAP.
4403 8. Two instances of the typo "uknown" in the source files are fixed.
4405 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4406 the Configure-Makefile script screwed up while processing it.
4408 10. Incorporated PCRE 4.4.
4410 11. The SMTP synchronization check was not operating right at the start of an
4411 SMTP session. For example, it could not catch a HELO sent before the client
4412 waited for the greeting. There is now a check for outstanding input at the
4413 point when the greeting is written. Because of the duplex, asynchronous
4414 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4415 way, but not yet received, when the check is performed.
4417 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4418 on TCP/IP sockets, because this apparently causes some broken clients to
4421 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4422 unchanged) from the Cygwin maintainer.
4424 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4425 was supported instead of testing for mbx. Effectively a typo.
4427 15. The spa authenticator server code was not checking that the input it
4428 received was valid base64.
4430 16. The debug output line for the "set" modifier in ACLs was not showing the
4431 name of the variable that was being set.
4433 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4435 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4436 lookup on its own. Something else was needed to trigger the lookup. For
4437 example, a match in host_lookup or the need for a host name in a host list.
4438 Now, if $sender_host_name is referenced and the host name has not yet been
4439 looked up, a lookup is performed. If the lookup fails, the variable remains
4440 empty, and $host_lookup_failed is set to "1".
4442 19. Added "eqi" as a case-independent comparison operator.
4444 20. The saslauthd authentication condition could segfault if neither service
4445 nor realm was specified.
4447 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4448 error message that was logged was misleading, and incoming SMTP
4449 connections were dropped. The message is now more accurate, and temporary
4450 errors are given to SMTP connections.
4452 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4453 (except RSET) with a 421 error, until QUIT is received. However, it was
4454 failing to send a response to QUIT.
4456 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4457 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4458 ACL is now run after the helo_try_verify_hosts code.
4460 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4461 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4462 case-independent, so other case variants are also recognized). Apparently
4463 some systems use these upper case variants.
4465 25. If more than two messages were waiting for the same host, and a transport
4466 filter was specified for the transport, Exim sent two messages over the
4467 same TCP/IP connection, and then failed with "socket operation on non-
4468 socket" when it tried to send the third.
4470 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4472 27. The extern definition of crypt16() in expand.c was not being excluded when
4473 the OS had its own crypt16() function.
4475 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4476 as a preferred synonym for return_size_limit, both as an option and as an
4479 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4481 30. Changed the default configuration ACL to relax the local part checking rule
4482 for addresses that are not in any local domains. For these addresses,
4483 slashes and pipe symbols are allowed within local parts, but the sequence
4484 /../ is explicitly forbidden.
4486 31. SPA server authentication was not clearing the challenge buffer before
4489 32. log_message in a "warn" ACL statement was writing to the reject log as
4490 well as to the main log, which contradicts the documentation and doesn't
4491 seem right (because no rejection is happening). So I have stopped it.
4493 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4494 However, I am unable to do any testing of this.
4496 34. Fixed an infelicity in the appendfile transport. When checking directories
4497 for a mailbox, to see if any needed to be created, it was accidentally
4498 using path names with one or more superfluous leading slashes; tracing
4499 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4501 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4502 contents are added to the log line that is written for every discarded
4503 recipient. (Previously a log_message setting was ignored.)
4505 36. The ${quote: operator now quotes the string if it is empty.
4507 37. The install script runs exim in order to find its version number. If for
4508 some reason other than non-existence or emptiness, which it checks, it
4509 could not run './exim', it was installing it with an empty version number,
4510 i.e. as "exim-". This error state is now caught, and the installation is
4513 38. An argument was missing from the function that creates an error message
4514 when Exim fails to connect to the socket for saslauthd authentication.
4515 This could cause Exim to crash, or give a corrupted message.
4517 39. Added isip, isip4, and isip6 to ${if conditions.
4519 40. The ACL variables $acl_xx are now saved with the message, and can be
4520 accessed later in routers, transports, and filters.
4522 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4523 strings in the file are not string-expanded.
4525 42. If a MAIL command specified a SIZE value that was too large to fit into an
4526 int variable, the check against message_size_limit failed. Such values are
4527 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4528 day this will have to be increased, but I don't think I want to be around
4529 when emails are that large.
4536 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4537 iconv() is not standard in FreeBSD.
4539 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4540 IPv6 enabled. The observed symptom was a segmentation fault on return from
4541 the function os_common_find_running_interfaces() in src/os.c.
