1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.586 2009/11/05 19:31:15 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
127 plus update to original patch.
129 NM/26 Bugzilla 799: Documentation correction for ratelimit
131 NM/27 Bugzilla 802: Improvements to local interface IP addr detection
132 Patch provided by David Brownlee
134 NM/28 Bugzilla 807: Improvements to LMTP delivery logging
136 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes
138 NM/30 Bugzilla 888: TLS documentation bugfixes
140 NM/31 Bugzilla 896: Dovecot buffer overrun fix
142 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
143 Unlike the original bugzilla I have changed all shell scripts in src tree
145 NM/33 Bugzilla 898: Transport filter timeout fix
146 Patch by Todd Rinaldo
152 TK/01 Add preliminary DKIM support. Currently requires a forked version of
153 ALT-N's libdkim that I have put here:
154 http://duncanthrax.net/exim-experimental/
156 Note to Michael Haardt: I had to rename some vars in sieve.c. They
157 were called 'true' and it seems that C99 defines that as a reserved
158 keyword to be used with 'bool' variable types. That means you could
159 not include C99-style headers which use bools without triggering
160 build errors in sieve.c.
162 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
163 as mailq or other aliases. Changed the --help handling significantly
164 to do whats expected. exim_usage() emits usage/help information.
166 SC/01 Added the -bylocaldomain option to eximstats.
168 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
170 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
172 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
178 PH/01 Another patch from the Sieve maintainer.
180 PH/02 When an IPv6 address is converted to a string for single-key lookup
181 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
182 dots are used instead of colons so that keys in lsearch files need not
183 contain colons. This was done some time before quoting was made available
184 in lsearch files. However, iplsearch files do require colons in IPv6 keys
185 (notated using the quote facility) so as to distinguish them from IPv4
186 keys. This meant that lookups for IP addresses in host lists did not work
187 for iplsearch lookups.
189 This has been fixed by arranging for IPv6 addresses to be expressed with
190 colons if the lookup type is iplsearch. This is not incompatible, because
191 previously such lookups could never work.
193 The situation is now rather anomolous, since one *can* have colons in
194 ordinary lsearch keys. However, making the change in all cases is
195 incompatible and would probably break a number of configurations.
197 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
200 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
201 conversion specification without a maximum field width, thereby enabling
202 a rogue spamd server to cause a buffer overflow. While nobody in their
203 right mind would setup Exim to query an untrusted spamd server, an
204 attacker that gains access to a server running spamd could potentially
205 exploit this vulnerability to run arbitrary code as the Exim user.
207 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
208 $primary_hostname instead of what libspf2 thinks the hosts name is.
210 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
211 a directory entry by the name of the lookup key. Previously, if a
212 symlink pointed to a non-existing file or a file in a directory that
213 Exim lacked permissions to read, a lookup for a key matching that
214 symlink would fail. Now it is enough that a matching directory entry
215 exists, symlink or not. (Bugzilla 503.)
217 PH/03 The body_linecount and body_zerocount variables are now exported in the
220 PH/04 Added the $dnslist_matched variable.
222 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
223 This means they are set thereafter only if the connection becomes
226 PH/06 Added the client_condition to authenticators so that some can be skipped
227 by clients under certain conditions.
229 PH/07 The error message for a badly-placed control=no_multiline_responses left
230 "_responses" off the end of the name.
232 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
234 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
235 (without spaces) instead of just copying the configuration text.
237 PH/10 Added the /noupdate option to the ratelimit ACL condition.
239 PH/11 Added $max_received_linelength.
241 PH/12 Added +ignore_defer and +include_defer to host lists.
243 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
244 way in which PCRE > 7.0 is built.
246 PH/14 Implemented queue_only_load_latch.
248 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
249 MAIL command. The effect was to mangle the value on 64-bit systems.
251 PH/16 Another patch from the Sieve maintainer.
253 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
255 PH/18 If a system quota error occurred while trying to create the file for
256 a maildir delivery, the message "Mailbox is full" was not appended to the
257 bounce if the delivery eventually timed out. Change 4.67/27 below applied
258 only to a quota excession during the actual writing of the file.
260 PH/19 It seems that peer DN values may contain newlines (and other non-printing
261 characters?) which causes problems in log lines. The DN values are now
262 passed through string_printing() before being added to log lines.
264 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
265 and InterBase are left for another time.)
267 PH/21 Added message_body_newlines option.
269 PH/22 Guard against possible overflow in moan_check_errorcopy().
271 PH/23 POSIX allows open() to be a macro; guard against that.
273 PH/24 If the recipient of an error message contained an @ in the local part
274 (suitably quoted, of course), incorrect values were put in $domain and
275 $local_part during the evaluation of errors_copy.
281 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
282 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
285 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
286 issue a MAIL command.
288 PH/02 In an ACL statement such as
290 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
292 if a client was not listed at all, or was listed with a value other than
293 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
294 the condition was not true (as it should be), so access was not denied.
295 The bug was that the ! inversion was incorrectly passed on to the second
296 item. This has been fixed.
298 PH/03 Added additional dnslists conditions == and =& which are different from
299 = and & when the dns lookup returns more than one IP address.
301 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
302 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
304 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
305 FSYNC, which compiles an option called disable_fsync that allows for
306 bypassing fsync(). The documentation is heavily laced with warnings.
308 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
310 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
311 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
312 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
313 including adding "make clean"; (3) Added -fPIC when compiling the test
314 dynamically loaded module, to get rid of a warning.
316 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
317 message fails, move_frozen_messages = true and ignore_bounce_errors_after
318 = 0s. The bug is otherwise harmless.
320 PH/07 There was a bug in the dovecot authenticator such that the value of
321 $auth1 could be overwritten, and so not correctly preserved, after a
322 successful authentication. This usually meant that the value preserved by
323 the server_setid option was incorrect.
325 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
327 PH/09 Installed PCRE release 7.0.
329 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
330 run for batched SMTP input. It is now run at the start of every message
331 in the batch. While fixing this I discovered that the process information
332 (output by running exiwhat) was not always getting set for -bs and -bS
333 input. This is fixed, and it now also says "batched" for BSMTP.
335 PH/11 Added control=no_pipelining.
337 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
338 patch, slightly modified), and move the expansion of helo_data till after
339 the connection is made in the smtp transport (so it can use these
342 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
344 PH/14 Added log_selector = +pid.
346 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
348 PH/16 Add ${if forany and ${if forall.
350 PH/17 Added dsn_from option to vary the From: line in DSNs.
352 PH/18 Flush SMTP output before performing a callout, unless control =
353 no_callout_flush is set.
355 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
356 was true (the default) a successful delivery failed to delete the retry
357 item, thus causing premature timeout of the address. The bug is now
360 PH/20 Added hosts_avoid_pipelining to the smtp transport.
362 PH/21 Long custom messages for fakedefer and fakereject are now split up
363 into multiline reponses in the same way that messages for "deny" and
364 other ACL rejections are.
366 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
367 with slight modification.
369 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
370 draft, changing the syntax and factoring some duplicate code".
372 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
373 for deliveries of the second and subsequent messages over the same SMTP
376 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
377 ${reduce, with only minor "tidies".
379 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
381 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
382 expansion side effects.
384 PH/27 When a message times out after an over-quota error from an Exim-imposed
385 quota, the bounce message says "mailbox is full". This message was not
386 being given when it was a system quota that was exceeded. It now should
389 MH/03 Made $recipients available in local_scan(). local_scan() already has
390 better access to the recipient list through recipients_list[], but
391 $recipients can be useful in postmaster-provided expansion strings.
393 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
394 in the case of a MAIL command with additional options following the
395 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
396 were accidentally chopped off.
398 PH/29 SMTP synchronization checks are implemented when a command is read -
399 there is a check that no more input is waiting when there shouldn't be
400 any. However, for some commands, a delay in an ACL can mean that it is
401 some time before the response is written. In this time, more input might
402 arrive, invalidly. So now there are extra checks after an ACL has run for
403 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
404 pipelining has not been advertised.
406 PH/30 MH's patch to allow iscntrl() characters to be list separators.
408 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
409 returned in the SMTP response when smtp_return_error_details was false.
412 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
413 instead of the C I/O that was originally supplied, because problems were
416 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
417 Exim which did not show up earlier: it was assuming that a call to
418 SSL_CTX_set_info_callback() might give an error value. In fact, there is
419 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
420 was a macro that became an assignment, so it seemed to work. This has
421 changed to a proper function call with a void return, hence the compile
422 error. Exim's code has been fixed.
424 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
427 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
429 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
431 PH/37 If a message is not accepted after it has had an id assigned (e.g.
432 because it turns out to be too big or there is a timeout) there is no
433 "Completed" line in the log. When some messages of this type were
434 selected by exigrep, they were listed as "not completed". Others were
435 picked up by some special patterns. I have improved the selection
436 criteria to be more general.
438 PH/38 The host_find_failed option in the manualroute router can now be set
439 to "ignore", to completely ignore a host whose IP address cannot be
440 found. If all hosts are ignored, the behaviour is controlled by the new
441 host_all_ignored option.
443 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
444 homing or because of multiple MX records with /mx) generated more than
445 one IP address, and the following item turned out to be the local host,
446 all the secondary addresses of the first item were incorrectly removed
447 from the list, along with the local host and any following hosts (which
448 is what is supposed to happen).
450 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
451 whoever called Exim into the -H file. In the case of the daemon it was
452 behaving confusingly. When first started, it used values for whoever
453 started the daemon, but after a SIGHUP it used the Exim user (because it
454 calls itself on a restart). I have changed the code so that it now always
457 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
458 message are rejected with the same error (e.g. no authentication or bad
459 sender address), and a DATA command is nevertheless sent (as can happen
460 with PIPELINING or a stupid MUA), the error message that was given to the
461 RCPT commands is included in the rejection of the DATA command. This is
462 intended to be helpful for MUAs that show only the final error to their
465 PH/42 Another patch from the Sieve maintainer.
467 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
468 Eximstats - Fixed some broken HTML links and added missing column headers
470 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
471 columns for Rejects, Temp Rejects, Ham, and Spam rows.
473 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
475 PH/43 Yet another patch from the Sieve maintainer.
477 PH/44 I found a way to check for a TCP/IP connection going away before sending
478 the response to the final '.' that terminates a message, but only in the
479 case where the client has not sent further data following the '.'
480 (unfortunately, this is allowed). However, in many cases there won't be
481 any further data because there won't be any more messages to send. A call
482 to select() can be used: if it shows that the input is "ready", there is
483 either input waiting, or the socket has been closed. An attempt to read
484 the next input character can distinguish the two cases. Previously, Exim
485 would have sent an OK response which the client would never have see.
486 This could lead to message repetition. This fix should cure that, at
487 least in a lot of common cases.
489 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
490 advertised in response to EHLO.
496 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
497 fixed by 4.65/MH/01 (is this a record?) are fixed:
499 (i) An empty string was always treated as zero by the numeric comparison
500 operators. This behaviour has been restored.
502 (ii) It is documented that the numeric comparison operators always treat
503 their arguments as decimal numbers. This was broken in that numbers
504 starting with 0 were being interpreted as octal.
506 While fixing these problems I realized that there was another issue that
507 hadn't been noticed. Values of message_size_limit (both the global option
508 and the transport option) were treated as octal if they started with 0.
509 The documentation was vague. These values are now always treated as
510 decimal, and I will make that clear in the documentation.
516 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
517 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
520 MH/01 Don't check that the operands of numeric comparison operators are
521 integers when their expansion is in "skipping" mode (fixes bug
522 introduced by 4.64-PH/07).
524 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
525 child addresses, Exim now panics and dies. Previously, because the count
526 is held in a short int, deliveries were likely to be lost. As such a
527 large number of recipients for a single message is ridiculous
528 (performance will be very, very poor), I have chosen to impose a limit
529 rather than extend the field.
535 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
536 leftover -K file (the existence of which was triggered by #402).
537 While we were at it, introduced process PID as part of the -K
538 filename. This should rule out race conditions when creating
541 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
542 processing considerably. Previous code took too long for large mails,
543 triggering a timeout which in turn triggers #401.
545 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
546 in the DK code in transports.c. sendfile() is not really portable,
547 hence the _LINUX specificness.
549 TF/01 In the add_headers option to the mail command in an Exim filter,
550 there was a bug that Exim would claim a syntax error in any
551 header after the first one which had an odd number of characters
554 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
555 callout verification, Exim cached a "reject" for the entire domain. This
556 is correct for most verifications, but it is not correct for a recipient
557 verification with use_sender or use_postmaster set, because in that case
558 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
559 case of MAIL FROM:<> rejection from other early rejections (e.g.
560 rejection of HELO). When verifying a recipient using a non-null MAIL
561 address, the cache is ignored if it shows MAIL FROM:<> rejection.
562 Whatever the result of the callout, the value of the domain cache is
563 left unchanged (for any other kind of callout, getting as far as trying
564 RCPT means that the domain itself is ok).
566 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
569 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
570 manifest itself as EPIPE rather than ECONNECT. When tidying away a
571 session, the daemon ignores ECONNECT errors and logs others; it now
572 ignores EPIPE as well.
574 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
575 (quoted-printable decoding).
577 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
578 later the small subsequent patch to fix an introduced bug.
580 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
582 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
584 PH/08 An error is now given if message_size_limit is specified negative.
586 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
587 to be given (somewhat) arbitrary names.
589 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
592 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
593 miscellaneous code fixes
595 PH/10 Added the log_reject_target ACL modifier to specify where to log
598 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
599 hostname. This is wrong, because it relates to the incoming message (and
600 probably the interface on which it is arriving) and not to the outgoing
601 callout (which could be using a different interface). This has been
602 changed to use the value of the helo_data option from the smtp transport
603 instead - this is what is used when a message is actually being sent. If
604 there is no remote transport (possible with a router that sets up host
605 addresses), $smtp_active_hostname is used.
607 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
608 tweaks were necessary in order to get it to work (see also 21 below):
609 (a) The code assumed that strncpy() returns a negative number on buffer
610 overflow, which isn't the case. Replaced with Exim's string_format()
612 (b) There were several signed/unsigned issues. I just did the minimum
613 hacking in of casts. There is scope for a larger refactoring.
614 (c) The code used strcasecmp() which is not a standard C function.
615 Replaced with Exim's strcmpic() function.
616 (d) The code set only $1; it now sets $auth1 as well.
617 (e) A simple test gave the error "authentication client didn't specify
618 service in request". It would seem that Dovecot has changed its
619 interface. Fortunately there's a specification; I followed it and
620 changed what the client sends and it appears to be working now.
622 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
625 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
626 address A is aliased to B and C, where B exists and C does not. Without
627 -v the output is "A verified" because verification stops after a
628 successful redirection if more than one address is generated. However,
629 with -v the child addresses are also verified. Exim was outputting "A
630 failed to verify" and then showing the successful verification for C,
631 with its parentage. It now outputs "B failed to verify", showing B's
632 parentage before showing the successful verification of C.
634 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
635 look up a TXT record in a specific list after matching in a combined
638 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
639 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
640 they consult the DNS. I had assumed they would set it the way they
641 wanted; and indeed my experiments on Linux seem to show that in some
642 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
643 To be on the safe side, however, I have now made the interface to
644 host_find_byname() similar to host_find_bydns(), with an argument
645 containing the DNS resolver options. The host_find_byname() function now
646 sets these options at its start, just as host_find_bydns() does. The smtp
647 transport options dns_qualify_single and dns_search_parents are passed to
648 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
649 of host_find_byname() use the default settings of RES_DEFNAMES
650 (qualify_single) but not RES_DNSRCH (search_parents).
652 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
653 spool_read_header() do less string testing, by means of a preliminary
654 switch on the second character of optional "-foo" lines. (This is
655 overdue, caused by the large number of possibilities that now exist.
656 Originally there were few.) While I was there, I also converted the
657 str(n)cmp tests so they don't re-test the leading "-" and the first
658 character, in the hope this might squeeze out yet more improvement.
660 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
661 flag allowing group syntax was set by the header_syntax check but not
662 turned off, possible causing trouble later; (2) The flag was not being
663 set at all for the header_verify test, causing "group"-style headers to
664 be rejected. I have now set it in this case, and also caused header_
665 verify to ignore an empty address taken from a group. While doing this, I
666 came across some other cases where the code for allowing group syntax
667 while scanning a header line wasn't quite right (mostly, not resetting
668 the flag correctly in the right place). These bugs could have caused
669 trouble for malformed header lines. I hope it is now all correct.
671 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
672 with the "reply" argument non-NULL. The code, however (which originally
673 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
674 but it didn't always do it. This confused somebody who was copying the
675 code for some other use. I have removed all the tests.
677 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
678 feature that was used to support insecure browsers during the U.S. crypto
679 embargo. It requires special client support, and Exim is probably the
680 only MTA that supported it -- and would never use it because real RSA is
681 always available. This code has been removed, because it had the bad
682 effect of slowing Exim down by computing (never used) parameters for the
683 RSA_EXPORT functionality.
685 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
686 authenticator to fail if there's a tab character in the incoming data
687 (there should never be unless someone is messing about, as it's supposed
688 to be base64-encoded). Also added, on Timo's advice, the "secured" option
689 if the connection is using TLS or if the remote IP is the same as the
690 local IP, and the "valid-client-cert option" if a client certificate has
693 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
694 authenticators. This can be used for authorization after authentication
695 succeeds. (In the case of plaintext, it servers for both authentication
698 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
699 if any retry times were supplied.
701 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
702 connection, where there is no client IP address to check. In this
703 situation, the verify now always succeeds.
705 PH/25 Applied John Jetmore's -Mset patch.
707 PH/26 Added -bem to be like -Mset, but loading a message from a file.
709 PH/27 In a string expansion for a processed (not raw) header when multiple
710 headers of the same name were present, leading whitespace was being
711 removed from all of them, but trailing whitespace was being removed only
712 from the last one. Now trailing whitespace is removed from each header
713 before concatenation. Completely empty headers in a concatenation (as
716 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
717 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
719 PH/29 [Removed. This was a change that I later backed out, and forgot to
720 correct the ChangeLog entry (that I had efficiently created) before
721 committing the later change.]
723 PH/30 Exim was sometimes attempting to deliver messages that had suffered
724 address errors (4xx response to RCPT) over the same connection as other
725 messages routed to the same hosts. Such deliveries are always "forced",
726 so retry times are not inspected. This resulted in far too many retries
727 for the affected addresses. The effect occurred only when there were more
728 hosts than the hosts_max_try setting in the smtp transport when it had
729 the 4xx errors. Those hosts that it had tried were not added to the list
730 of hosts for which the message was waiting, so if all were tried, there
731 was no problem. Two fixes have been applied:
733 (i) If there are any address or message errors in an SMTP delivery, none
734 of the hosts (tried or untried) are now added to the list of hosts
735 for which the message is waiting, so the message should not be a
736 candidate for sending over the same connection that was used for a
737 successful delivery of some other message. This seems entirely
738 reasonable: after all the message is NOT "waiting for some host".