4543 3. In the check_special_case() function in daemon.c I had used "errno" as an
4544 argument name, which causes warnings on some systems. This was basically a
4545 typo, since it was named "eno" in the comments!
4547 4. The code that waits for the clock to tick (at a resolution of some fraction
4548 of a second) so as to ensure message-id uniqueness was always waiting for
4549 at least one whole tick, when it could have waited for less. [This is
4550 almost certainly not relevant at current processor speeds, where it is
4551 unlikely to ever wait at all. But we try to future-proof.]
4553 5. The function that sleeps for a time interval that includes fractions of a
4554 second contained a race. It did not block SIGALRM between setting the
4555 timer, and suspending (a couple of lines later). If the interval was short
4556 and the sigsuspend() was delayed until after it had expired, the suspension
4557 never ended. On busy systems this could lead to processes getting stuck for
4560 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4561 process, before it forks any delivery processes. The open lookup caching
4562 mechanism meant that the open file or database connection was passed into
4563 the delivery process. The problem was that delivery processes always tidy
4564 up cached lookup data. This could cause a problem for the next delivery
4565 process started by the queue runner, because the external queue runner
4566 process does not know about the closure. So the next delivery process
4567 still has data in the lookup cache. In the case of a file lookup, there was
4568 no problem because closing a file descriptor in a subprocess doesn't affect
4569 the parent. However, if the lookup was caching a connection to a database,
4570 the connection was closed, and the second delivery process was likely to
4571 see errors such as "PGSQL: query failed: server closed the connection
4572 unexpectedly". The problem has been fixed by closing all cached lookups
4573 in a queue runner before running a delivery process.
4575 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4576 seem to have the "const" qualifier which it has on other OS. I've
4579 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4580 *of the same type* (client or server) with the same public name that an
4581 error should be diagnosed.
4583 9. When Exim looked up a host name for an IP address, but failed to find the
4584 original IP address when looking up the host name (a safety check), it
4585 output the message "<ip address> does not match any IP for NULL", which was
4586 confusing, to say the least. The bug was that the host name should have
4587 appeared instead of "NULL".
4589 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4590 user that is built into the binary, and the -C or -D options is used, root
4591 privilege is dropped before the configuration file is read. In addition,
4592 logging is switched to stderr instead of the normal log files. If the
4593 configuration then re-defines the Exim user, the unprivileged environment
4594 is probably not what is expected, so Exim logs a panic warning message (but
4597 However, if deliver_drop_privilege is set, the unprivileged state may well
4598 be exactly what is intended, so the warning has been cut out in that case,
4599 and Exim is allowed to try to write to its normal log files.
4605 1. smtp_return_error_details was not giving details for temporary sender
4606 or receiver verification errors.
4608 2. Diagnose a configuration error if two authenticators have the same public
4611 3. Exim used not to create the message log file for a message until the first
4612 delivery attempt. This could be confusing when incoming messages were held
4613 for policy or load reasons. The message log file is now created at the time
4614 the message is received, and an initial "Received" line is written to it.
4616 4. The automatically generated man page for command line options had a minor
4617 bug that caused no ill effects; however, a more serious problem was that
4618 the procedure for building the man page automatically didn't always
4619 operate. Consequently, release 4.20 contains an out-of-date version. This
4620 shouldn't happen again.
4622 5. When building Exim with embedded Perl support, the script that builds the
4623 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4624 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4626 6. The freeze_tell option was not being used for messages that were frozen on
4627 arrival, either by an ACL or by local_scan().
4629 7. Added the smtp_incomplete_transaction log selector.
4631 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4632 was accepting AUTH without a new EHLO.
4634 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4635 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4638 10. Logging of TCP/IP connections (when configured) now happens in the main
4639 daemon process instead of the child process, so that the TCP/IP connection
4640 count is more accurate (but it can never be perfect).
4642 11. The use of "drop" in a nested ACL was not being handled correctly in the
4643 outer ACL. Now, if condition failure induced by the nested "drop" causes
4644 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4645 or "require"), the connection is dropped.
4647 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4648 that yield "discard" can now be used with an "accept" or a "discard" verb,
4649 but an error is generated for any others (because I can't see a useful way
4650 to define what should happen).
4652 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4653 are now processed in the same way as lines in the Exim configuration file.
4654 In particular, continuation lines are supported.
4656 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4658 15. Added -ti meaning -t -i.
4660 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4661 domains, and warn by logging if others are found.
4663 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4664 returned from GIFCONF when Exim is trying to find the list of interfaces on
4665 a host. The code in os.c has been modified to copy each ifreq to an aligned
4666 structure in all cases.
4668 Also, in some cases, the returned ifreq's were being copied to a 'struct
4669 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4670 means the last couple of bytes of an IPv6 address could be chopped if the
4671 ifreq contained only a normal sockaddr (14 bytes storage).
4673 18. Named domain lists were not supported in the hosts_treat_as_local option.
4674 An entry such as +xxxx was not recognized, and was treated as a literal
4677 19. Ensure that header lines added by a DATA ACL are included in the reject log
4678 if the ACL subsequently rejects the message.
4680 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4681 MD5 (which is deprecated).
4683 21. When testing a filter file using -bf, Exim was writing a message when it
4684 took the sender from a "From " line in the message, but it was not doing so
4685 when it took $return_path from a Return-Path: header line. It now does.
4687 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4688 with a valid header line field name (a series of printing characters
4689 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4691 23. Changed "disc" in the source to "disk" to conform to the documentation and
4692 the book and for uniformity.
4694 24. Ignore Sendmail's -Ooption=value command line item.
4696 25. When execve() failed while trying to run a command in a pipe transport,
4697 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4698 could be confused with a return value of 69 from the command itself. This
4699 has been changed to 127, the value the shell returns if it is asked to run
4700 a non-existent command. The wording for the related log line suggests a
4701 non-existent command as the problem.
4703 26. If received_header_text expands to an empty string, do not add a Received:
4704 header line to the message. (Well, it adds a token one on the spool, but
4705 marks it "old" so that it doesn't get used or transmitted.)
4707 27. Installed eximstats 1.28 (addition of -nt option).
4709 28. There was no check for failure on the call to getsockname() in the daemon
4710 code. This can fail if there is a shortage of resources on the system, with
4711 ENOMEM, for example. A temporary error is now given on failure.
4713 29. Contrary to the C standard, it seems that in some environments, the
4714 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4715 program. Exim now does this explicitly; it affects the formatting of
4716 timestamps using strftime().
4718 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4719 complaints. I've now initialized all the variables, to avoid this.
4721 32. Header lines added by a system filter were not being "seen" during
4722 transport-time rewrites.
4724 33. The info_callback() function passed to OpenSSL is set up with type void
4725 (*)(SSL *, int, int), as described somewhere. However, when calling the
4726 function (actually a macro) that sets it up, the type void(*)() is
4727 expected. I've put in a cast to prevent warnings from picky compilers.
4729 34. If a DNS black list lookup found a CNAME record, but there were no A
4730 records associated with the domain it pointed at, Exim crashed.
4732 35. If a DNS black list lookup returned more than one A record, Exim ignored
4733 all but the first. It now scans all returned addresses if a particular IP
4734 value is being sought. In this situation, the contents of the
4735 $dnslist_value variable are a list of all the addresses, separated by a
4738 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4739 to accept a host name and all its aliases if the forward lookup for any of
4740 them yielded the IP address of the incoming connection. Now it accepts only
4741 those names whose forward lookup yields the correct IP address. Any other
4742 names are discarded. This closes a loophole whereby a rogue DNS
4743 administrator could create reverse DNS records to break through a
4744 wildcarded host restriction in an ACL.
4746 37. If a user filter or a system filter that ran in a subprocess used any of
4747 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4748 the wrong values were passed to the pipe command ($thisaddress had the
4749 value of $0, $0 had the value of $1, etc). This bug was introduced by
4750 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4752 38. Improved the line breaking for long SMTP error messages from ACLs.
4753 Previously, if there was no break point between 40 and 75 characters, Exim
4754 left the rest of the message alone. Two changes have been made: (a) I've
4755 reduced the minimum length to 35 characters; (b) if it can't find a break
4756 point between 35 and 75 characters, it looks ahead and uses the first one
4757 that it finds. This may give the occasional overlong line, but at least the
4758 remaining text gets split now.
4760 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4761 file descriptors might be low, and that setting 1000 would always raise it.
4762 It turns out that in some environments, the limit is already over 1000 and
4763 that lowering it causes trouble. So now Exim takes care not to decrease it.