739 This is so "obvious" that I'm not sure why it wasn't done
740 previously. Hope I haven't missed anything, but it can't do any
741 harm, as the worst effect is to miss an optimization.
743 (ii) If, despite (i), such a delivery is accidentally attempted, the
744 routing retry time is respected, so at least it doesn't keep
745 hammering the server.
747 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
748 in ${readsocket because some servers need this prod.
750 PH/32 Added some extra debug output when updating a wait-xxx database.
752 PH/33 The hint "could be header name not terminated by colon", which has been
753 given for certain expansion errors for a long time, was not being given
754 for the ${if def:h_colon_omitted{... case.
756 PH/34 The spec says: "With one important exception, whenever a domain list is
757 being scanned, $domain contains the subject domain." There was at least
758 one case where this was not true.
760 PH/35 The error "getsockname() failed: connection reset by peer" was being
761 written to the panic log as well as the main log, but it isn't really
762 panic-worthy as it just means the connection died rather early on. I have
763 removed the panic log writing for the ECONNRESET error when getsockname()
766 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
767 runs only) independently of the message's sender address. This meant
768 that, if the 4xx error was in fact related to the sender, a different
769 message to the same recipient with a different sender could confuse
770 things. In particualar, this can happen when sending to a greylisting
771 server, but other circumstances could also provoke similar problems.
772 I have changed the default so that the retry time for these errors is now
773 based a combination of the sender and recipient addresses. This change
774 can be overridden by setting address_retry_include_sender=false in the
777 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
778 remote server are returned as part of bounce messages. This was not
779 happening for LMTP over a pipe (the lmtp transport), but now it is the
780 same for both kinds of LMTP.
782 PH/38 Despite being documented as not happening, Exim was rewriting addresses
783 in header lines that were in fact CNAMEs. This is no longer the case.
785 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
786 and queue runs started by the daemon processed all messages. This has
787 been fixed so that -R and -S can now usefully be given with -q<time>.
789 PH/40 Import PCRE release 6.7 (fixes some bugs).
791 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
793 PH/42 Give an error if -q is specified more than once.
795 PH/43 Renamed the variables $interface_address and $interface_port as
796 $received_ip_address and $received_port, to make it clear that these
797 values apply to message reception, and not to the outgoing interface when
798 a message is delivered. (The old names remain recognized, of course.)
800 PH/44 There was no timeout on the connect() call when using a Unix domain
801 socket in the ${readsocket expansion. There now is.
803 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
804 be meaningful with "accept".
806 SC/01 Eximstats V1.43
807 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
809 SC/02 Eximstats V1.44
810 Use a glob alias rather than an array ref in the generated
811 parser. This improves both readability and performance.
813 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
814 Collect SpamAssassin and rejection statistics.
815 Don't display local sender or destination tables unless
816 there is data to show.
817 Added average volumes into the top table text output.
819 SC/04 Eximstats V1.46
820 Collect data on the number of addresses (recipients)
821 as well as the number of messages.
823 SC/05 Eximstats V1.47
824 Added 'Message too big' to the list of mail rejection
825 reasons (thanks to Marco Gaiarin).
827 SC/06 Eximstats V1.48
828 Mainlog lines which have GMT offsets and are too short to
829 have a flag are now skipped.
831 SC/07 Eximstats V1.49 (Alain Williams)
832 Added the -emptyok flag.
834 SC/08 Eximstats V1.50
835 Fixes for obtaining the IP address from reject messages.
837 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
838 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
839 whitesspace changes from 4.64-PH/27
841 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
844 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
845 are found, allow negative numbers in numeric criteria)
847 JJ/06 exipick.20061117.2, added new $message_body_missing variable
849 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
850 to match changes made in 4.64-PH/43
852 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
854 PH/47 Put in an explicit test for a DNS lookup of an address record where the
855 "domain" is actually an IP address, and force a failure. This locks out
856 those revolvers/nameservers that support "A-for-A" lookups, in
857 contravention of the specifications.
859 PH/48 When a host name was looked up from an IP address, and the subsequent
860 forward lookup of the name timed out, the host name was left in
861 $sender_host_name, contrary to the specification.
863 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
864 restricted to single-key lookups, Exim was not diagnosing an error if
865 * or *@ was used with a query-style lookup.
867 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
869 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
870 long ago, but noone interested enough thought of it. Let's just say that
871 the "1.1" means that there are some new functions that weren't there at
872 some point in the past.
874 PH/51 Error processing for expansion failure of helo_data from an smtp
875 transport during callout processing was broken.
877 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
878 tested/used via the -bh/-bhc/-bs options.
880 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
881 bug, fixed in subsequent PCRE releases).
883 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
884 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
886 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
892 SC/01 Use a glob alias rather than an array ref in eximstats generated
893 parser. This improves both readability and performance.
895 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
896 Don't display local sender or destination tables in eximstats unless
897 there is data to show.
898 Added average volumes into the eximstats top table text output.
900 SC/03 Collect data on the number of addresses (recipients) as well
901 as the number of messages in eximstats.
903 TF/01 Correct an error in the documentation for the redirect router. Exim
904 does (usually) call initgroups() when daemonizing.
906 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
907 with consistent privilege compared to when running as a daemon.
909 TF/03 Note in the spec that $authenticated_id is not set for local
910 submissions from trusted users.
912 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
913 Thanks to Dean Brooks <dean@iglou.com> for the patch.
915 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
916 by adding some example configuration directives to the default
917 configuration file. A little bit of work is required to uncomment the
918 directives and define how usernames and passwords are checked, but
919 there is now a framework to start from.
921 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
922 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
923 without this. I don't know how relevant this is to other LDAP libraries.
925 PH/02 Add the verb name to the "unknown ACL verb" error.
927 PH/03 Magnus Holmgren's patch for filter_prepend_home.
929 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
931 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
932 directory not expanded when it should be if an expanded home directory
933 was set for the address (which is overridden by the transport).
935 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
938 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
939 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
940 because it is too late at that time, and has no effect.
942 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
943 security issue with \' (bugzilla #107). I could not use the
944 PQescapeStringConn() function, because it needs a PGconn value as one of
947 PH/08 When testing addresses using -bt, indicate those final addresses that
948 are duplicates that would not cause an additional delivery. At least one
949 person was confused, thinking that -bt output corresponded to deliveries.
950 (Suppressing duplicates isn't a good idea as you lose the information
951 about possibly different redirections that led to the duplicates.)
953 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
954 systems where poll() doesn't work, in particular OS X.
956 PH/10 Added more information to debugging output for retry time not reached.
958 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
959 operations in malware.c.
961 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
964 PH/13 If write_rejectlog was set false when logging was sent to syslog with
965 syslog_duplication set false, log lines that would normally be written
966 both the the main log and to the reject log were not written to syslog at
969 PH/14 In the default configuration, change the use of "message" in ACL warn
970 statements to "add_header".
972 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
973 not followed by a command (e.g. "seen endif").
975 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
976 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
979 PH/17 Added extra conditions to the default value of delay_warning_condition
983 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
984 { match{$h_precedence:}{(?i)bulk|list|junk} } \
985 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
988 The Auto-Submitted: and various List- headers are standardised, whereas I
989 don't think Precedence: ever was.
991 PH/18 Refactored debugging code in route_finduser() to show more information,
992 in particular, the error code if getpwnam() issues one.
994 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
995 This is apparently needed in addition to the PH/07 change above to avoid
996 any possible encoding problems.
998 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
999 but not after initializing Perl.
1001 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1002 output them only if debugging. By default they are written stderr,
1003 apparently, which is not desirable.
1005 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1008 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1011 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1013 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1014 authenticated or an ident call has been made. Suppress the default
1015 values for $authenticated_id and $authenticated_sender (but permit -oMai
1016 and -oMas) when testing with -bh.
1018 PH/24 Re-jigged the order of the tests in the default configuration so that the
1019 tests for valid domains and recipients precede the DNS black list and CSA
1020 tests, on the grounds that those ones are more expensive.
1022 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1023 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1024 This bug exists in every version of Exim that I still have, right back to
1027 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1028 However, an attempt to turn on case-sensitivity in a regex key by
1029 including (?-i) didn't work because the subject string was already
1030 lowercased, and the effects were non-intuitive. It turns out that a
1031 one-line patch can be used to allow (?-i) to work as expected.
1037 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1038 other effects) broke the use of negated acl sub-conditions.
1040 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1043 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1044 "Deny" causes Exim to reject the incoming connection with a 554 error.
1045 Unfortunately, if there is a major crisis, such as a disk failure,
1046 tcp-wrappers gives "deny", whereas what one would like would be some
1047 kind of temporary error. A kludge has been added to help with this.
1048 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1049 554 error is used if errno is still zero or contains ENOENT (which occurs
1050 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1053 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1055 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1056 errors. Otherwise a message that provokes a temporary error (when other
1057 messages do not) can cause a whole host to time out.
1059 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1060 addresses were routed directly to files or pipes from a redirect router.
1061 File deliveries just didn't batch; pipe deliveries might have suffered
1064 PH/06 A failure to get a lock for a hints database would erroneously always say
1065 "Failed to get write lock", even when it was really a read lock.
1067 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1068 of 0600. This has been changed to use the value of the lockfile_mode
1069 option (which defaults to 0600).
1071 PH/08 Applied small patch from the Sieve maintainer.
1073 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1074 folder from quota calculations, a direct delivery into this folder messed
1075 up the contents of the maildirsize file. This was because the regex was
1076 used only to exclude .Trash (or whatever) when the size of the mailbox
1077 was calculated. There was no check that a delivery was happening into an
1078 excluded directory. This bug has been fixed by ignoring all quota
1079 processing for deliveries into excluded directories.
1081 PH/10 Added the maildirfolder_create_regex option to appendfile.
1087 PH/01 The code for finding all the local interface addresses on a FreeBSD
1088 system running IPv6 was broken. This may well have applied to all BSD
1089 systems, as well as to others that have similar system calls. The broken
1090 code found IPv4 interfaces correctly, but gave incorrect values for the
1091 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1092 that it would not match correctly against @[] and not recognize the IPv6
1095 PH/02 The ipliteral router was not recognizing addresses of the form user@
1096 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1098 PH/03 Added disable_ipv6.
1100 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1101 decoded form, because it is most often used to construct To: headers
1102 lines in autoreplies, and the decoded form may well be syntactically
1103 invalid. However, $reply_address has leading white space removed, and all
1104 newlines turned into spaces so that the autoreply transport does not
1107 PH/05 If group was specified without a user on a router, and no group or user
1108 was specified on a transport, the group from the router was ignored.
1110 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1111 for visible compile-time settings that can be used to change these
1112 numbers, for those that want even more. Backwards compatibility with old
1113 spool files has been maintained. However, going back to a previous Exim
1114 release will lost any variables that are in spool files.
1116 PH/07 Two small changes when running in the test harness: increase delay when
1117 passing a TCP/IP connection to a new process, in case the original
1118 process has to generate a bounce, and remove special handling of
1119 127.0.0.2 (sic), which is no longer necessary.
1121 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1122 be the same on different OS.
1124 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1127 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1128 whether --show-vars was specified or not
1130 JJ/02 exipick: Added support for new ACL variable spool format introduced
1133 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1134 syntactically invalid From: or Reply-to: line, and a filter used this to
1135 generate an autoreply, and therefore failed to obtain an address for the
1136 autoreply, Exim could try to deliver to a non-existent relative file
1137 name, causing unrelated and misleading errors. What now happens is that
1138 it logs this as a hard delivery error, but does not attempt to create a
1141 PH/11 The exinext utility has a -C option for testing purposes, but although
1142 the given file was scanned by exinext itself; it wasn't being passed on
1143 when Exim was called.
1145 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1146 an end-of-file indication when reading a command response.
1148 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1149 compiled. In many other places in Exim, IPv6 addresses are always
1150 recognized, so I have changed this. It also means that IPv4 domain
1151 literals of the form [IPV4:n.n.n.n] are now always recognized.
1153 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1154 used if the router is not running as root, for example, when verifying at
1155 ACL time, or when using -bh. The debugging output from this situation was
1156 non-existent - all you got was a failure to exec. I have made two
1159 (a) Failures to set uid/gid, the current directory, or a process leader
1160 in a subprocess such as that created by queryprogram now generate
1161 suitable debugging ouput when -d is set.
1163 (b) The queryprogram router detects when it is not running as root,
1164 outputs suitable debugging information if -d is set, and then runs
1165 the subprocess without attempting to change uid/gid.
1167 PH/15 Minor change to Makefile for building test_host (undocumented testing
1170 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1171 additional section of a DNS packet that returns MX or SRV records.
1172 Instead, it always explicitly searches for A/AAAA records. This avoids
1173 major problems that occur when a DNS server includes only records of one
1174 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1175 fixed another bug: if SRV records were looked up and the corresponding
1176 address records were *not* found in the additional section, the port
1177 values from the SRV records were lost.
1179 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1180 using the correct key (the original address) when searching the retry
1181 rules in order to find which one to use for generating the retry hint.
1183 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1184 adding the default one. Similarly, if it contains a Reply-To: header, the
1185 errors_reply_to option, if set, is not used.
1187 PH/19 When calculating a retry time, Exim used to measure the "time since
1188 failure" by looking at the "first failed" field in the retry record. Now
1189 it does not use this if it is later than than the arrival time of the
1190 message. Instead it uses the arrival time. This makes for better
1191 behaviour in cases where some deliveries succeed, thus re-setting the
1192 "first failed" field. An example is a quota failure for a huge message
1193 when small messages continue to be delivered. Without this change, the
1194 "time since failure" will always be short, possible causing more frequent
1195 delivery attempts for the huge message than are intended.
1196 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1198 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1199 $1, $2, $3) because the numerical variables can be reset during some
1200 expansion items (e.g. "match"), thereby losing the authentication data.
1202 PH/21 Make -bV show the size of off_t variables so that the test suite can
1203 decide whether to run tests for quotas > 2G.
1205 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1206 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1207 filecount value is greater than 2G or if a quota value is greater than 2G
1208 on a system where the size of off_t is not greater than 4, a panic error
1211 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1212 never match. The debug and -bh output now contains an explicit error
1213 message indicating a malformed IPv4 address or mask.
1215 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1216 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1217 PH/23 above applies.
1219 PH/25 Do not write to syslog when running in the test harness. The only
1220 occasion when this arises is a failure to open the main or panic logs
1221 (for which there is an explicit test).
1223 PH/26 Added the /no_tell option to "control=freeze".
1225 PH/27 If a host name lookup failed very early in a connection, for example, if
1226 the IP address matched host_lookup and the reverse lookup yielded a name
1227 that did not have a forward lookup, an error message of the form "no IP
1228 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1229 could be logged. Now it outputs the IP address instead of "NULL".
1231 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1232 allows the sender and the authenticated sender to be set when
1233 submitting a message from within Exim. Since child_open_exim() is
1234 documented for local_scan(), the new function should be too.
1236 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1237 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1238 results in an empty string is now treated as unset.
1240 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1242 PH/31 Added sender_verify_fail logging option.
1244 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1245 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1246 tidied the source and removed it altogether.
1248 PH/33 When a queue run was abandoned because the load average was too high, a
1249 log line was always written; now it is written only if the queue_run log
1250 selector is set. In addition, the log line for abandonment now contains
1251 information about the queue run such as the pid. This is always present
1252 in "start" and "stop" lines but was omitted from the "abandon" line.
1254 PH/34 Omit spaces between a header name and the colon in the error message that
1255 is given when verify = headers_syntax fails (if there are lots of them,
1256 the message gets confusing).
1258 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1259 names, as there are now some PTR records that contain slashes. This check
1260 is only to protect against broken name servers that fall over on strange
1261 characters, so the fact that it applies to all lookups doesn't matter.
1263 PH/36 Now that the new test suite is complete, we can remove some of the
1264 special code in Exim that was needed for the old test suite. For example,
1265 sorting DNS records because real resolvers return them in an arbitrary
1266 order. The new test suite's fake resolver always returns records in the
1269 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1270 bounces) except when queue_only is set, to avoid logging races between
1271 the different processes.
1273 PH/38 Panic-die if .include specifies a non-absolute path.
1275 PH/39 A tweak to the "H" retry rule from its user.
1277 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1278 a label. They prevented compilation on older perls.
1280 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1281 a warning to be raised on newish perls.
1283 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1284 on queue. Changes to match documented behaviour of showing count of
1285 messages matching specified criteria.
1287 PH/40 Changed the default ident timeout from 30s to 5s.
1289 PH/41 Added support for the use of login_cap features, on those BSD systems
1290 that have them, for controlling the resources used by pipe deliveries.
1292 PH/42 The content-scanning code uses fopen() to create files in which to put
1293 message data. Previously it was not paying any attention to the mode of
1294 the files. Exim runs with umask(0) because the rest of the code creates
1295 files with open(), and sets the required mode explicitly. Thus, these
1296 files were ending up world-writeable. This was not a big issue, because,
1297 being within the spool directory, they were not world-accessible. I have
1298 created a function called modefopen, which takes an additional mode
1299 argument. It sets umask(777), creates the file, chmods it to the required
1300 mode, then resets the umask. All the relevant calls to fopen() in the
1301 content scanning code have been changed to use this function.
1303 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1304 to 24 hours. This avoids potential overflow problems when processing G
1305 and H retry rules. I suspect nobody ever tinkers with this value.
1307 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1309 PH/45 When the plaintext authenticator is running as a client, the server's
1310 challenges are checked to ensure they are valid base64 strings. By
1311 default, the authentication attempt is cancelled if an invalid string is
1312 received. Setting client_ignore_invalid_base64 true ignores these errors.
1313 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1314 they are received. Thus, the responses can be made to depend on the
1315 challenges. If an invalid string is ignored, an empty string is placed in
1318 PH/46 Messages that are created by the autoreply transport now contains a
1319 References: header, in accordance with RFCs 2822 and 3834.
1321 PH/47 Added authenticated_sender_force to the smtp transport.
1323 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1325 PH/49 Installed latest patch from the Sieve maintainer.
1327 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1328 was also set, the appendfile transport was unnecessarily scanning a
1329 directory of message files (e.g. for maildir delivery) to find the count
1330 of files (along with the size), even though it did not need this
1331 information. It now does the scan only if it needs to find either the
1332 size of the count of files.
1334 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1336 PH/52 Two bugs concerned with error handling when the smtp transport is
1339 (i) Exim was not creating retry information for temporary errors given
1340 for individual recipients after the DATA command when the smtp transport
1341 was used in LMTP mode. This meant that they could be retried too
1342 frequently, and not timed out correctly.
1344 (ii) Exim was setting the flag that allows error details to be returned
1345 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1346 recipients that were returned after the DATA command.
1348 PH/53 This is related to PH/52, but is more general: for any failing address,
1349 when detailed error information was permitted to be returned to the
1350 sender, but the error was temporary, then after the final timeout, only
1351 "retry timeout exceeded" was returned. Now it returns the full error as
1352 well as "retry timeout exceeded".
1354 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1355 do this, and (what is worse) MTAs that accept it.
1357 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1358 will now be deprecated.
1360 PH/56 New os.c-cygwin from the Cygwin maintainer.
1362 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1363 formats (previously only available in exim formats via -bpr, -bpru,
1364 and -bpra. Now also available in native and exiqgrep formats)
1366 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1367 with very large, slow to parse queues
1369 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1371 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1373 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1374 responses to authentication challenges, though it was showing the
1375 challenges; (ii) I've removed the CR characters from the debug output for
1378 PH/58 Allow for the insertion of a newline as well as a space when a string
1379 is turned into more than one encoded-word during RFC 2047 encoding. The
1380 Sieve code now uses this.
1382 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1383 data_4xx, lost_connection, tls_required.
1385 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1386 message was being sent as an SMTP response.
1388 PH/61 Add -l and -k options to exicyclog.
1390 PH/62 When verifying, if an address was redirected to one new address, so that
1391 verification continued, and the new address failed or deferred after
1392 having set something in $address_data, the value of $address_data was not
1393 passed back to the ACL. This was different to the case when no
1394 redirection occurred. The value is now passed back in both cases.
1396 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1397 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1398 use login_cap.h, so on its own it isn't the distinguishing feature. The
1399 new name refers directly to the setclassresources() function.
1401 PH/65 Added configuration files for NetBSD3.
1403 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1405 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1406 is preferred over IPv4.
1408 PH/68 The bounce_return_message and bounce_return_body options were not being
1409 honoured for bounces generated during the reception of non-SMTP messages.
1410 In particular, this applied to messages rejected by the ACL. This bug has
1411 been fixed. However, if bounce_return_message is true and bounce_return_
1412 body is false, the headers that are returned for a non-SMTP message
1413 include only those that have been read before the error was detected.
1414 (In the case of an ACL rejection, they have all been read.)
1416 PH/69 The HTML version of the specification is now built in a directory called
1417 spec_html instead of spec.html, because the latter looks like a path with
1418 a MIME-type, and this confuses some software.
1420 PH/70 Catch two compiler warnings in sieve.c.
1422 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1423 function verify_get_ident() calls ip_connect() to connect a socket, but
1424 if the "connect()" function timed out, ip_connect() used to close the
1425 socket. However, verify_get_ident() also closes the socket later, and in
1426 between Exim writes to the log, which may get opened at this point. When
1427 the socket was closed in ip_connect(), the log could get the same file
1428 descriptor number as the socket. This naturally causes chaos. The fix is
1429 not to close the socket in ip_connect(); the socket should be closed by
1430 the function that creates it. There was only one place in the code where
1431 this was missing, in the iplookup router, which I don't think anybody now
1432 uses, but I've fixed it anyway.
1434 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1435 well as to direct DNS lookups. Otherwise the handling of names in host
1436 lists is inconsistent and therefore confusing.
1442 PH/01 Two changes to the default runtime configuration:
1444 (1) Move the checks for relay_from_hosts and authenticated clients from
1445 after to before the (commented out) DNS black list checks.
1447 (2) Add control=submission to the relay_from_hosts and authenticated
1448 clients checks, on the grounds that messages accepted by these
1449 statements are most likely to be submissions.
1451 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1453 (1) Generate an error if the third argument for the ${prvs expansion is
1456 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1459 (3) Reset the variables that are obtained from the first argument of
1460 ${prvscheck and used in the second argument before leaving the code,
1461 because their memory is reclaimed, so using them afterwards may do
1464 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1465 one (it's much easier than Tom thought :-).
1467 (5) Because of (4), we can now allow for the use of $prvscheck_result
1468 inside the third argument.
1470 PH/03 For some reason, the default setting of PATH when running a command from
1471 a pipe transport was just "/usr/bin". I have changed it to
1474 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1475 anything to be listed in the output from -bV.
1477 PH/05 When a filter generated an autoreply, the entire To: header line was
1478 quoted in the delivery log line, like this:
1480 => >A.N.Other <ano@some.domain> <original@ddress> ...
1482 This has been changed so that it extracts the operative address. There
1483 may be more than one such address. If so, they are comma-separated, like
1486 => >ano@some.domain,ona@other.domain <original@ddress> ...
1488 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1489 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1490 not being looked up in the rDNS to get a host name, Exim was showing the
1491 IP address twice in Received: lines, even though the IP addresses were
1492 identical. For example:
1494 Received: from [1.2.3.4] (helo=[1.2.3.4])
1496 However, if the real host name was known, it was omitting the HELO data
1497 if it matched the actual IP address. This has been tidied up so that it
1498 doesn't show the same IP address twice.
1500 PH/07 When both +timestamp and +memory debugging was on, the value given by
1501 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1502 called by the debug printing, thereby overwriting the timestamp buffer.
1503 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1505 PH/08 When the original message was included in an autoreply transport, it
1506 always said "this is a copy of the message, including all the headers",
1507 even if body_only or headers_only was set. It now gives an appropriate
1510 PH/09 Applied a patch from the Sieve maintainer which:
1512 o fixes some comments
1513 o adds the (disabled) notify extension core
1514 o adds some debug output for the result of if/elsif tests
1515 o points to the current vacation draft in the documentation
1516 and documents the missing references header update
1520 o fixes a bug in processing the envelope test (when testing
1521 multiple envelope elements, the last element determinted the
1524 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1525 Electronic Mail") by including:
1527 Auto-submitted: auto-generated
1529 in the messages that it generates (bounce messages and others, such as
1530 warnings). In the case of bounce messages for non-SMTP mesages, there was
1531 also a typo: it was using "Auto_submitted" (underscore instead of
1532 hyphen). Since every message generated by Exim is necessarily in response
1533 to another message, thes have all been changed to:
1535 Auto-Submitted: auto-replied
1537 in accordance with these statements in the RFC:
1539 The auto-replied keyword:
1541 - SHOULD be used on messages sent in direct response to another
1542 message by an automatic process,
1544 - MUST NOT be used on manually-generated messages,
1546 - MAY be used on Delivery Status Notifications (DSNs) and Message
1547 Disposition Notifications (MDNs),
1549 - MUST NOT be used on messages generated by automatic or periodic
1550 processes, except for messages which are automatic responses to
1553 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1554 to the default Received: header definition.
1556 PH/12 Added log selector acl_warn_skipped (default on).
1558 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1559 variables because (a) they are in the wrong storage pool and (b) even if
1560 they were copied, it wouldn't work properly because of the caching.
1562 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1563 checking when decoding. Apparently there are clients that generate
1564 overlong encoded strings. Why am I not surprised?
1566 PH/15 If the first argument of "${if match_address" was not empty, but did not
1567 contain an "@" character, Exim crashed. Now it writes a panic log message
1568 and treats the condition as false.
1570 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1572 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1573 extension "envelope-auth". The code is finished and in agreement with
1574 other implementations, but there is no documentation so far and in fact,
1575 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1576 not changing the active code.
1578 Print executed "if" and "elsif" statements when debugging is used. This
1579 helps a great deal to understand what a filter does.
1581 Document more things not specified clearly in RFC3028. I had all this
1582 sorted out, when out of a sudden new issues came to my mind. Oops."
1584 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1587 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1588 canonical form (as documented). However, after a host name lookup from
1589 the IP address, check_host() was doing a simple string comparison with
1590 addresses acquired from the DNS when checking that the found name did
1591 have the original IP as one of its addresses. Since any found IPv6
1592 addresses are likely to be in abbreviated form, the comparison could
1593 fail. Luckily, there already exists a function for doing the comparison
1594 by converting both addresses to binary, so now that is used instead of
1595 the text comparison.
1597 PH/20 There was another similar case to PH/19, when a complete host name was
1598 given in a host list; looking up its IP address could give an abbreviated
1599 form, whereas the current host's name might or might not be abbreviated.
1600 The same fix has been applied.
1606 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1607 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1610 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1611 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1613 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1615 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1616 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1617 Also, for "Auto-Submitted", treat anything other than "no" as
1618 non-personal, in accordance with RFC 3834. (Previously it treated
1619 anything starting "auto-" as non-personal.)
1621 TF/01 The control=submission/name=... option had a problem with syntax
1622 errors if the name included a slash character. The /name= option
1623 now slurps the rest of the string, so it can include any characters
1624 but it must come last in the list of options (after /sender_retain
1627 PH/05 Some modifications to the interface to the fake nameserver for the new
1635 TK/01 Added the "success_on_redirect" address verification option. See
1636 NewStuff for rationale and an example.
1638 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1640 PH/02 Patch to exigrep to allow it to work on syslog lines.
1642 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1643 fread() to skip over the body file's header line, because in Cygwin the
1644 header line is locked and is inaccessible.
1646 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1647 co-exist for some time) to make it clear that it is the Exim ID that is
1648 referenced, not the Message-ID: header line.
1650 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1651 string_format() function, because snprintf() does not exist on all
1654 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1655 ${stat: expansion item.
1657 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1658 protocol synchronization error", to keep the pedants happy.
1660 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1661 well as for IRIX systems, when gcc is being used. See the host.c source
1664 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1666 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1669 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1670 transport and to the smtp transport in LMTP mode.
1672 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1674 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1675 run a filter in a subprocess. This could lead to confusion in subsequent
1676 lookups in the parent process. There should also be a search_tidyup() at
1677 the end of the subprocess.
1679 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1680 only if the host matched helo_try_verify_hosts, which caused the
1681 verification to occur when the EHLO/HELO command was issued. The ACL just
1682 tested the remembered result. Now, if a previous verification attempt has
1683 not happened, "verify = helo" does it there and then.
1685 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1687 TK/03 Fix log output including CR from clamd.
1689 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1690 exist provoked a memory error which could cause a segfault.
1692 PH/15 Installed PCRE 6.2
1694 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1696 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1697 of the problem. Specifically, suggested +O2 rather than +O1 for the
1700 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1702 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1703 record, the port was not correctly logged when the outgoing_port log
1704 selector was set (it logged the transort's default port).
1706 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1707 fallback_hosts, and "hosts" in the smtp transport.
1709 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1710 host errors such as "Connection refused".
1712 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1713 authentication with radiusclient 0.4.9:
1715 - Error returned from rc_read_config was caught wrongly
1716 - Username/password not passed on to radius server due to wrong length.
1718 The presumption is that some radiusclient API changes for 4.51/PH/17
1719 were not taken care of correctly. The code is still untested by me (my
1720 Linux distribution still has 0.3.2 of radiusclient), but it was
1721 contributed by a Radius user.
1723 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1724 expanding the "port" option of the smtp transport.
1726 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1727 while reading a MIME header. Thanks to Tom Hughes for a patch.
1729 PH/24 Include config.h inside local_scan.h so that configuration settings are
1732 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1733 that in an ACL for RCPT (for example), you can examine exactly what was
1736 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1737 commands, but it was not correctly comparing the address with the actual
1738 client host address. Thus, it would show the EHLO address in Received:
1739 header lines when this was not necessary.
1741 PH/27 Added the % operator to ${eval:}.
1743 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1744 it should be ignoring failures (because with -C, for example, it has lost
1745 privilege). It wasn't ignoring creation failures other than "already
1748 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1751 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1752 because an input line was too long, either on its own, or by virtue of
1753 too many continuations, the temporary file was not being removed, and the
1754 return code was incorrect.
1756 PH/31 Missing "BOOL" in function definition in filtertest.c.
1758 PH/32 Applied Sieve patches from the maintainer.
1760 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1762 PH/33 Added "verify = not_blind".
1764 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1765 Local/Makefile (with some defaults set). These are used in built scripts
1766 such as exicyclog, but they have never been used in the exim_install
1767 script (though there are many overriding facilities there). I have
1768 arranged that the exim_install script now takes note of these two
1771 PH/35 Installed configuration files for Dragonfly.
1773 PH/36 When a locally submitted message by a trusted user did not contain a
1774 From: header, and the sender address was obtained from -f or from an SMTP
1775 MAIL command, and the trusted user did not use -F to supply a sender
1776 name, $originator_name was incorrectly used when constructing a From:
1777 header. Furthermore, $originator_name was used for submission mode
1778 messages from external hosts without From: headers in a similar way,
1779 which is clearly wrong.
1781 PH/37 Added control=suppress_local_fixups.
1783 PH/38 When log_selector = +received_sender was set, and the addition of the
1784 sender made the log line's construction buffer exactly full, or one byte
1785 less than full, an overflow happened when the terminating "\n" was
1788 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1789 when the result of a list match is failure because a DNS lookup failed.
1791 PH/40 RM_COMMAND is now used in the building process.
1793 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1794 the "build-* directories that it finds.
1796 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1797 address in a domain literal was a prefix of an interface address.
1799 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1800 when verifying a sender address, unless rewrite_headers is false.
1802 PH/44 Wrote a long comment about why errors_to addresses are verified as
1803 recipients, not senders.
1805 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1806 the ratelimit ACL was added.
1808 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1810 PH/46 Added extra information about PostgreSQL errors to the error string.
1812 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1813 avoiding the need to install special zones in a real server. This is
1814 backwards compatible; if it can't find the fake resolver, it drops back.
1815 Thus, both old and new test suites can be run.
1817 TF/02 Added util/ratelimit.pl
1819 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1820 clock is set back in time.
1822 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1823 Candler <B.Candler@pobox.com>.
1825 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1826 OK for addresses that are the result of redirections.
1828 PH/48 A number of further additions for the benefit of the new test suite,
1829 including a fake gethostbyname() that interfaces to the fake DNS resolver
1832 TF/06 The fix for widen_domains has also been applied to qualify_single and
1833 search_parents which are the other dnslookup options that can cause
1836 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1839 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1841 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1842 service" if it encountered an MX record with an empty target hostname.
1843 The message is now "an MX or SRV record indicated no SMTP service".
1845 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1846 if the DNS of the sending site is misconfigured. This is quite a
1847 common situation. This change restores the behaviour of treating a
1848 helo verification defer as a failure.
1850 PH/51 If self=fail was set on a router, the bounce message did not include the
1851 actual error message.
1857 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1859 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1860 command itself ended in error, the underlying message about the transport
1861 filter timeout was being overwritten with the pipe command error. Now the
1862 underlying error message should be appended to the second error message.
1864 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1866 PH/02 Reduce the amount of output that "make" produces by default. Full output
1867 can still be requested.
1869 PH/03 The warning log line about a condition test deferring for a "warn" verb
1870 was being output only once per connection, rather than after each
1871 occurrence (because it was using the same function as for successful
1872 "warn" verbs). This seems wrong, so I have changed it.
1874 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1875 it should not have, which might have caused a crash in the right
1876 circumstances, but probably never did.
1878 PH/04 Installed a modified version of Tony Finch's patch to make submission
1879 mode fix the return path as well as the Sender: header line, and to
1880 add a /name= option so that you can make the user's friendly name appear
1883 TF/03 Added the control = fakedefer ACL modifier.
1885 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1886 Mark Lowes for thorough testing.
1888 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1890 TK/03 Merged latest SRS patch from Miles Wilton.
1892 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1893 with the definition in sysexits.h (which is #included earlier).
1894 Fortunately, Exim does not actually use EX_OK. The code used to try to
1895 preserve the sysexits.h value, by assumimg that macro definitions were
1896 scanned for macro replacements. I have been disabused of this notion,
1897 so now the code just undefines EX_OK before #including unistd.h.
1899 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1900 in the smtp transport. When a block could not be written in a single
1901 write() function, the timeout was being re-applied to each part-write.
1902 This seems wrong - if the receiver was accepting one byte at a time it
1903 would take for ever. The timeout is now adjusted when this happens. It
1904 doesn't have to be particularly precise.
1906 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1907 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1909 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1910 without a domain if the check to <postmaster@domain> fails.
1912 SC/01 Eximstats: added -xls and the ability to specify output files
1913 (patch written by Frank Heydlauf).
1915 SC/02 Eximstats: use FileHandles for outputing results.
1917 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1919 SC/04 Eximstats: fixed display of large numbers with -nvr option
1921 SC/05 Eximstats: fixed merging of reports with empty tables.
1923 SC/06 Eximstats: added the -include_original_destination flag
1925 SC/07 Eximstats: removed tabs and trailing whitespace.
1927 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1929 TK/06 MBOX spool code: Add real "From " MBOX separator line
1930 so the .eml file is really in mbox format (even though
1931 most programs do not really care). Patch from Alex Miller.
1933 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1934 The latter is generated from $received_to and is only set if the
1935 message has one envelope recipient. SA can use these headers,
1936 obviously out-of-the-box. Patch from Alex Miller.
1938 PH/08 The ${def test on a variable was returning false if the variable's
1939 value was "0", contrary to what the specification has always said!
1940 The result should be true unless the variable is empty.
1942 PH/09 The syntax error of a character other than { following "${if
1943 def:variable_name" (after optional whitespace) was not being diagnosed.
1944 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1945 accidental colon was present, for example, could give incorrect results.
1947 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1948 result is used (not including appendfile, where other changes are about
1951 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1952 This involved changing a lot of size variables from int to off_t. It
1953 should work with maildirs and everything.
1955 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1956 spamd dying while we are connected to it.
1958 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1961 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1962 function for BDB 4.3.
1964 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1966 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1967 the sources for PCRE have been re-arranged and more files are now
1970 PH/15 The code I had for printing potentially long long variables in PH/11
1971 above was not the best (it lost precision). The length of off_t variables
1972 is now inspected at build time, and an appropriate printing format (%ld
1973 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1974 to be "long long int" or "long int". This is needed for the internal
1975 formatting function string_vformat().
1977 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1978 the configuration file to be ":syslog", then the script "guesses" where
1979 the logs files are, rather than using the compiled in default. In our
1980 case the guess is not the same as the compiled default, so the script
1981 suddenly stopped working when I started to use syslog. The patch checks
1982 to see if log_file_path is "". If so, it attempts to read it from exim
1983 with no configuration file to get the compiled in version, before it
1984 falls back to the previous guessing code."
1986 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1987 implementing BATV in an Exim configuration. See NewStuff for the gory
1990 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1991 Makefile that are specific to HP-UX.
1993 PH/18 If the "use_postmaster" option was set for a recipient callout together
1994 with the "random" option, the postmaster address was used as the MAIL
1995 FROM address for the random test, but not for the subsequent recipient
1996 test. It is now used for both.