4765 40. When delivering a message, the value of $return_path is set to $sender_
4766 address at the start of routing (routers may change the value). By an
4767 oversight, this default was not being set up when an address was tested by
4768 -bt or -bv, which affected the outcome if any router or filter referred to
4771 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4772 log only when "message" or "log_message" are set. However, if one of the
4773 conditions was an address verification, or a call to a nested ACL, the
4774 messages generated by the underlying test were being passed through. This
4775 no longer happens. The underlying message is available in $acl_verify_
4776 message for both "message" and "log_message" expansions, so it can be
4777 passed through if needed.
4779 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4780 new expansion $bh_ to give the encoded byte string without charset
4781 translation. Translation happens only if iconv() is available; HAVE_ICONV
4782 indicates this at build time. HEADERS_CHARSET gives the charset to
4783 translate to; headers_charset can change it in the configuration, and
4784 "headers charset" can change it in an individual filter file.
4786 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4787 that creates RFC 2047 encoded "words" labels them as that charset instead
4788 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4789 expansion operator; (ii) when Exim creates a From: line for a local
4790 message; (iii) when a header line is rewritten to include a "phrase" part.
4792 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4793 buggy, causing it to skip the first lines of messages whose message ID
4794 ended in 'D'. This would not have bitten before Exim release 4.14, because
4795 message IDs were unlikely to end in 'D' before then. The effect was to have
4796 incorrect size information for certain domains.
4798 45. #include "config.h" was missing at the start of the crypt16.c module. This
4799 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4802 46. If there was a timeout during a "random" callout check, Exim treated it as
4803 a failure of the random address, and carried on sending RSET and the real
4804 address. If the delay was just some slowness somewhere, the response to the
4805 original RCPT would be taken as a response to RSET and so on, causing
4806 mayhem of various kinds.
4808 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4809 when I implemented it. It didn't allow for the fact that some option values
4810 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4811 the right test for positive values.
4813 48. Domain names in DNS records are case-independent. Exim always looks them up
4814 in lower case. Some resolvers return domain names in exactly the case they
4815 appear in the zone file, that is, they may contain uppercase letters. Not
4816 all resolvers do this - some return always lower case. Exim was treating a
4817 change of case by a resolver as a change of domain, similar to a widening
4818 of a domain abbreviation. This triggered its re-routing code and so it was
4819 trying to route what was effectively the same domain again. This normally
4820 caused routing to fail (because the router wouldn't handle the domain
4821 twice). Now Exim checks for this case specially, and just changes the
4822 casing of the domain that it ultimately uses when it transmits the message
4825 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4828 50. If a filter generated a file delivery with a non-absolute name (possible if
4829 no home directory exists for the router), the forbid_file option was not
4832 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4833 the existing equality matching.
4835 52. Exim was using ints instead of ino_t variables in some places where it was
4836 dealing with inode numbers.
4838 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4839 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4840 called TMPDIR, and if it finds it is different, it changes its value.
4842 54. The smtp_printf() function is now made available to local_scan() so
4843 additional output lines can be written before returning. There is also an
4844 smtp_fflush() function to enable the detection of a dropped connection.
4845 The variables smtp_input and smtp_batched_input are exported to
4848 55. Changed the default runtime configuration: the message "Unknown user"
4849 has been removed from the ACL, and instead placed on the localuser router,
4850 using the cannot_route_message feature. This means that any verification
4851 failures that generate their own messages won't get overridden. Similarly,
4852 the "Unrouteable address" message that was in the ACL for unverifiable
4853 relay addresses has also been removed.
4855 56. Added hosts_avoid_esmtp to the smtp transport.
4857 57. The exicyclog script was not checking for the esoteric option
4858 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4859 will work only if exicyclog is run under the appropriate euid.
4861 58. Following a discussion on the list, the rules by which Exim recognises line
4862 endings on incoming messages have been changed. The -dropcr and drop_cr
4863 options are now no-ops, retained only for backwards compatibility. The
4864 following line terminators are recognized: LF CRLF CR. However, special
4865 processing applies to CR:
4867 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4868 nor a local message in the state where . is a terminator.
4870 (ii) If a bare CR is encountered in a header line, an extra space is added
4871 after the line terminator so as not to end the header. The reasoning
4872 behind this is that bare CRs in header lines are most likely either
4873 to be mistakes, or people trying to play silly games.
4875 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4876 was being incorrectly given as 18 bytes larger than it should have been.