1998 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1999 patch removes a few documentation additions to RFC 3028, because the
2000 latest draft now contains them. It adds the new en;ascii-case comparator
2001 and a new error check for 8bit text in MIME parts. Comparator and
2002 require names are now matched exactly. I enabled the subaddress
2003 extension, but it is not well tested yet (read: it works for me)."
2005 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2006 rework some of the code of TK/09 above to avoid the hardwired use of
2007 "%lld" and "long long". Replaced the call to snprintf() with a call to
2010 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2011 records point to non-existent hosts", "retry timeout exceeded", and
2012 "retry time not reached for any host after a long failure period".
2014 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2015 experimental DomainKeys support:
2017 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2018 (2) On an error such as an illegally used "control", the wrong name for
2019 the control was given.
2021 These problems did NOT occur unless DomainKeys support was compiled.
2023 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2025 PH/24 Added ${if match_ip condition.
2027 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2028 databases so that it will be absolutely obvious if a crash occurs in the
2029 DB library. This is a regular occurrence (often caused by mis-matched
2032 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2033 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2034 picked up on a user's system that detects such things. There doesn't seem
2035 to be a gcc warning option for this - only an attribute that has to be
2036 put on the function's prototype. It seems that in Fedora Core 4 they have
2037 set this on a number of new functions. No doubt there will be more in due
2040 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2041 specify a transport. However, if an address that was verified by such a
2042 router was the subject of a callout, Exim crashed because it tried to
2043 read the rcpt_include_affixes from the non-existent transport. Now it
2044 just assumes that the setting of that option is false. This bug was
2045 introduced by 4.51/PH/31.
2047 PH/28 Changed -d+all to exclude +memory, because that information is very
2048 rarely of interest, but it makes the output a lot bigger. People tend to
2049 do -d+all out of habit.
2051 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2052 code in os-type was giving problems when libc.so lives in lib64, like on
2055 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2056 aren't the modern standard, and it seems that some systems' include files
2057 don't always have them. Exim was already checking for some of the newer
2058 ones like T_AAAA, and defining it itself. I've added checks for all the
2059 record types that Exim uses.
2061 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2062 not automatically generating a new one, as it is supposed to. This
2063 prevented TLS from working. If the file did exist, but contained invalid
2064 data, a new version was generated, as expected. It was only the case of a
2065 non-existent file that was broken.
2067 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2068 with a change in libdomainkeys > 0.64.
2070 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2071 from DNS. If the selector record carries the flag, it now has
2072 precedence over the domain-wide flag.
2074 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2076 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2077 the use of a port name that isn't defined in /etc/services) occurred, the
2078 message was deferred as in a normal delivery, and thus remained on the
2079 spool, instead of being failed because of the mua_wrapper setting. This
2080 is now fixed, and I tidied up some of the mua_wrapper messages at the
2083 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2084 the messages in a hash of arrays rather than using individual hashes.
2085 This is a bit cleaner and results in dramatic memory savings, albeit
2086 at a slight CPU cost.
2088 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2089 as requested by Marc Sherman.
2091 SC/10 Eximstats: added histograms for user specified patterns as requested
2094 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2096 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2097 fopen() in the content-scanning modules that did not already have it.
2103 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2104 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2106 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2108 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2110 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2111 to test Sieve filters that use "vacation".
2113 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2114 that changes the way the GnuTLS parameters are stored in the cache file.
2115 The new format can be generated externally. For backward compatibility,
2116 if the data in the cache doesn't make sense, Exim assumes it has read an
2117 old-format file, and it generates new data and writes a new file. This
2118 means that you can't go back to an older release without removing the
2121 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2122 work if there are any delivery delays because "one_time" forces the
2123 parent to be marked "delivered", so its unseen clone is never tried
2124 again. For this reason, Exim now forbids the simultaneous setting of
2127 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2128 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2129 entry if you want to know the details. The fix, however, neglected to
2130 consider the case when local delivery batching is involved. The test for
2131 "previously delivered" was not happening when checking to see if an
2132 address could be batched with a previous (undelivered) one; under
2133 certain circumstances this could lead to multiple deliveries to the same
2136 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2137 in its include files, and this causes problems building Exim.
2139 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2140 header_syntax) but Exim was just ignoring anything given after a slash.
2141 In particular, this caused confusion with an attempt to use "verify =
2142 reverse_host_lookup/defer_ok". An error is now given when options are
2143 supplied for verify items that do not have them. (Maybe reverse_host_
2144 lookup should have a defer_ok option, but that's a different point.)
2146 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2147 defined by RFC 821) to 2048, because there were problems with some AUTH
2148 commands, and RFC 1869 says the size should be increased for extended
2149 SMTP commands that take arguments.
2151 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2154 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2155 "unknown" error; now it says that the functionality isn't in the binary.
2157 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2158 an address' error message when a string expansion fails (syntax or
2159 whatever). Otherwise the password may appear in the log. Following change
2160 PH/42 below, there is no longer a chance of it appearing in a bounce
2163 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2165 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2166 of its addresses was ever tried. (Bugzilla bug #2.)
2168 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2169 the result incorrectly in the debug output. (It correctly added a newline
2170 to what was transported.)
2172 TF/01 Added $received_time.
2174 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2175 commented out examples of how to interface to a virus scanner and to
2176 SpamAssassin. Also added commented examples of av_scanner and
2177 spamd_address settings.
2179 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2180 and controls are allowed in which ACLs. There were a couple of minor
2181 errors. Some of the entries in the conditions table (which is a table of
2182 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2183 negation of where the condition IS allowed.
2185 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2187 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2188 header file does not have a version number, so I've had to invent a new
2189 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2190 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2191 radiusclient), but it was contributed by a Radius user.
2193 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2194 files or directories, for OpenSSL.
2196 PH/19 When an Exim process that is running as root has to create an Exim log
2197 file, it does so in a subprocess that runs as exim:exim so as to get the
2198 ownership right at creation (otherwise, other Exim processes might see
2199 the file with the wrong ownership). There was no test for failure of this
2200 fork() call, which would lead to the process getting stuck as it waited
2201 for a non-existent subprocess. Forks do occasionally fail when resources
2202 run out. I reviewed all the other calls to fork(); they all seem to check
2205 PH/20 When checking for unexpected SMTP input at connect time (before writing
2206 the banner), Exim was not dealing correctly with a non-positive return
2207 from the read() function. If the client had disconnected by this time,
2208 the result was a log entry for a synchronization error with an empty
2209 string after "input=" when read() returned zero. If read() returned -1
2210 (an event I could not check), uninitialized data bytes were printed.
2211 There were reports of junk text (parts of files, etc) appearing after
2214 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2216 PH/22 Added support for macro redefinition, and (re)definition in between
2217 driver and ACL definitions.
2219 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2220 forgetting to use the resulting value; it was using the unexpanded value.
2222 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2223 hadn't been configured. The fix is from Juergen Kreileder, who
2224 understands it better than I do:
2226 "Here's what I see happening with three configured cyrus_sasl
2227 authenticators configured (plain, login, cram-md5):
2229 On startup auth_cyrus_sasl_init() gets called for each of these.
2230 This means three calls to sasl_listmech() without a specified mech_list.
2231 => SASL tests which mechs of all available mechs actually work
2232 => three warnings about OTP not working
2233 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2235 With the patch, sasl_listmech() also gets called three times. But now
2236 SASL's mech_list option is set to the server_mech specified in the the
2237 authenticator. Or in other words, the answer from sasl_listmech()
2238 gets limited to just the mech you're testing for (which is different
2240 => the return list contains just 'plain' or 'login', 'cram-md5' or
2241 nothing depending on the value of ob->server_mech.
2243 I've just tested the patch: Authentication still works fine,
2244 unavailable mechs specified in the exim configuration are still
2245 caught, and the auth.log warnings about OTP are gone."
2247 PH/25 When debugging is enabled, the contents of the command line are added
2248 to the debugging output, even when log_selector=+arguments is not
2251 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2252 answer is "GNU", and only if the return is "GNU/something" is the answer
2255 PH/27 $acl_verify_message is now set immediately after the failure of a
2256 verification in an ACL, and so is available in subsequent modifiers. In
2257 particular, the message can be preserved by coding like this:
2259 warn !verify = sender
2260 set acl_m0 = $acl_verify_message
2262 Previously, $acl_verify_message was set only while expanding "message"
2263 and "log_message" when a very denied access.
2265 PH/28 Modified OS/os.c-Linux with
2267 -#ifndef OS_LOAD_AVERAGE
2268 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2270 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2271 nomenclature these days.)
2273 PH/29 Installed patch from the Sieve maintainer that adds the options
2274 sieve_useraddress and sieve_subaddress to the redirect router.
2276 PH/30 In these circumstances:
2277 . Two addresses routed to the same list of hosts;
2278 . First host does not offer TLS;
2279 . First host accepts first address;
2280 . First host gives temporary error to second address;
2281 . Second host offers TLS and a TLS session is established;
2282 . Second host accepts second address.
2283 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2284 and peerdn, if requested) that were in fact used only for the second
2287 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2288 attention to any local part prefix or suffix that was matched by the
2289 router that accepted the address. It now behaves in the same way as it
2290 does for delivery: the affixes are removed from the local part unless
2291 rcpt_include_affixes is set on the transport.
2293 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2294 timeout during the DATA phase of an incoming message.
2296 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2297 applied a patch sanctioned by the Sieve maintainer.
2299 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2300 the uid or gid is negative. A case of a negative gid caused this to be
2301 noticed. The fix allows for either to be negative.
2303 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2304 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2307 PH/36 The change PH/12 above was broken. Fixed it.
2309 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2310 the grounds that routing the same address twice would always produce the
2311 same answer. This might have been true once, but it is certainly no
2312 longer true now. Routing a child address may depend on the previous
2313 routing that produced that child. Some complicated redirection strategies
2314 went wrong when messages had multiple recipients, and made Exim's
2315 behaviour dependent on the order in which the addresses were given.
2317 I have moved the duplicate checking until after the routing is complete.
2318 Exim scans the addresses that are assigned to local and remote
2319 transports, and removes any duplicates. This means that more work will be
2320 done, as duplicates will always all be routed, but duplicates are
2321 presumably rare, so I don't expect this is of any significance.
2323 For deliveries to pipes, files, and autoreplies, the duplicate checking
2324 still happens during the routing process, since they are not going to be
2327 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2328 It corrects a timeout issue with spamd. This is Ian's comment: "The
2329 background is that sometimes spamd either never reads data from a
2330 connection it has accepted, or it never writes response data. The exiscan
2331 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2332 blindly assumes that writes won't block so it may never time out."
2334 PH/39 Allow G after quota size as well as K and M.
2336 PH/40 The value set for $authenticated_id in an authenticator may not contain
2337 binary zeroes or newlines because the value is written to log lines and
2338 to spool files. There was no check on this. Now the value is run through
2339 the string_printing() function so that such characters are converted to
2340 printable escape sequences.
2342 PH/41 $message_linecount is a new variable that contains the total number of
2343 lines in the message. Compare $body_linecount, which is the count for the
2346 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2347 bounce and delay warning messages, except in certain special cases, which
2350 (a) An SMTP error message from a remote host;
2351 (b) A message specified in a :fail: redirection;
2352 (c) A message specified in a "fail" command in a system filter;
2353 (d) A message specified in a FAIL return from the queryprogram router;
2354 (e) A message specified by the cannot_route_message router option.
2356 In these cases only, Exim does include the error details in bounce and
2357 warning messages. There are also a few cases where bland messages such
2358 as "unrouteable address" or "local delivery error" are given.
2360 PH/43 $value is now also set for the "else" part of a ${run expansion.
2362 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2363 being worked on, but at least Exim now implements the latest version to
2366 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2367 process to complete was treated as a delivery failure, a timeout while
2368 writing the message to the pipe was logged, but erroneously treated as a
2369 successful delivery. Such timeouts include transport filter timeouts. For
2370 consistency with the overall process timeout, these timeouts are now
2371 treated as errors, giving rise to delivery failures by default. However,
2372 there is now a new Boolean option for the pipe transport called
2373 timeout_defer, which, if set TRUE, converts the failures into defers for
2374 both kinds of timeout. A transport filter timeout is now identified in
2377 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2378 systems where "make" and "gmake" are different, calling "gmake" at top
2379 level broke things. I've arranged for the value of $(MAKE) to be passed
2380 from the Makefile to this script so that it can call the same version of
2384 A note about Exim versions 4.44 and 4.50
2385 ----------------------------------------
2387 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2388 changes of various kinds. As a consequence, a big documentation update was
2389 needed. This delayed the release for rather longer than seemed good, especially
2390 in the light of a couple of (minor) security issues. Therefore, the changes
2391 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2392 release. So 4.44 and 4.50 are in effect two different branches that both start
2395 I have left the 4.50 change log unchanged; it contains all the changes since
2396 4.43. The change log for 4.44 is below; many of its items are identical to
2397 those for 4.50. This seems to be the most sensible way to preserve the
2398 historical information.
2404 1. Minor wording change to the doc/README.SIEVE file.
2406 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2407 computation of the current number of files was incorrect.
2409 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2410 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2413 4. Give more explanation in the error message when the command for a transport
2414 filter fails to execute.
2416 5. There are several places where Exim runs a non-Exim command in a
2417 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2418 was being done only for the command run by the queryprogram router. It is
2419 now done for all such subprocesses. The other cases are: ${run, transport
2420 filters, and the commands run by the lmtp and pipe transports.
2422 6. Added CONFIGURE_GROUP build-time option.
2424 7. Some older OS have a limit of 256 on the maximum number of file
2425 descriptors. Exim was using setrlimit() to set 1000 as a large value
2426 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2427 systems. I've change it so that if it can't get 1000, it tries for 256.
2429 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2430 was an oversight, and furthermore, ever since the addition of extra
2431 controls (e.g. 4.43/32), the checks on when to allow different forms of
2432 "control" were broken. There should now be diagnostics for all cases when a
2433 control that does not make sense is encountered.
2435 9. Added the /retain_sender option to "control=submission".
2437 10. $recipients is now available in the predata ACL (oversight).
2439 11. Tidy the search cache before the fork to do a delivery from a message
2440 received from the command line. Otherwise the child will trigger a lookup
2441 failure and thereby defer the delivery if it tries to use (for example) a
2442 cached ldap connection that the parent has called unbind on.
2444 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2445 of $address_data from the recipient verification was clobbered by the
2446 sender verification.
2448 13. The value of address_data from a sender verification is now available in
2449 $sender_address_data in subsequent conditions in the ACL statement.
2451 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2453 15. Added a new option "connect=<time>" to callout options, to set a different
2456 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2457 was its contents. (It was OK if the option was not defined at all.)
2459 17. A "Completed" log line is now written for messages that are removed from
2460 the spool by the -Mrm option.
2462 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2463 information about exactly what failed.
2465 19. Added -dd to debug only the daemon process.
2467 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2468 handles timeouts, both on the server side and network timeouts. Renamed the
2469 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2471 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2472 It is now set to "smtps".
2474 22. $host_address is now set to the target address during the checking of
2475 ignore_target_hosts.
2477 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2478 being passed; this would have caused $sender_host_name to have been used if
2479 matching the list had actually called for a host name (not very likely,
2480 since this list is usually IP addresses). A host name is now passed as
2483 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2484 code that specifies a non-restarting handler (typically sigaction() in
2485 modern systems) in an attempt to fix a rare and obscure crash bug.
2487 25. Narrowed the window for a race in the daemon that could cause it to ignore
2488 SIGCHLD signals. This is not a major problem, because they are used only to
2489 wake it up if nothing else does.
2491 26. A malformed maildirsize file could cause Exim to calculate negative values
2492 for the mailbox size or file count. Odd effects could occur as a result.
2493 The maildirsize information is now recalculated if the size or filecount
2496 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2497 support for a long time. Removed HAVE_SYS_VFS_H.
2499 28. Installed the latest version of exipick from John Jetmore.
2501 29. In an address list, if the pattern was not a regular expression, an empty
2502 subject address (from a bounce message) matched only if the pattern was an
2503 empty string. Non-empty patterns were not even tested. This was the wrong
2504 because it is perfectly reasonable to use an empty address as part of a
2505 database query. An empty address is now tested by patterns that are
2506 lookups. However, all the other forms of pattern expect the subject to
2507 contain a local part and a domain, and therefore, for them, an empty
2508 address still always fails if the pattern is not itself empty.
2510 30. Exim went into a mad DNS loop when attempting to do a callout where the
2511 host was specified on an smtp transport, and looking it up yielded more
2512 than one IP address.
2514 31. Re-factored the code for checking spool and log partition space into a
2515 function that finds that data and another that does the check. The former
2516 is then used to implement four new variables: $spool_space, $log_space,
2517 $spool_inodes, and $log_inodes.
2519 32. The RFC2047 encoding function was originally intended for short strings
2520 such as real names; it was not keeping to the 75-character limit for
2521 encoded words that the RFC imposes. It now respects the limit, and
2522 generates multiple encoded words if necessary. To be on the safe side, I
2523 have increased the buffer size for the ${rfc2047: expansion operator from
2526 33. It is now permitted to omit both strings after an "if" condition; if the
2527 condition is true, the result is "true". As before, when the second string
2528 is omitted, a false condition yields an empty string. This makes it less
2529 cumbersome to write custom ACL and router conditions.
2531 34. Failure to deliver a bounce message always caused it to be frozen, even if
2532 there was an errors_to setting on the router. The errors_to setting is now
2535 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2536 canonical form (fully expanded) before being placed in
2537 $sender_host_address.
2539 36. The table in the code that translates DNS record types into text (T_A to
2540 "A" for instance) was missing entries for NS and CNAME. It is just possible
2541 that this could have caused confusion if both these types were looked up
2542 for the same domain, because the text type is used as part of Exim's
2543 per-process caching. But the chance of anyone hitting this buglet seems
2546 37. The dnsdb lookup has been extended in a number of ways.
2548 (1) There is a new type, "zns", which walks up the domain tree until it
2549 finds some nameserver records. It should be used with care.
2551 (2) There is a new type, "mxh", which is like "mx" except that it returns
2552 just the host names, not the priorities.
2554 (3) It is now possible to give a list of domains (or IP addresses) to be
2555 looked up. The behaviour when one of the lookups defers can be
2556 controlled by a keyword.
2558 (4) It is now possible to specify the separator character for use when
2559 multiple records are returned.
2561 38. The dnslists ACL condition has been extended: it is now possible to supply
2562 a list of IP addresses and/or domains to be looked up in a particular DNS
2565 39. Added log_selector=+queue_time_overall.
2567 40. When running the queue in the test harness, wait just a tad after forking a
2568 delivery process, to get repeatability of debugging output.
2570 41. Include certificate and key file names in error message when GnuTLS fails
2571 to set them up, because the GnuTLS error message doesn't include the name
2572 of the failing file when there is a problem reading it.
2574 42. Allow both -bf and -bF in the same test run.
2576 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2578 44. The "Exiscan patch" is now merged into the mainline Exim source.
2580 45. Sometimes the final signoff response after QUIT could fail to get
2581 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2582 < 0 before doing a fflush(). This bug looks as though it goes back to the
2583 introduction of TLS in release 3.20, but "sometimes" must have been rare
2584 because the tests only now provoked it.
2586 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2587 (this can affect the format of dates).
2589 47. exim_tidydb, when checking for the continued existence of a message for
2590 which it has found a message-specific retry record, was not finding
2591 messages that were in split spool directories. Consequently, it was
2592 deleting retry records that should have stayed in existence.
2594 48. Steve fixed some bugs in eximstats.
2596 49. The SPA authentication driver was not abandoning authentication and moving
2597 on to the next authenticator when an expansion was forced to fail,
2598 contradicting the general specification for all authenticators. Instead it
2599 was generating a temporary error. It now behaves as specified.
2601 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2602 (the order specifies the preference for clients). The order is now AES256,
2603 AES128, 3DES, ARCFOUR128.
2605 51. Small patch to Sieve code - explicitly set From: when generating an
2608 52. Exim crashed if a remote delivery caused a very long error message to be
2609 recorded - for instance if somebody sent an entire SpamAssassin report back
2610 as a large number of 550 error lines. This bug was coincidentally fixed by
2611 increasing the size of one of Exim's internal buffers (big_buffer) that
2612 happened as part of the Exiscan merge. However, to be on the safe side, I
2613 have made the code more robust (and fixed the comments that describe what
2616 53. Now that there can be additional text after "Completed" in log lines (if
2617 the queue_time_overall log selector is set), a one-byte patch to exigrep
2618 was needed to allow it to recognize "Completed" as not the last thing in
2621 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2622 patch that reportedly fixes this has been added. I am not expert enough to
2623 create a test for it. This is what the patch creator wrote:
2625 "I found a little strange behaviour of ldap code when working with
2626 Windows 2003 AD Domain, where users was placed in more than one
2627 Organization Units. When I tried to give exim partial DN, the exit code
2628 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2629 But simultaneously result of request was absolutely normal ldap result,
2630 so I produce this patch..."
2632 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2633 REFERENCE, so I have modified the code to exclude the patch when that macro
2636 55. Some experimental protocols are using DNS PTR records for new purposes. The
2637 keys for these records are domain names, not reversed IP addresses. The
2638 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2639 leaves it alone. Component reversal etc. now happens only for IP addresses.
2642 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2644 57. Double the size of the debug message buffer (to 2048) so that more of very
2645 long debug lines gets shown.
2647 58. The exicyclog utility now does better if the number of log files to keep
2648 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2650 59. Two changes related to the smtp_active_hostname option:
2652 (1) $smtp_active_hostname is now available as a variable.
2653 (2) The default for smtp_banner uses $smtp_active_hostname instead
2654 of $primary_hostname.
2656 60. The host_aton() function is supposed to be passed a string that is known
2657 to be a valid IP address. However, in the case of IPv6 addresses, it was
2658 not checking this. This is a hostage to fortune. Exim now panics and dies
2659 if the condition is not met. A case was found where this could be provoked
2660 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2661 components; fortuitously, this particular loophole had already been fixed
2662 by change 4.50/55 above.
2664 If there are any other similar loopholes, the new check in host_aton()
2665 itself should stop them being exploited. The report I received stated that
2666 data on the command line could provoke the exploit when Exim was running as
2667 exim, but did not say which command line option was involved. All I could
2668 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2669 running as the user.
2672 61. There was a buffer overflow vulnerability in the SPA authentication code
2673 (which came originally from the Samba project). I have added a test to the
2674 spa_base64_to_bits() function which I hope fixes it.
2677 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2678 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2680 63. The daemon start-up calls getloadavg() while still root for those OS that
2681 need the first call to be done as root, but it missed one case: when
2682 deliver_queue_load_max is set with deliver_drop_privilege. This is
2683 necessary for the benefit of the queue runner, because there is no re-exec
2684 when deliver_drop_privilege is set.
2686 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2687 This has been fixed.
2689 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2690 was in use, was not putting the data itself into the right store pool;
2691 consequently, it could be overwritten for a subsequent message in the same
2692 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2695 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2697 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2698 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2699 the function were treating the return as a boolean value, which happened to
2700 work because 0=false and not-0=true, but is not correct code.
2702 68. The host_aton() function was not handling scoped IPv6 addresses (those
2703 with, for example, "%eth0" on the end) correctly.
2705 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2706 negated items (that is, ~something) in unsigned ints. Some compilers
2707 apparently mutter when there is no cast.
2709 70. If an address verification called from an ACL failed, and did not produce a
2710 user-specific message (i.e. there was only a "system" message), nothing was
2711 put in $acl_verify_message. In this situation, it now puts the system
2714 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2715 session; change 4.31/43 added the unwanted input to the log line - except
2716 that it did not do this in the start of session case. It now does.
2718 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2719 This is wrong and can cause the other end to generate a synchronization
2720 error if it is another Exim or anything else that does the synchronization
2721 check. A QUIT command is no longer sent after a timeout.
2723 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2724 during host lookups.
2726 74. The defer_ok option of callout verification was not working if it was used
2727 when verifying addresses in header lines, that is, for this case:
2729 verify = header_sender/callout=defer_ok
2731 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2732 those file descriptors could be used for SMTP connections. If anything
2733 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2734 could be sent to the SMTP client, causing chaos. The daemon now opens
2735 stdin, stdout, and stderr to /dev/null when it puts itself into the
2738 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2739 log by default. The user can override this with suitable Perl magic.
2741 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2742 the log message when discard triggers, was not working for the DATA ACL or
2743 for the non-SMTP ACL.
2745 78. Error message wording change in sieve.c.
2747 79. If smtp_accept_max_per_host was set, the number of connections could be
2748 restricted to fewer than expected, because the daemon was trying to set up
2749 a new connection before checking whether the processes handling previous
2750 connections had finished. The check for completed processes is now done
2751 earlier. On busy systems, this bug wouldn't be noticed because something
2752 else would have woken the daemon, and it would have reaped the completed
2755 80. If a message was submitted locally by a user whose login name contained one
2756 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2757 It caused a spool format error. I have fixed the spool reading code. A
2758 related problem was that the "from" clause in the Received: line became
2759 illegal because of the space(s). It is now covered by ${quote_local_part.
2761 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2764 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2766 83. Patch from Sieve maintainer to fix unterminated string problem in
2767 "vacation" handling.
2769 84. Some minor changes to the Linux configuration files to help with other
2770 OS variants using glibc.
2772 85. One more patch for Sieve to update vacation handling to latest spec.
2775 ----------------------------------------------------
2776 See the note above about the 4.44 and 4.50 releases.
2777 ----------------------------------------------------
2783 1. Change 4.43/35 introduced a bug that caused file counts to be
2784 incorrectly computed when quota_filecount was set in an appendfile
2787 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2788 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2791 3. Give more explanation in the error message when the command for a transport
2792 filter fails to execute.
2794 4. There are several places where Exim runs a non-Exim command in a
2795 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2796 was being done only for the command run by the queryprogram router. It is
2797 now done for all such subprocesses. The other cases are: ${run, transport
2798 filters, and the commands run by the lmtp and pipe transports.
2800 5. Some older OS have a limit of 256 on the maximum number of file
2801 descriptors. Exim was using setrlimit() to set 1000 as a large value
2802 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2803 systems. I've change it so that if it can't get 1000, it tries for 256.
2805 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2806 was an oversight, and furthermore, ever since the addition of extra
2807 controls (e.g. 4.43/32), the checks on when to allow different forms of
2808 "control" were broken. There should now be diagnostics for all cases when a
2809 control that does not make sense is encountered.
2811 7. $recipients is now available in the predata ACL (oversight).
2813 8. Tidy the search cache before the fork to do a delivery from a message
2814 received from the command line. Otherwise the child will trigger a lookup
2815 failure and thereby defer the delivery if it tries to use (for example) a
2816 cached ldap connection that the parent has called unbind on.
2818 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2819 of $address_data from the recipient verification was clobbered by the
2820 sender verification.
2822 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2823 was its contents. (It was OK if the option was not defined at all.)
2825 11. A "Completed" log line is now written for messages that are removed from
2826 the spool by the -Mrm option.
2828 12. $host_address is now set to the target address during the checking of
2829 ignore_target_hosts.
2831 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2832 being passed; this would have caused $sender_host_name to have been used if
2833 matching the list had actually called for a host name (not very likely,
2834 since this list is usually IP addresses). A host name is now passed as
2837 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2838 code that specifies a non-restarting handler (typically sigaction() in
2839 modern systems) in an attempt to fix a rare and obscure crash bug.
2841 15. Narrowed the window for a race in the daemon that could cause it to ignore
2842 SIGCHLD signals. This is not a major problem, because they are used only to
2843 wake it up if nothing else does.
2845 16. A malformed maildirsize file could cause Exim to calculate negative values
2846 for the mailbox size or file count. Odd effects could occur as a result.
2847 The maildirsize information is now recalculated if the size or filecount
2850 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2851 support for a long time. Removed HAVE_SYS_VFS_H.
2853 18. Updated exipick to current release from John Jetmore.
2855 19. Allow an empty sender to be matched against a lookup in an address list.
2856 Previously the only cases considered were a regular expression, or an
2859 20. Exim went into a mad DNS lookup loop when doing a callout where the
2860 host was specified on the transport, if the DNS lookup yielded more than
2863 21. The RFC2047 encoding function was originally intended for short strings
2864 such as real names; it was not keeping to the 75-character limit for
2865 encoded words that the RFC imposes. It now respects the limit, and
2866 generates multiple encoded words if necessary. To be on the safe side, I
2867 have increased the buffer size for the ${rfc2047: expansion operator from
2870 22. Failure to deliver a bounce message always caused it to be frozen, even if
2871 there was an errors_to setting on the router. The errors_to setting is now
2874 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2875 canonical form (fully expanded) before being placed in
2876 $sender_host_address.
2878 24. Updated eximstats to version 1.33
2880 25. Include certificate and key file names in error message when GnuTLS fails
2881 to set them up, because the GnuTLS error message doesn't include the name
2882 of the failing file when there is a problem reading it.
2884 26. Expand error message when OpenSSL has problems setting up cert/key files.
2887 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2888 (this can affect the format of dates).
2890 28. exim_tidydb, when checking for the continued existence of a message for
2891 which it has found a message-specific retry record, was not finding
2892 messages that were in split spool directories. Consequently, it was
2893 deleting retry records that should have stayed in existence.
2895 29. eximstats updated to version 1.35
2896 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2897 1.35 - bugfix such that pie charts by volume are generated correctly
2899 30. The SPA authentication driver was not abandoning authentication and moving
2900 on to the next authenticator when an expansion was forced to fail,
2901 contradicting the general specification for all authenticators. Instead it
2902 was generating a temporary error. It now behaves as specified.
2904 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2905 (the order specifies the preference for clients). The order is now AES256,
2906 AES128, 3DES, ARCFOUR128.
2908 31. Small patch to Sieve code - explicitly set From: when generating an
2911 32. Exim crashed if a remote delivery caused a very long error message to be
2912 recorded - for instance if somebody sent an entire SpamAssassin report back
2913 as a large number of 550 error lines. This bug was coincidentally fixed by
2914 increasing the size of one of Exim's internal buffers (big_buffer) that
2915 happened as part of the Exiscan merge. However, to be on the safe side, I
2916 have made the code more robust (and fixed the comments that describe what
2919 33. Some experimental protocols are using DNS PTR records for new purposes. The
2920 keys for these records are domain names, not reversed IP addresses. The
2921 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2922 leaves it alone. Component reversal etc. now happens only for IP addresses.
2925 34. The host_aton() function is supposed to be passed a string that is known
2926 to be a valid IP address. However, in the case of IPv6 addresses, it was
2927 not checking this. This is a hostage to fortune. Exim now panics and dies
2928 if the condition is not met. A case was found where this could be provoked
2929 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2930 components; fortuitously, this particular loophole had already been fixed
2931 by change 4.50/55 or 4.44/33 above.
2933 If there are any other similar loopholes, the new check in host_aton()
2934 itself should stop them being exploited. The report I received stated that
2935 data on the command line could provoke the exploit when Exim was running as
2936 exim, but did not say which command line option was involved. All I could
2937 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2938 running as the user.
2941 35. There was a buffer overflow vulnerability in the SPA authentication code
2942 (which came originally from the Samba project). I have added a test to the
2943 spa_base64_to_bits() function which I hope fixes it.
2946 36. The daemon start-up calls getloadavg() while still root for those OS that
2947 need the first call to be done as root, but it missed one case: when
2948 deliver_queue_load_max is set with deliver_drop_privilege. This is
2949 necessary for the benefit of the queue runner, because there is no re-exec
2950 when deliver_drop_privilege is set.
2952 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2953 was in use, was not putting the data itself into the right store pool;
2954 consequently, it could be overwritten for a subsequent message in the same
2955 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2958 38. Sometimes the final signoff response after QUIT could fail to get
2959 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2960 < 0 before doing a fflush(). This bug looks as though it goes back to the
2961 introduction of TLS in release 3.20, but "sometimes" must have been rare
2962 because the tests only now provoked it.
2968 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2969 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2970 Now it returns an integer. A number of calls were still expecting a T/F
2971 return. Fortuitously, in all cases, the tests worked in OK situations,
2972 which is the norm. However, things would have gone wrong on any write
2973 failures on the smtp file descriptor. This function is used when sending
2974 messages over SMTP and also when doing verify callouts.
2976 2. When Exim is called to do synchronous delivery of a locally submitted
2977 message (the -odf or -odi options), it no longer closes stderr before doing
2980 3. Implemented the mua_wrapper option.
2982 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2984 5. Implemented the functions header_remove(), header_testname(),
2985 header_add_at_position(), and receive_remove_recipient(), and exported them
2988 6. If an ACL "warn" statement specified the addition of headers, Exim already
2989 inserted X-ACL-Warn: at the start if there was no header name. However, it
2990 was not making this test for the second and subsequent header lines if
2991 there were newlines in the string. This meant that an invalid header could
2992 be inserted if Exim was badly configured.
2994 7. Allow an ACL "warn" statement to add header lines at the start or after all
2995 the Received: headers, as well as at the end.
2997 8. Added the rcpt_4xx retry error code.
2999 9. Added postmaster_mailfrom=xxx to callout verification option.
3001 10. Added mailfrom=xxxx to the callout verification option, for verify=
3004 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3005 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3007 12. Inserted some casts to stop certain compilers warning when using pointer
3008 differences as field lengths or precisions in printf-type calls (mostly
3009 affecting debugging statements).
3011 13. Added optional readline() support for -be (dynamically loaded).
3013 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3014 same clock tick as a message's arrival, so that its received time was the
3015 same as the "first fail" time on the retry record, and that message
3016 remained on the queue past the ultimate address timeout, every queue runner
3017 would try a delivery (because it was past the ultimate address timeout) but
3018 after another failure, the ultimate address timeout, which should have then
3019 bounced the address, did not kick in. This was a "< instead of <=" error;
3020 in most cases the first failure would have been in the next clock tick
3021 after the received time, and all would be well.
3023 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3024 being recognized when the domain list was tested by the match_domain
3025 condition in an expansion string.
3027 16. Added the ${str2b64: operator.
3029 17. Exim was always calling setrlimit() to set a large limit for the number of
3030 processes, without checking whether the existing limit was already
3031 adequate. (It did check for the limit on file descriptors.) Furthermore,
3032 errors from getrlimit() and setrlimit() were being ignored. Now they are
3033 logged to the main and panic logs, but Exim does carry on, to try to do its
3034 job under whatever limits there are.
3036 18. Imported PCRE 5.0.
3038 19. Trivial typo in log message " temporarily refused connection" (the leading
3041 20. If the log selector return_path_on_delivery was set and an address was
3042 redirected to /dev/null, the delivery process crashed because it assumed
3043 that a return path would always be set for a "successful" delivery. In this
3044 case, the whole delivery is bypassed as an optimization, and therefore no
3047 21. Internal re-arrangement: the function for sending a challenge and reading
3048 a response while authentication was assuming a zero-terminated challenge
3049 string. It's now changed to take a pointer and a length, to allow for
3050 binary data in such strings.
3052 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3054 23. Exim was not respecting finduser_retries when seeking the login of the
3055 uid under which it was called; it was always trying 10 times. (The default
3056 setting of finduser_retries is zero.) Also, it was sleeping after the final
3057 failure, which is pointless.
3059 24. Implemented tls_on_connect_ports.
3061 25. Implemented acl_smtp_predata.
3063 26. If the domain in control=submission is set empty, Exim assumes that the
3064 authenticated id is a complete email address when it generates From: or
3065 Sender: header lines.
3067 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3068 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3069 chown and chgrp in /bin and hostname in /usr/bin.
3071 28. Exim was keeping the "process log" file open after each use, just as it
3072 does for the main log. This opens the possibility of it remaining open for
3073 long periods when the USR1 signal hits a daemon. Occasional processlog
3074 errors were reported, that could have been caused by this. Anyway, it seems
3075 much more sensible not to leave this file open at all, so that is what now
3078 29. The long-running daemon process does not normally write to the log once it
3079 has entered its main loop, and it closes the log before doing so. This is
3080 so that log files can straightforwardly be renamed and moved. However,
3081 there are a couple of unusual error situations where the daemon does write
3082 log entries, and I had neglected to close the log afterwards.
3084 30. The text of an SMTP error response that was received during a remote
3085 delivery was being truncated at 512 bytes. This is too short for some of
3086 the long messages that one sometimes sees. I've increased the limit to
3089 31. It is now possible to make retry rules that apply only when a message has a
3090 specific sender, in particular, an empty sender.
3092 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3093 it possible to be selective about when SMTP synchronization is enforced.
3095 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3097 32. Implemented hosts_connection_nolog.
3099 33. Added an ACL for QUIT.
3101 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3104 35. Added mailbox_size and mailbox_filecount to appendfile.
3106 36. Added control = no_multiline_responses to ACLs.
3108 37. There was a bug in the logic of the code that waits for the clock to tick
3109 in the case where the clock went backwards by a substantial amount such
3110 that the microsecond fraction of "now" was more than the microsecond
3111 fraction of "then" (but the whole seconds number was less).
3113 38. Added support for the libradius Radius client library this is found on
3114 FreeBSD (previously only the radiusclient library was supported).
3120 1. When certain lookups returned multiple values in the form name=value, the
3121 quoting of the values was not always being done properly. Specifically:
3122 (a) If the value started with a double quote, but contained no whitespace,
3124 (b) If the value contained whitespace other than a space character (i.e.
3125 tabs or newlines or carriage returns) it was not quoted.
3126 This fix has been applied to the mysql and pgsql lookups by writing a
3127 separate quoting function and calling it from the lookup code. The fix
3128 should probably also be applied to nisplus, ibase and oracle lookups, but
3129 since I cannot test any of those, I have not disturbed their existing code.
3131 2. A hit in the callout cache for a specific address caused a log line with no
3132 reason for rejecting RCPT. Now it says "Previous (cached) callout
3133 verification failure".
3135 3. There was an off-by-one bug in the queryprogram router. An over-long
3136 return line was truncated at 256 instead of 255 characters, thereby
3137 overflowing its buffer with the terminating zero. As well as fixing this, I
3138 have increased the buffer size to 1024 (and made a note to document this).
3140 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3141 when Exim is waiting for an SMTP response from a remote server, Exim
3142 restarts its select() call on the socket, thereby resetting its timeout.
3143 This is not a problem when such interrupts are rare. Somebody set up a cron
3144 job to run exiwhat every 2 minutes, which is less than the normal select()
3145 timeout (5 or 10 minutes). This meant that the select() timeout never
3146 kicked in because it was always reset. I have fixed this by comparing the
3147 time when an interrupt arrives with the time at the start of the first call
3148 to select(). If more time than the timeout has elapsed, the interrupt is
3149 treated as a timeout.
3151 5. Some internal re-factoring in preparation for the addition of Sieve
3152 extensions (by MH). In particular, the "personal" test is moved to a
3153 separate function, and given an option for scanning Cc: and Bcc: (which is
3154 not set for Exim filters).
3156 6. When Exim created an email address using the login of the caller as the
3157 local part (e.g. when creating a From: or Sender: header line), it was not
3158 quoting the local part when it contained special characters such as @.
3160 7. Installed new OpenBSD configuration files.
3162 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3163 try to make them clearer.
3165 9. Callout options, other than the timeout value, were being ignored when
3166 verifying sender addresses in header lines. For example, when using
3168 verify = header_sender/callout=no_cache
3170 the cache was (incorrectly) being used.
3172 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3173 only the Cygwin environment.
3175 11. When return_path_on_delivery was set as a log selector, if different remote
3176 addresses in the same message used different return paths and parallel
3177 remote delivery occurred, the wrong values would sometimes be logged.
3178 (Whenever a remote delivery process finished, the return path value from
3179 the most recently started remote delivery process was logged.)
3181 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3182 lines when AUTH and/or TLS are in use. This is the "received protocol"
3183 field. Exim used to use "asmtp" for authenticated SMTP, without any
3184 indication (in the protocol name) for TLS use. Now it follows the RFC and
3185 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3186 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3187 names appear in log lines as well as in Received: header lines.
3189 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3190 extensions, and comparison tests, and to fix some bugs.
3192 14. Changes to the "personal" filter test:
3194 (1) The test was buggy in that it was just doing the equivalent of
3195 "contains" tests on header lines. For example, if a user's address was
3196 anne@some.where, the "personal" test would incorrectly be true for
3198 To: susanne@some.where
3200 This test is now done by extracting each address from the header in turn,
3201 and checking the entire address. Other tests that are part of "personal"
3202 are now done using regular expressions (for example, to check local parts
3203 of addresses in From: header lines).
3205 (2) The list of non-personal local parts in From: addresses has been
3206 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3207 taken from the Sieve specification recommendations.
3209 (3) If the message contains any header line starting with "List-" it is
3210 treated as non-personal.
3212 (4) The test for "circular" in the Subject: header line has been removed
3213 because it now seems ill-conceived.
3215 15. Minor typos in src/EDITME comments corrected.
3217 16. Installed latest exipick from John Jetmore.
3219 17. If headers_add on a router specified a text string that was too long for
3220 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3221 of string_sprintf() is now avoided.
3223 18. $message_body_size was not set (it was always zero) when running the DATA
3224 ACL and the local_scan() function.
3226 19. For the "mail" command in an Exim filter, no default was being set for
3227 the once_repeat time, causing a random time value to be used if "once" was
3228 specified. (If the value happened to be <= 0, no repeat happened.) The
3229 default is now 0s, meaning "never repeat". The "vacation" command was OK
3230 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3231 (I found it when inspecting the code).
3233 20. There is now an overall timeout for performing a callout verification. It
3234 defaults to 4 times the callout timeout, which applies to individual SMTP
3235 commands during the callout. The overall timeout applies when there is more
3236 than one host that can be tried. The timeout is checked before trying the
3237 next host. This prevents very long delays if there are a large number of
3238 hosts and all are timing out (e.g. when the network connections are timing
3239 out). The value of the overall timeout can be changed by specifying an
3240 additional sub-option for "callout", called "maxwait". For example:
3242 verify = sender/callout=5s,maxwait=20s
3244 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3245 to the end before writing, but this should make it even safer).
3247 22. Exim was forgetting that it had advertised PIPELINING for the second and
3248 subsequent messages on an SMTP connection. It was also not resetting its
3249 memory on STARTTLS and an internal HELO.
3251 23. When Exim logs an SMTP synchronization error within a session, it now
3252 records whether PIPELINING has been advertised or not.
3254 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3255 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3256 rather than long int.
3258 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3260 26. Added the never_mail option to autoreply.
3266 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3267 crash if the getsockname() call failed; this can happen if a connection is
3268 closed very soon after it is established. The problem was simply in the
3269 order in which certain operations were done, causing Exim to try to write
3270 to the SMTP stream before it had set up the file descriptor. The bug has
3271 been fixed by making things happen in the correct order.
3277 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3278 before the connection was closed, thus losing the rejection response.
3280 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3281 some early Solaris releases, but causes trouble in current releases where
3282 socklen_t is defined.
3284 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3287 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3290 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3291 that releasing the top of it at the end releases what was used for sub-
3292 expansions (unless the block got too big). However, discard this block if
3293 the first thing is a variable or header, so that we can use its block when
3294 it is dynamic (useful for very large $message_headers, for example).
3296 6. Lookups now cache *every* query, not just the most recent. A new, separate
3297 store pool is used for this. It can be recovered when all lookup caches are
3298 flushed. Lookups now release memory at the end of their result strings.
3299 This has involved some general refactoring of the lookup sources.
3301 7. Some code has been added to the store_xxx() functions to reduce the amount
3302 of flapping under certain conditions.
3304 8. log_incoming_interface used to affect only the <= reception log lines. Now
3305 it causes the local interface and port to be added to several more SMTP log
3306 lines, for example "SMTP connection from", and rejection lines.
3308 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3310 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3312 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3313 could be overwritten at the end of the current message (or the start of a
3314 new message if it was set in a HELO ACL). The value is now preserved for
3315 the duration of the SMTP connection.
3317 12. If a transport had a headers_rewrite setting, and a matching header line
3318 contained an unqualified address, that address was qualified, even if it
3319 did not match any rewriting rules. The underlying bug was that the values
3320 of the flags that permit the existence of unqualified sender and recipient
3321 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3322 non-local messages, and by -bnq for local messages) were not being
3323 preserved with the message after it was received.
3325 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3326 "next input=" as part of the text comprising the host identity instead of
3327 the correct text. The code was using the same buffer for two different
3328 strings. However, depending on which order the printing function evaluated
3329 its arguments, the bug did not always show up. Under Linux, for example, my
3330 test suite worked just fine.
3332 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3333 doesn't work with some older versions of Perl. It has been changed to "my",
3334 which in any case is probably the better facility to use.
3336 15. A really picky compiler found some instances of statements for creating
3337 error messages that either had too many or two few arguments for the format
3340 16. The size of the buffer for calls to the DNS resolver has been increased
3341 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3342 for addresses that have a lot of PTR records. This alleviates a problem; it
3343 does not fully solve it.
3345 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3346 buffer now truncates the list and logs the incident, which is the same
3347 action as happens when Exim is looking up a host name and its aliases.
3348 Previously in this situation something unpredictable would happen;
3349 sometimes it was "internal error: store_reset failed".
3351 18. If a server dropped the connection unexpectedly when an Exim client was
3352 using GnuTLS and trying to read a response, the client delivery process
3353 crashed while trying to generate an error log message.
3355 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3356 string, for example:
3358 warn message = H1: something\nH2: something
3360 the text was added as a single header line from Exim's point of view
3361 though it ended up OK in the delivered message. However, searching for the
3362 second and subsequent header lines using $h_h2: did not work. This has been
3363 fixed. Similarly, if a system filter added multiple headers in this way,
3364 the routers could not see them.
3366 20. Expanded the error message when iplsearch is called with an invalid key to
3367 suggest using net-iplsearch in a host list.
3369 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3370 ACLs are no longer actually imposed (and a message to that effect is
3373 22. If a "gecos" field in a passwd entry contained escaped characters, in
3374 particular, if it contained a \" sequence, Exim got it wrong when building
3375 a From: or a Sender: header from that name. A second bug also caused
3376 incorrect handling when an unquoted " was present following a character
3377 that needed quoting.
3379 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3380 was not being matched caselessly.
3382 24. Arranged for all hyphens in the exim.8 source to be escaped with
3385 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3386 properly. Recipient callout cache records were still being keyed to include
3387 the sender, even when use_sender was set false. This led to far more
3388 callouts that were necessary. The sender is no longer included in the key
3389 when use_sender is false.
3391 26. Added "control = submission" modifier to ACLs.
3393 27. Added the ${base62d: operator to decode base 62 numbers.
3395 28. dnsdb lookups can now access SRV records.
3397 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3398 the configuration file.
3400 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3401 (-v) mode. This makes the output for a verbose queue run more intelligible.
3403 31. Added a use_postmaster feature to recipient callouts.
3405 32. Added the $body_zerocount variable, containing the number of binary zero
3406 bytes in the message body.
3408 33. The time of last modification of the "new" subdirectory is now used as the
3409 "mailbox time last read" when there is a quota error for a maildir
3412 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3414 35. Added +ignore_unknown as a special item in host lists.
3416 36. Code for decoding IPv6 addresses in host lists is now included, even if
3417 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3418 address was recognized as an IP address, but was then not correctly decoded
3419 into binary, causing unexpected and incorrect effects when compared with
3426 1. Very minor rewording of debugging text in manualroute to say "list of
3427 hosts" instead of "hostlist".
3429 2. If verify=header_syntax was set, and a header line with an unqualified
3430 address (no domain) and a large number of spaces between the end of the
3431 name and the colon was received, the reception process suffered a buffer
3432 overflow, and (when I tested it) crashed. This was caused by some obsolete
3433 code that should have been removed. The fix is to remove it!
3435 3. When running in the test harness, delay a bit after writing a bounce
3436 message to get a bit more predictability in the log output.
3438 4. Added a call to search_tidyup() just before forking a reception process. In
3439 theory, someone could use a lookup in the expansion of smtp_accept_max_
3440 per_host which, without the tidyup, could leave open a database connection.
3442 5. Added the variables $recipient_data and $sender_data which get set from a
3443 lookup success in an ACL "recipients" or "senders" condition, or a router
3444 "senders" option, similar to $domain_data and $local_part_data.
3446 6. Moved the writing of debug_print from before to after the "senders" test
3449 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3450 problems for message scanning, either using a data ACL, or using
3451 local_scan() because the Received: header was not generated till after they
3452 were called (in order to set the time as the time of reception completion).
3453 I have revised the way this works. The header is now generated after the
3454 body is received, but before the ACL or local_scan() are called. After they
3455 are run, the timestamp in the header is updated.
3461 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3462 before starting a queue runner without re-exec. This happened only when
3463 deliver_drop_privilege was set or when the Exim user was set to root. The
3464 effect of the bug was that timeouts during subsequent deliveries caused
3465 crashes instead of being properly handled. The handler is now left at its
3466 default (and expected) setting.
3468 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3469 message, again when deliver_drop_privilege is set or Exim is run as root.
3470 The bug described in (1) was not present in this case, but the tidying up
3471 of the other signals was missing. I have made the two cases consistent.
3473 3. The ignore_target_hosts setting on a manualroute router was being ignored
3474 for hosts that were looked up using the /MX notation.
3476 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3479 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3480 operated on the sender address. After changing the $sender_address to <>
3481 for the sender address verify, Exim was re-instated it as the original
3482 (before rewriting) address, but remembering that it had rewritten it, so it
3483 wasn't rewriting it again. This bug also had the effect of breaking the
3484 sender address verification caching when the sender address was rewritten.
3486 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3487 This has been changed so that if the ip literal address matches
3488 ignore_target_hosts, the router declines.
3490 7. Added expansion conditions match_domain, match_address, and match_local_
3491 part (NOT match_host).
3493 8. The placeholder for the Received: header didn't have a length field set.
3495 9. Added code to Exim itself and to exim_lock to test for a specific race
3496 condition that could lead to file corruption when using MBX delivery. The
3497 issue is with the lockfile that is created in /tmp. If this file is removed
3498 after a process has opened it but before that process has acquired a lock,
3499 there is the potential for a second process to recreate the file and also
3500 acquire a lock. This could lead to two Exim processes writing to the file
3501 at the same time. The added code performs the same test as UW imapd; it
3502 checks after acquiring the lock that its file descriptor still refers to
3503 the same named file.
3505 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3506 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3509 11. Added the smtp_active_hostname option. If used, this will typically be made
3510 to depend on the incoming interface address. Because $interface_address is
3511 not set up until the daemon has forked a reception process, error responses
3512 that can happen earlier (such as "too many connections") no longer contain
3515 12. If an expansion in a condition on a "warn" statement fails because a lookup
3516 defers, the "warn" statement is abandoned, and the next ACL statement is
3517 processed. Previously this caused the whole ACL to be aborted.
3519 13. Added the iplsearch lookup type.
3521 14. Added ident_timeout as a log selector.
3523 15. Added tls_certificate_verified as a log selector.
3525 16. Added a global option tls_require_ciphers (compare the smtp transport
3526 option of the same name). This controls incoming TLS connections.
3528 17. I finally figured out how to make tls_require_ciphers do a similar thing
3529 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3530 before starting the TLS session.
3532 18. Tabs are now shown as \t in -bP output.
3534 19. If the log selector return_path_on_delivery was set, Exim crashed when
3535 bouncing a message because it had too many Received: header lines.
3537 20. If two routers both had headers_remove settings, and the first one included
3538 a superfluous trailing colon, the final name in the first list and the
3539 first name in the second list were incorrectly joined into one item (with a
3540 colon in the middle).
3546 1. Added -C and -D options to the exinext utility, mainly to make it easier
3547 to include in the automated testing, but these could be helpful when
3548 multiple configurations are in use.
3550 2. The exinext utility was not formatting the output nicely when there was
3551 an alternate port involved in the retry record key, nor when there was a
3552 message id as well (for retries that were specific to a specific message
3553 and a specific host). It was also confused by IPv6 addresses, because of
3554 the additional colons they contain. I have fixed the IPv4 problem, and
3555 patched it up to do a reasonable job for IPv6.
3557 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3558 delivery, the log line now contains "pipelined" if PIPELINING was used.
3560 4. An SMTP transport process used to panic and die if the bind() call to set
3561 an explicit outgoing interface failed. This has been changed; it is now
3562 treated in the same way as a connect() failure.
3564 5. A reference to $sender_host_name in the part of a conditional expansion
3565 that was being skipped was still causing a DNS lookup. This no longer
3568 6. The def: expansion condition was not recognizing references to header lines
3569 that used bh_ and bheader_.
3571 7. Added the _cache feature to named lists.
3573 8. The code for checking quota_filecount in the appendfile transport was
3574 allowing one more file than it should have been.
3576 9. For compatibility with Sendmail, the command line option
3584 and sets the incoming protocol and host name (for trusted callers). The
3585 host name and its colon can be omitted when only the protocol is to be set.
3586 Note the Exim already has two private options, -pd and -ps, that refer to
3587 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3588 "s", but I don't think that's a major issue.
3590 10. A number of refactoring changes to the code, none of which should affect
3593 (a) The number of logging options was getting close to filling up the
3594 32-bit word that was used as a bit map. I have split them into two classes:
3595 those that are passed in the argument to log_write(), and those that are
3596 only ever tested independently outside of that function. These are now in
3597 separate 32-bit words, so there is plenty of room for expansion again.
3598 There is no change in the user interface or the logging behaviour.
3600 (b) When building, for example, log lines, the code previously used a
3601 macro that called string_cat() twice, in order to add two strings. This is
3602 not really sufficiently general. Furthermore, there was one instance where
3603 it was actually wrong because one of the argument was used twice, and in
3604 one call a function was used. (As it happened, calling the function twice
3605 did not affect the overall behaviour.) The macro has been replaced by a
3606 function that can join an arbitrary number of extra strings onto a growing
3609 (c) The code for expansion conditions now uses a table and a binary chop
3610 instead of a serial search (which was left over from when there were very
3611 few conditions). Also, it now recognizes conditions like "pam" even when
3612 the relevant support is not compiled in: a suitably worded error message is
3613 given if an attempt is made to use such a condition.
3615 11. Added ${time_interval:xxxxx}.
3617 12. A bug was causing one of the ddress fields not to be passed back correctly
3618 from remote delivery subprocesses. The field in question was not being
3619 subsequently used, so this caused to problems in practice.
3621 13. Added new log selectors queue_time and deliver_time.
3623 14. Might have fixed a bug in maildirsizefile handling that threw up
3624 "unexpected character" debug warnings, and recalculated the data
3625 unnecessarily. In any case, I expanded the warning message to give more
3628 15. Added the message "Restricted characters in address" to the statements in
3629 the default ACL that block characters like @ and % in local parts.
3631 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3632 Three changes have been made:
3634 (a) There was a serious bug; a negative response to MAIL caused the whole
3635 recipient domain to be cached as invalid, thereby blocking all messages
3636 to all local parts at the same domain, from all senders. This bug has
3637 been fixed. The domain is no longer cached after a negative response to
3638 MAIL if the sender used is not empty.
3640 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3643 (c) A new callout option, "use_sender" has been added for people who want
3644 the modified behaviour.
3650 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3653 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3654 indeed breaks things for older releases.
3656 3. Added additional logging to the case where there is a problem reading data
3657 from a filter that is running in a subprocess using a pipe, in order to
3658 try to track down a specific problem.
3660 4. Testing facility fudge: when running in the test harness and attempting
3661 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3662 getting "No route to host". Convert this to a timeout.
3664 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3667 6. Some OS don't have socklen_t but use size_t instead. This affects the
3668 fifth argument of getsockopt() amongst other things. This is now
3669 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3670 can be set for individual OS. I have set it for SunOS5, OSF1, and
3671 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3672 some earlier ones do not.
3674 7. Change 4.30/15 was not doing the test caselessly.
3676 8. The standard form for an IPv6 address literal was being rejected by address
3677 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3678 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3679 this, as well as the form without the "IPv6" on the front (but only when
3680 address literals are enabled, of course).
3682 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3684 10. Exim crashed if a message with an empty sender address specified by -f
3685 encountered a router with an errors_to setting. This could be provoked only
3686 by a command such as
3690 where an empty string was supplied; "<>" did not hit this bug.
3692 11. Installed PCRE release 4.5.
3694 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3695 remained set. It is now erased.
3697 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3698 times from message ids (which are base 36 rather than the normal 62).
3700 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3701 were being counted as actual protocol errors, and logged if the log
3702 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3703 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3704 and DATA following a set of rejected RCPTs do not count as protocol errors.
3705 In other words, Exim assumes they were pipelined, though this may not
3706 actually be the case. Of course, in all cases the client gets an
3707 appropriate error code.
3709 15. If a lookup fails in an ACL condition, a message about the failure may
3710 be available; it is used if testing the ACL cannot continue, because most
3711 such messages specify what the cause of the deferral is. However, some
3712 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3713 that caused an old message to be retained and used if a later statement
3714 caused a defer, replacing the real cause of the deferral.
3716 16. If an IP address had so many PTR records that the DNS lookup buffer
3717 was not large enough to hold them, Exim could crash while trying to process
3718 the truncated data. It now detects and logs this case.
3720 17. Further to 4.21/58, another change has been made: if (and only if) the
3721 first line of a message (the first header line) ends with CRLF, a bare LF
3722 in a subsequent header line has a space inserted after it, so as not to
3723 terminate the header.
3725 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3726 unnecessarily, used atoi() instead of strtol(), and didn't check the
3727 termination when getting file sizes from file names by regex.
3729 19. Completely re-implemented the support for maildirsize files, in the light
3730 of a number of problems with the previous contributed implementation
3731 (4.30/29). In particular:
3733 . If the quota is zero, the maildirsize file is maintained, but no quota is
3736 . If the maildir directory does not exist, it is created before any attempt
3737 to write a maildirsize file.
3739 . The quota value in the file is just a cache; if the quota is changed in
3740 the transport, the new value overrides.
3742 . A regular expression is available for excluding directories from the
3745 20. The autoreply transport checks the characters in options that define the
3746 message's headers; it allows continued headers, but it was checking with
3747 isspace() after an embedded newline instead of explicitly looking for a
3750 21. If all the "regular" hosts to which an address was routed had passed their
3751 expiry times, and had not reached their retry times, the address was
3752 bounced, even if fallback hosts were defined. Now Exim should go on to try
3755 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3756 equivalent code in the SMTP transport. Some hosts send humungous responses
3757 to HELO/EHLO, more than 1024 it seems.
3759 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3760 gives compiler warnings in some environments. I've now done it "properly",
3763 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3764 (because of problems with the built-in one) was declared to return uschar *
3765 instead of char *, causing compiler failure.
3767 25. Fixed a file descriptor leak when processing alias/forward files.
3769 26. Fixed a minor format string issue in dbfn.c.
3771 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3773 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3774 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3777 29. When a sender address is verified, it is cached, to save repeating the test
3778 when there is more than one recipient in a message. However, when the
3779 verification involves a callout, it is possible for different callout
3780 options to be set for different recipients. It is too complicated to keep
3781 track of this in the cache, so now Exim always runs a verification when a
3782 callout is required, relying on the callout cache for the optimization.
3783 The overhead is duplication of the address routing, but this should not be
3786 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3787 to be verified with callout=postmaster, and the main callout worked but the
3788 postmaster check failed, the verification correctly failed. However, if a
3789 subsequent RCPT command asked for sender verification *without* the
3790 postmaster check, incorrect caching caused this verification also to fail,
3793 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3794 it was not caching the DNS options (qualify_single, search_parents) that
3795 were used when the lookup failed. A subsequent lookup with different
3796 options therefore always gave the same answer, though there were cases
3797 where it should not have. (Example: a "domains = !$mx_any" option on a
3798 dnslookup router: the "domains" option is always processed without any
3799 widening, but the router might have qualify_single set.) Now Exim uses the
3800 cached value only when the same options are set.
3802 32. Added John Jetmore's "exipick" utility to the distribution.
3804 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3805 than a timeout (e.g. a certificate is required, and is not provided), an
3806 Exim server now closes the connection immediately. Previously it waited for
3807 the client to close - but if the client is SSL, it seems that they each
3808 wait for each other, leading to a delay before one of them times out.
3810 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3811 maintained 0.8.x compatibility because I don't think many are using it, and
3812 it is clearly obsolete.
3814 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3817 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3818 client certificate was expired. A simple patch fixes this, though I don't
3819 understand the full logic of why the verify callback is called multiple
3822 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3823 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3824 which causes problems with some clients (such as the Certicom SSL Plus
3825 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3826 disables the coutermeasure allowing Eudora to connect."
3828 38. Exim was not checking that a write() to a log file succeeded. This could
3829 lead to Bad Things if a log got too big, in particular if it hit a file
3830 size limit. Exim now panics and dies if it cannot write to a log file, just
3831 as it does if it cannot open a log file.
3833 39. Modified OS/Makefile-Linux so that it now contains
3835 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3837 The two -D definitions ensure that Exim is compiled with large file
3838 support, which makes it possible to handle log files that are bigger than
3841 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3842 instance) a domain was checked against a named list that involved a lookup,
3843 causing $domain_data to be set, then another domain was checked against the
3844 same list, then the first domain was re-checked, the value of $domain_data
3845 after the final check could be wrong. In particular, if the second check
3846 failed, it could be set empty. This bug probably also applied to
3849 41. The strip_trailing_dot option was not being applied to the address given
3850 with the -f command-line option.
3852 42. The code for reading a message's header from the spool was incrementing
3853 $received_count, but never initializing it. This meant that the value was
3854 incorrect (doubled) while delivering a message in the same process in which
3855 it was received. In the most common configuration of Exim, this never
3856 happens - a fresh exec is done - but it can happen when
3857 deliver_drop_privilege is set.
3859 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3860 it now includes up to 150 characters of the unexpected data in the log
3863 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3864 and building data strings. The size of both of these buffers was 10 000
3865 bytes - far larger than anybody would *ever* want, thought I. Needless to
3866 say, somebody hit the limit. I have increased the maximum line length to
3867 20 000 and the maximum data length of concatenated lines to 100 000. I have
3868 also fixed two bugs, because there was no checking on these buffers. Tsk,
3869 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3870 buffer is too small.
3872 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3873 lsearch lookups. Now it does.
3875 46. When parsing a route_list item in a manualroute router, a fixed-length
3876 buffer was used for the list of hosts. I made this 1024 bytes long,
3877 thinking that nobody would ever have a list of hosts that long. Wrong.
3878 Somebody had a whole pile of complicated expansion conditions, and the
3879 string was silently truncated, leading to an expansion error. It turns out
3880 that it is easier to change to an unlimited length (owing to other changes
3881 that have happened since this code was originally written) than to build
3882 structure for giving a limitation error. The length of the item that
3883 expands into the list of hosts is now unlimited.
3885 47. The lsearch lookup could not handle data where the length of text line was
3886 more than 4095 characters. Such lines were truncated, leading to shortened
3887 data being returned. It should now handle lines of any length.
3889 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3890 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3893 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3894 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3895 no longer happens in comment lines. A list of replacements is now placed
3896 at the head of all of the source files, except those whose only change is
3897 to replace PERL_COMMAND in the very first #! line.
3899 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3900 messages on the queue, with a bottom-up merge sort, using code contributed
3901 by Michael Haardt. This should make operations like -bp somewhat faster on
3902 large queues. It won't affect queue runners, except when queue_run_in_order
3905 51. Installed eximstats 1.31 in the distribution.
3907 52. Added support for SRV lookups to the dnslookup router.
3909 53. If an ACL referred to $message_body or $message_body_end, the value was not
3910 reset for any messages that followed in the same SMTP session.
3912 54. The store-handling optimization for building very long strings was not
3913 differentiating between the different store pools. I don't think this
3914 actually made any difference in practice, but I've tidied it.
3916 55. While running the routers to verify a sender address, $sender_address
3917 was still set to the sender address. This is wrong, because when routing to
3918 send a bounce to the sender, it would be empty. Therefore, I have changed
3919 it so that, while verifying a sender address, $sender_address is set to <>.
3920 (There is no change to what happens when verifying a recipient address.)
3922 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3923 target A or AAAA records (if not already returned) without resetting the
3924 qualify_single or search_parents options of the DNS resolver. These are
3925 inappropriate in this case because the targets of MX and SRV records must
3926 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3927 target that, when qualified, matched something in the local domain. These
3928 two options are now turned off when doing these lookups.
3930 57. It seems that at least some releases of Reiserfs (which does not have the
3931 concept of a fixed number of inodes) returns zero and not -1 for the
3932 number of available inodes. This interacted badly with check_spool_inodes,
3933 which assumed that -1 was the "no such thing" setting. What I have done is
3934 to check that the total number of inodes is greater than zero before doing
3935 the test of how many are available.
3937 58. When a "warn" ACL statement has a log_message modifier, the message is
3938 remembered, and not repeated. This is to avoid a lot of repetition when a
3939 message has many recipients that cause the same warning to be written.
3940 Howewer, Exim was preserving the list of already written lines for an
3941 entire SMTP session, which doesn't seem right. The memory is now reset if a
3942 new message is started.
3944 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3945 debugging output unless log_rewrite was also set.
3947 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3948 of (int)(handle) when we know that handle contains (void *)(-1).
3950 61. The Exim daemon panic-logs an error return when it closes the incoming
3951 connection. However "connection reset by peer" seems to be common, and
3952 isn't really an error worthy of noting specially, so that particular error
3955 62. When Exim is trying to find all the local interfaces, it used to panic and
3956 die if the ioctl to get the interface flags failed. However, it seems that
3957 on at least one OS (Solaris 9) it is possible to have an interface that is
3958 included in the list of interfaces, but for which you get a failure error
3959 for this call. This happens when the interface is not "plumbed" into a
3960 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3961 failure of the "get flags" call assumes that the interface is down.
3963 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3964 makes life easier for people who are doing arithmetic on fields extracted
3965 from dates, where you often get leading zeros that should not be
3966 interpreted as octal.
3968 64. Added qualify_domain to the redirect router, to override the global
3971 65. If a pathologically long header line contained very many addresses (the
3972 report of this problem mentioned 10 000) and each of them was rewritten,
3973 Exim could use up a very large amount of memory. (It kept on making new
3974 copies of the header line as it rewrote, and never released the old ones.)
3975 At the expense of a bit more processing, the header rewriting function has
3976 been changed so that it no longer eats memory in this way.
3978 66. The generation of the Received: header has been moved from the time that a
3979 message starts to be received, to the time that it finishes. The timestamp
3980 in the Received: header should now be very close to that of the <= log
3981 line. There are two side-effects of this change:
3983 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3984 logged header lines no longer include the local Received: line, because
3985 it has not yet been created. The same applies to a copy of the message
3986 that is returned to a non-SMTP sender when a message is rejected.
3988 (b) When a filter file is tested using -bf, no additional Received: header
3989 is added to the test message. After some thought, I decided that this
3992 This change does not affect the value of $received_for. It is still set
3993 after address rewriting, but before local_scan() is called.
3995 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3997 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3998 gave an unhelpful panic error message, and a defer error. I have managed to
3999 change this behaviour so that it now rejects any supplied certificate,
4000 which seems right, as the list of acceptable certificates is empty.
4002 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4003 gave an unhelpful defer error. I have not managed to make this reject any
4004 supplied certificates, but the error message it gives is "no certificate
4005 supplied", which is not helpful.
4007 70. exigrep's output now also includes lines that are not associated with any
4008 message, but which match the given pattern. Implemented by a patch from
4009 Martin Sluka, which also tidied up the Perl a bit.
4011 71. Recipient callout verification, like sender verification, was using <> in
4012 the MAIL FROM command. This isn't really the right thing, since the actual
4013 sender may affect whether the remote host accepts the recipient or not. I
4014 have changed it to use the actual sender in the callout; this means that
4015 the cache record is now keyed on a recipient/sender pair, not just the
4016 recipient address. There doesn't seem to be a real danger of callout loops,
4017 since a callout by the remote host to check the sender would use <>.
4018 [SEE ABOVE: changed after hitting problems.]
4020 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4021 temporary errors. However, in the case of such a code being given after
4022 the end of a data transmission (i.e. after ".") Exim was failing to write
4023 a retry record for the message. (Yes, there was some broken host that was
4024 actually sending 8xx at this point.)
4026 73. An unknown lookup type in a host list could cause Exim to panic-die when
4027 the list was checked. (An example that provoked this was putting <; in the
4028 middle of a list instead of at the start.) If this happened during a DATA
4029 ACL check, a -D file could be left lying around. This kind of configuration
4030 error no longer causes Exim to die; instead it causes a defer errror. The
4031 incident is still logged to the main and panic logs.
4033 74. Buglet left over from Exim 3 conversion. The message "too many messages
4034 in one connection" was written to the rejectlog but not the mainlog, except
4035 when address rewriting (yes!) was being logged.
4037 75. Added write_rejectlog option.
4039 76. When a system filter was run not as root (that is, when system_filter_user
4040 was set), the values of the $n variables were not being returned to the
4041 main process; thus, they were not subsequently available in the $sn
4044 77. Added +return_path_on_delivery log selector.
4046 78. A connection timeout was being treated differently from recipients deferred
4047 when testing hosts_max_try with a message that was older than the host's
4048 retry timeout. (The host should not be counted, thus allowing all hosts to
4049 be tried at least once before bouncing.) This may have been the cause of an
4050 occasionally reported bug whereby a message would remain on the queue
4051 longer than the retry timeout, but would be bounced if a delivery was
4052 forced. I say "may" because I never totally pinned down the problem;
4053 setting up timeout/retry tests is difficult. See also the next item.
4055 79. The ultimate address timeout was not being applied to errors that involved
4056 a combination of host plus message (for example, a timeout on a MAIL
4057 command). When an address resolved to a number of possible hosts, and they
4058 were not all tried for each delivery (e.g. because of hosts_max_try), a
4059 message could remain on the queue longer than the retry timeout.
4061 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4064 81. Fixed an obscure SMTP outgoing bug which required at least the following
4065 conditions: (a) there was another message waiting for the same server;
4066 (b) the server returned 5xx to all RCPT commands in the first message so
4067 that the message was not completed; (c) the server dropped the connection
4068 or gave a negative response to the RSET that Exim sends to abort the
4069 transaction. The observed case was a dropped connection after DATA that had
4070 been sent in pipelining mode. That is, the server had advertised PIPELINING
4071 but was not implementing it correctly. The effect of the bug was incorrect
4072 behaviour, such as trying another host, and this could lead to a crash.
4078 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4079 and daemon.c were passed as pointers to ints; they should have been
4080 pointers to socklen_t variables (which are typically unsigned ints).
4082 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4085 3. Fixed a really odd bug that affected only the testing scheme; patching a
4086 certain fixed string in the binary changed the value of another string that
4087 happened to be identical to the end of the original first string.
4089 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4090 name", it returns that address as the IP address. On some operating
4091 systems (e.g. Solaris), it also passes back the IP address string as the
4092 "host name". However, on others (e.g. Linux), it passes back an empty
4093 string. Exim wasn't checking for this, and was changing the host name to an
4094 empty string, assuming it had been canonicized.
4096 5. Although rare, it is permitted to have more than one PTR record for a given
4097 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4098 all the names associated with an address, because they do in Solaris.
4099 However, it seems that they do not in Linux for data that comes from the
4100 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4101 I found this out when I moved to a new Linux workstation and tried to run
4102 the Exim test suite.
4104 To get round this problem I have changed the code so that it now does its
4105 own call to the DNS to look up PTR records when searching for a host name.
4106 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4107 addresses that are only in /etc/hosts are still found.
4109 This behaviour is, however, controlled by an option called host_lookup_
4110 order, which defaults to "bydns:byaddr". If people want to use the other
4111 order, or indeed, just use one or the other means of lookup, they can
4112 specify it in this variable.
4114 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4115 some operating systems, this comes back from gethostbyaddr() as an empty
4116 string, and this is what Exim used to test for. However, it seems that in
4117 other systems, "." is yielded. Exim now tests for this case too.
4119 7. The values of check_spool_space and check_log_space are now held internally
4120 as a number of kilobytes instead of an absolute number of bytes. If a
4121 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4122 kilobyte. This means that much larger values can be stored.
4124 8. Exim monitor: an attempt to get the action menu when not actually pointing
4125 at a message produces an empty menu entitled "No message selected". This
4126 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4127 no entries in it ("Shell widget menu has zero width and/or height"). So I
4128 have added a single, blank menu entry in this case.
4130 9. Added ${quote_local_part.
4132 10. MIME decoding is now applied to the contents of Subject: header lines when
4135 11. Now that a reference to $sender_host_address automatically causes a reverse
4136 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4137 host lookup before query-style lookups in lists that might use this
4138 variable. This has therefore been abolished, and the "net-" prefix is no
4139 longer necessary for query-style lookups.
4141 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4142 have been a typo for LFLAGS, so it has been changed.
4144 13. The install script calls Exim with "-C /dev/null" in order to find the
4145 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4146 to be output. Howeve, since Exim outputs its version number before the
4147 error, it didn't break the script. It just looked ugly. I fixed this by
4148 always allowing "-C /dev/null" if the caller is root.
4150 14. Ignore overlarge ACL variable number when reading spool file - insurance
4151 against a later release with more variables having written the file.
4153 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4154 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4155 this, as well as the form without the "IPv6" on the front.
4157 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4158 OS/Makefile-Darwin file.
4160 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4161 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4163 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4164 variable. However, some OS use ints for inodes. I've added cast to long int
4165 to get rid of the compiler warning.
4167 19. I had forgotten to lock out "/../" in configuration file names when
4168 ALT_CONFIG_PREFIX was set.
4170 20. Routers used for verification do not need to specify transports. However,
4171 if such a router generated a host list, and callout was configured, Exim
4172 crashed, because it could not find a port number from the (non-existent)
4173 transport. It now assumes port 25 in this circumstance.
4175 21. Added the -t option to exigrep.
4177 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4178 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4179 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4180 them is set, LOOKUP_LSEARCH is forced.
4182 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4184 24. Added sender and host information to the "rejected by local_scan()" log
4185 line; previously there was no indication of these.
4187 25. Added .include_if_exists.
4189 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4190 Linux. It turns out that not all file systems support this. Apparently some
4191 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4192 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4193 not supported on the file descriptor, is now ignored when Exim is trying to
4194 sync a directory. This applies only to Linux.
4196 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4198 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4199 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4202 29. Added support for maildirsize files from supplied patch (modified a bit).
4204 30. The use of :fail: followed by an empty string could lead Exim to respond to
4205 sender verification failures with (e.g.):
4207 550 Verification failed for <xxx>
4208 550 Sender verify failed
4210 where the first response line was missing the '-' that indicates it is not
4211 the final line of the response.
4213 31. The loop for finding the name of the user that called Exim had a hardwired
4214 limit of 10; it now uses the value of finduser_retries, which is used for
4215 all other user lookups.
4217 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4220 33. Exim was neglecting to zero errno before one call of strtol() when
4221 expanding a string and expecting an integer value. On some systems this
4222 resulted in spurious "integer overflow" errors. Also, it was casting the
4223 result into an int without checking.
4225 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4226 did not work. The code looks as if it has *never* worked, though it appears
4227 to have been documented since at least releast 1.62. I have made it work.
4229 35. The "timeout_DNS" error in retry rules, also documented since at least
4230 1.62, also never worked. As it isn't clear exactly what this means, and
4231 clearly it isn't a major issue, I have abolished the feature by treating it
4232 as "timeout", and writing a warning to the main and panic logs.
4234 36. The display of retry rules for -brt wasn't always showing the error code
4237 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4238 timeout_connect_A, timeout_connect_MX.
4240 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4241 to the empty sender.
4243 39. The daemon was not analysing the content of -oX till after it had closed
4244 stderr and disconnected from the controlling terminal. This meant that any
4245 syntax errors were only noted on the panic log, and the return code from
4246 the command was 0. By re-arranging the code a little, I've made the
4247 decoding happen first, so such errors now appear on stderr, and the return
4248 code is 1. However, the actual setting up of the sockets still happens in
4249 the disconnected process, so errors there are still only recorded on the
4252 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4253 connections (as happens on some IP stacks) was logged at start up time as
4254 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4255 it from "IPv6 and IPv4", which means that two separate sockets are being
4258 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4259 says whether AF_INET or AF_INET6 was passed as an argument.
4261 42. Exiwhat output was messed up when time zones were included in log
4264 43. Exiwhat now gives more information about the daemon's listening ports,
4265 and whether -tls-on-connect was used.
4267 44. The "port" option of the smtp transport is now expanded.
4269 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4270 silently ignored. Now an error message is written to the main and panic
4273 46. There's a new ACL modifier called "logwrite" which writes to a log file
4274 as soon as it is encountered.
4276 47. Added $local_user_uid and $local_user_gid at routing time.
4278 48. Exim crashed when trying to verify a sender address that was being
4281 49. Exim was recognizing only a space character after ".include". It now also
4282 recognizes a tab character.
4284 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4285 extracting the relevant information from the specification. The man page no
4286 longer contains scrambled data for the -d option, and I've added a section
4287 at the front about calling Exim under different names.
4289 51. Added "extra_headers" argument to the "mail" command in filter files.
4291 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4294 53. Installed eximstats 1.29.
4296 54. Added transport_filter_timeout as a generic transport option.
4298 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4299 Cc: header lines. This was required by RFC 822, but it not required by RFC
4302 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4303 incoming messages that did not have them. Now it does so only if the
4304 message originates locally, that is, if there is no associated remote host
4305 address. When Resent- header lines are present, this applies to the Resent-
4306 lines rather than the non-Resent- lines.
4308 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4309 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4311 58. Messages for configuration errors now include the name of the main
4312 configuration file - useful now that there may be more than one file in a
4313 list (.included file names were always shown).
4315 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4316 for those rare installations that do not start the daemon as root or run it
4317 setuid root. I've cut out the call to initgroups() if the daemon is not
4320 60. The Exim user and group can now be bound into the binary as text strings
4321 that are looked up at the start of Exim's processing.
4323 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4325 62. Added $mailstore_basename variable.
4327 63. Installed patch to sieve.c from Michael Haardt.
4329 64. When Exim failed to open the panic log after failing to open the main log,
4330 the original message it was trying to log was written to stderr and debug
4331 output, but if they were not available (the usual case in production), it
4332 was lost. Now it is written to syslog before the two lines that record the
4333 failures to open the logs.
4335 65. Users' Exim filters run in subprocesses under the user's uid. It is
4336 possible for a "deliver" command or an alias in a "personal" command to
4337 provoke an address rewrite. If logging of address rewriting is configured,
4338 this fails because the process is not running as root or exim. There may be
4339 a better way of dealing with this, but for the moment (because 4.30 needs
4340 to be released), I have disabled address rewrite logging when running a
4341 filter in a non-root, non-exim process.
4347 1. The buildconfig auxiliary program wasn't quoting the value set for
4348 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4349 not defined. This bug was masked in 4.22 by the effect that was fixed in
4352 2. Some messages that were rejected after a message id was allocated were
4353 shown as "incomplete" by exigrep. It no longer does this for messages that
4354 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4356 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4357 have allow_domain_literals set, the ID did not get logged in the <= line.
4358 Domain literals are now always recognized in Message-ID: header lines.
4360 4. The first argument for a ${extract expansion item is the key name or field
4361 number. Leading and trailing spaces in this item were not being ignored,
4362 causing some misleading effects.
4364 5. When deliver_drop_privilege was set, single queue runner processes started
4365 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4366 same command in the process it spins off) were not dropping privilege.
4368 6. When the daemon running as "exim" started a queue runner, it always
4369 re-executed Exim in the spun-off process. This is a waste of effort when
4370 deliver_drop_privilege is set. The new process now just calls the
4371 queue-runner function directly.
4377 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4380 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4381 ignored. Though the use of -oP was forcing the writing of a pid file, it
4382 was always written to the default place.
4384 3. If the message "no IP address found for host xxxx" is generated during
4385 incoming verification, it is now followed by identification of the incoming
4386 connection (so you can more easily find what provoked it).
4388 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4390 5. Added some features to "harden" Exim a bit more against certain attacks:
4392 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4393 be put in Local/Makefile. This is like the never_users runtime option,
4394 but it cannot be overridden. The default setting is "root".
4396 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4397 prefix string with which any file named in a -C command line option
4400 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4401 is retained for -C and -D only if the caller of Exim is root. Without
4402 it, the exim user may also use -C and -D and retain privilege.
4404 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4405 command line option is disabled.
4407 6. Macro names set by the -D option must start with an upper case letter, just
4408 like macro names defined in the configuration file.
4410 7. Added "dereference=" facility to LDAP.
4412 8. Two instances of the typo "uknown" in the source files are fixed.
4414 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4415 the Configure-Makefile script screwed up while processing it.
4417 10. Incorporated PCRE 4.4.
4419 11. The SMTP synchronization check was not operating right at the start of an
4420 SMTP session. For example, it could not catch a HELO sent before the client
4421 waited for the greeting. There is now a check for outstanding input at the
4422 point when the greeting is written. Because of the duplex, asynchronous
4423 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4424 way, but not yet received, when the check is performed.
4426 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4427 on TCP/IP sockets, because this apparently causes some broken clients to
4430 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4431 unchanged) from the Cygwin maintainer.
4433 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4434 was supported instead of testing for mbx. Effectively a typo.
4436 15. The spa authenticator server code was not checking that the input it
4437 received was valid base64.
4439 16. The debug output line for the "set" modifier in ACLs was not showing the
4440 name of the variable that was being set.
4442 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4444 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4445 lookup on its own. Something else was needed to trigger the lookup. For
4446 example, a match in host_lookup or the need for a host name in a host list.
4447 Now, if $sender_host_name is referenced and the host name has not yet been
4448 looked up, a lookup is performed. If the lookup fails, the variable remains
4449 empty, and $host_lookup_failed is set to "1".
4451 19. Added "eqi" as a case-independent comparison operator.
4453 20. The saslauthd authentication condition could segfault if neither service
4454 nor realm was specified.
4456 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4457 error message that was logged was misleading, and incoming SMTP
4458 connections were dropped. The message is now more accurate, and temporary
4459 errors are given to SMTP connections.
4461 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4462 (except RSET) with a 421 error, until QUIT is received. However, it was
4463 failing to send a response to QUIT.
4465 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4466 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4467 ACL is now run after the helo_try_verify_hosts code.
4469 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4470 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4471 case-independent, so other case variants are also recognized). Apparently
4472 some systems use these upper case variants.
4474 25. If more than two messages were waiting for the same host, and a transport
4475 filter was specified for the transport, Exim sent two messages over the
4476 same TCP/IP connection, and then failed with "socket operation on non-
4477 socket" when it tried to send the third.
4479 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4481 27. The extern definition of crypt16() in expand.c was not being excluded when
4482 the OS had its own crypt16() function.
4484 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4485 as a preferred synonym for return_size_limit, both as an option and as an
4488 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4490 30. Changed the default configuration ACL to relax the local part checking rule
4491 for addresses that are not in any local domains. For these addresses,
4492 slashes and pipe symbols are allowed within local parts, but the sequence
4493 /../ is explicitly forbidden.
4495 31. SPA server authentication was not clearing the challenge buffer before
4498 32. log_message in a "warn" ACL statement was writing to the reject log as
4499 well as to the main log, which contradicts the documentation and doesn't
4500 seem right (because no rejection is happening). So I have stopped it.
4502 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4503 However, I am unable to do any testing of this.
4505 34. Fixed an infelicity in the appendfile transport. When checking directories
4506 for a mailbox, to see if any needed to be created, it was accidentally
4507 using path names with one or more superfluous leading slashes; tracing
4508 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4510 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4511 contents are added to the log line that is written for every discarded
4512 recipient. (Previously a log_message setting was ignored.)
4514 36. The ${quote: operator now quotes the string if it is empty.
4516 37. The install script runs exim in order to find its version number. If for
4517 some reason other than non-existence or emptiness, which it checks, it
4518 could not run './exim', it was installing it with an empty version number,
4519 i.e. as "exim-". This error state is now caught, and the installation is
4522 38. An argument was missing from the function that creates an error message
4523 when Exim fails to connect to the socket for saslauthd authentication.
4524 This could cause Exim to crash, or give a corrupted message.
4526 39. Added isip, isip4, and isip6 to ${if conditions.
4528 40. The ACL variables $acl_xx are now saved with the message, and can be
4529 accessed later in routers, transports, and filters.
4531 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4532 strings in the file are not string-expanded.
4534 42. If a MAIL command specified a SIZE value that was too large to fit into an
4535 int variable, the check against message_size_limit failed. Such values are
4536 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4537 day this will have to be increased, but I don't think I want to be around
4538 when emails are that large.
4545 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4546 iconv() is not standard in FreeBSD.
4548 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4549 IPv6 enabled. The observed symptom was a segmentation fault on return from
4550 the function os_common_find_running_interfaces() in src/os.c.
4552 3. In the check_special_case() function in daemon.c I had used "errno" as an
4553 argument name, which causes warnings on some systems. This was basically a
4554 typo, since it was named "eno" in the comments!
4556 4. The code that waits for the clock to tick (at a resolution of some fraction
4557 of a second) so as to ensure message-id uniqueness was always waiting for
4558 at least one whole tick, when it could have waited for less. [This is
4559 almost certainly not relevant at current processor speeds, where it is
4560 unlikely to ever wait at all. But we try to future-proof.]
4562 5. The function that sleeps for a time interval that includes fractions of a
4563 second contained a race. It did not block SIGALRM between setting the
4564 timer, and suspending (a couple of lines later). If the interval was short
4565 and the sigsuspend() was delayed until after it had expired, the suspension
4566 never ended. On busy systems this could lead to processes getting stuck for
4569 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4570 process, before it forks any delivery processes. The open lookup caching
4571 mechanism meant that the open file or database connection was passed into
4572 the delivery process. The problem was that delivery processes always tidy
4573 up cached lookup data. This could cause a problem for the next delivery
4574 process started by the queue runner, because the external queue runner
4575 process does not know about the closure. So the next delivery process
4576 still has data in the lookup cache. In the case of a file lookup, there was
4577 no problem because closing a file descriptor in a subprocess doesn't affect
4578 the parent. However, if the lookup was caching a connection to a database,
4579 the connection was closed, and the second delivery process was likely to
4580 see errors such as "PGSQL: query failed: server closed the connection
4581 unexpectedly". The problem has been fixed by closing all cached lookups
4582 in a queue runner before running a delivery process.
4584 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4585 seem to have the "const" qualifier which it has on other OS. I've
4588 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4589 *of the same type* (client or server) with the same public name that an
4590 error should be diagnosed.
4592 9. When Exim looked up a host name for an IP address, but failed to find the
4593 original IP address when looking up the host name (a safety check), it
4594 output the message "<ip address> does not match any IP for NULL", which was
4595 confusing, to say the least. The bug was that the host name should have
4596 appeared instead of "NULL".
4598 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4599 user that is built into the binary, and the -C or -D options is used, root
4600 privilege is dropped before the configuration file is read. In addition,
4601 logging is switched to stderr instead of the normal log files. If the
4602 configuration then re-defines the Exim user, the unprivileged environment
4603 is probably not what is expected, so Exim logs a panic warning message (but
4606 However, if deliver_drop_privilege is set, the unprivileged state may well
4607 be exactly what is intended, so the warning has been cut out in that case,
4608 and Exim is allowed to try to write to its normal log files.
4614 1. smtp_return_error_details was not giving details for temporary sender
4615 or receiver verification errors.
4617 2. Diagnose a configuration error if two authenticators have the same public
4620 3. Exim used not to create the message log file for a message until the first
4621 delivery attempt. This could be confusing when incoming messages were held
4622 for policy or load reasons. The message log file is now created at the time
4623 the message is received, and an initial "Received" line is written to it.
4625 4. The automatically generated man page for command line options had a minor
4626 bug that caused no ill effects; however, a more serious problem was that
4627 the procedure for building the man page automatically didn't always
4628 operate. Consequently, release 4.20 contains an out-of-date version. This
4629 shouldn't happen again.
4631 5. When building Exim with embedded Perl support, the script that builds the
4632 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4633 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4635 6. The freeze_tell option was not being used for messages that were frozen on
4636 arrival, either by an ACL or by local_scan().
4638 7. Added the smtp_incomplete_transaction log selector.
4640 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4641 was accepting AUTH without a new EHLO.
4643 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4644 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4647 10. Logging of TCP/IP connections (when configured) now happens in the main
4648 daemon process instead of the child process, so that the TCP/IP connection
4649 count is more accurate (but it can never be perfect).
4651 11. The use of "drop" in a nested ACL was not being handled correctly in the
4652 outer ACL. Now, if condition failure induced by the nested "drop" causes
4653 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4654 or "require"), the connection is dropped.
4656 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4657 that yield "discard" can now be used with an "accept" or a "discard" verb,
4658 but an error is generated for any others (because I can't see a useful way
4659 to define what should happen).
4661 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4662 are now processed in the same way as lines in the Exim configuration file.
4663 In particular, continuation lines are supported.
4665 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4667 15. Added -ti meaning -t -i.
4669 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4670 domains, and warn by logging if others are found.
4672 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4673 returned from GIFCONF when Exim is trying to find the list of interfaces on
4674 a host. The code in os.c has been modified to copy each ifreq to an aligned
4675 structure in all cases.
4677 Also, in some cases, the returned ifreq's were being copied to a 'struct
4678 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4679 means the last couple of bytes of an IPv6 address could be chopped if the
4680 ifreq contained only a normal sockaddr (14 bytes storage).
4682 18. Named domain lists were not supported in the hosts_treat_as_local option.
4683 An entry such as +xxxx was not recognized, and was treated as a literal
4686 19. Ensure that header lines added by a DATA ACL are included in the reject log
4687 if the ACL subsequently rejects the message.
4689 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4690 MD5 (which is deprecated).
4692 21. When testing a filter file using -bf, Exim was writing a message when it
4693 took the sender from a "From " line in the message, but it was not doing so
4694 when it took $return_path from a Return-Path: header line. It now does.
4696 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4697 with a valid header line field name (a series of printing characters
4698 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4700 23. Changed "disc" in the source to "disk" to conform to the documentation and
4701 the book and for uniformity.
4703 24. Ignore Sendmail's -Ooption=value command line item.
4705 25. When execve() failed while trying to run a command in a pipe transport,
4706 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4707 could be confused with a return value of 69 from the command itself. This
4708 has been changed to 127, the value the shell returns if it is asked to run
4709 a non-existent command. The wording for the related log line suggests a
4710 non-existent command as the problem.
4712 26. If received_header_text expands to an empty string, do not add a Received:
4713 header line to the message. (Well, it adds a token one on the spool, but
4714 marks it "old" so that it doesn't get used or transmitted.)
4716 27. Installed eximstats 1.28 (addition of -nt option).
4718 28. There was no check for failure on the call to getsockname() in the daemon
4719 code. This can fail if there is a shortage of resources on the system, with
4720 ENOMEM, for example. A temporary error is now given on failure.
4722 29. Contrary to the C standard, it seems that in some environments, the
4723 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4724 program. Exim now does this explicitly; it affects the formatting of
4725 timestamps using strftime().
4727 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4728 complaints. I've now initialized all the variables, to avoid this.
4730 32. Header lines added by a system filter were not being "seen" during
4731 transport-time rewrites.
4733 33. The info_callback() function passed to OpenSSL is set up with type void
4734 (*)(SSL *, int, int), as described somewhere. However, when calling the
4735 function (actually a macro) that sets it up, the type void(*)() is
4736 expected. I've put in a cast to prevent warnings from picky compilers.
4738 34. If a DNS black list lookup found a CNAME record, but there were no A
4739 records associated with the domain it pointed at, Exim crashed.
4741 35. If a DNS black list lookup returned more than one A record, Exim ignored
4742 all but the first. It now scans all returned addresses if a particular IP
4743 value is being sought. In this situation, the contents of the
4744 $dnslist_value variable are a list of all the addresses, separated by a
4747 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4748 to accept a host name and all its aliases if the forward lookup for any of
4749 them yielded the IP address of the incoming connection. Now it accepts only
4750 those names whose forward lookup yields the correct IP address. Any other
4751 names are discarded. This closes a loophole whereby a rogue DNS
4752 administrator could create reverse DNS records to break through a
4753 wildcarded host restriction in an ACL.
4755 37. If a user filter or a system filter that ran in a subprocess used any of
4756 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4757 the wrong values were passed to the pipe command ($thisaddress had the
4758 value of $0, $0 had the value of $1, etc). This bug was introduced by
4759 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4761 38. Improved the line breaking for long SMTP error messages from ACLs.
4762 Previously, if there was no break point between 40 and 75 characters, Exim
4763 left the rest of the message alone. Two changes have been made: (a) I've
4764 reduced the minimum length to 35 characters; (b) if it can't find a break
4765 point between 35 and 75 characters, it looks ahead and uses the first one
4766 that it