4877 This is a VOB (very old bug).
4879 60. This may never have affected anything current, but just in case it has:
4880 When the local host is found other than at the start of a list of hosts,
4881 the local host, those with the same MX, and any that follow, are discarded.
4882 When the list in question was part of a longer list of hosts, the following
4883 hosts (not currently being processed) were also being discarded. This no
4884 longer happens. I'm not sure if this situation could ever has previously
4887 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4890 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4891 header. This is something that is recommended in a new Internet Draft, and
4892 is something that is documented as being done by Sendmail. There are two
4893 possible values. For messages generated by the autoreply transport, Exim
4896 Auto-Submitted: auto-replied
4898 whereas for all other generated messages (e.g. bounces) it adds
4900 Auto-Submitted: auto-generated
4902 63. The "personal" condition in filters now includes a test for the
4903 Auto-Submitted: header. If it contains the string "auto-" the message it
4904 not considered personal.
4906 64. Added rcpt_include_affixes as a generic transport option.
4908 65. Added queue_only_override (default true).
4910 66. Added the syslog_duplication option.
4912 67. If what should have been the first header line of a message consisted of
4913 a space followed by a colon, Exim was mis-interpreting it as a header line.
4914 It isn't of course - it is syntactically invalid and should therefore be
4915 treated as the start of the message body. The misbehaviour could have
4916 caused a number of strange effects, including loss of data in subsequent
4917 header lines, and spool format errors.
4919 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4920 client host had authenticated. This control can now be exercised by an ACL
4921 for more flexibility.
4923 69. By default, callouts do not happen when testing with -bh. There is now a
4924 variant, -bhc, which does actually run the callout code, including
4925 consulting and updating the callout cache.
4927 70. Added support for saslauthd authentication, courtesy of Alexander
4930 71. If statvfs() failed on the spool or log directories while checking their
4931 size for availability, Exim confusingly gave the error "space shortage".
4932 Furthermore, in debugging mode it crashed with a floating point exception.
4933 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4934 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4935 serious problem, Exim now writes to the main and panic logs when this
4936 happens, with details of the failure. It then refuses to accept the
4937 incoming message, giving the message "spool directory problem" or "log
4938 directory problem" with a 421 code for SMTP messages.
4940 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4941 0, 1, and 2 exist, because some OS complain for execs without them (see
4942 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4943 descriptors. However, the code omitted to check that the open succeeded,
4944 causing mysterious errors if for some reason the permissions on /dev/null
4945 got screwed. Now Exim writes a message to the main and panic logs, and
4946 bombs out if it can't open /dev/null.
4948 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4949 interact so that it is all more flexible. It is supposed to remain
4950 backwards compatible. Also added extra_local_interfaces.
4952 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4953 to bomb out with an assertion failure - to the client this appears as a
4954 connection drop. This problem occurs in the part of the code that was taken
4955 from the Samba project. Fortunately, the assertion is in a very simple
4956 function, so I have fixed this by reproducing the function inline in the
4957 one place where it is called, and arranging for authentication to fail
4958 instead of killing the process with assert().
4960 75. The SPA client code was not working when the server requested OEM rather
4961 than Unicode encoding.
4963 76. Added code to make require_files with a specific uid setting more usable in
4964 the case where statting the file as root fails - usually a non-root-mounted
4965 NFS file system. When this happens and the failure is EACCES, Exim now
4966 forks a subprocess and does the per-uid checking as the relevant uid.
4968 77. Added process_log_path.
4970 78. If log_file_path was not explicitly set, a setting of check_log_space or
4971 check_log_inodes was ignored.
4973 79. If a space check for the spool or log partitions fails, the incident is now
4974 logged. Of course, in the latter case the data may get lost...
4976 80. Added the %p formatting code to string_format() so that it can be used to
4977 print addresses in debug_print(). Adjusted all the address printing in the
4978 debugging in store.c to use %p rather than %d.
4980 81. There was a concern that a line of code in smtp_in.c could overflow a
4981 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4982 initially expressed, the concern was not well-founded, because trailing
4983 spaces are removed early. However, if the trailing spaces were followed by
4984 a NULL, they did not get removed, so the overflow was possible. Two fixes
4987 (a) I re-wrote the offending code in a cleaner fashion.
4988 (b) If an incoming SMTP command contains a NULL character, it is rejected
4991 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4992 runs initgroups(), so that if the Exim user is in any additional groups,
4993 they will be used during message reception.
4999 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob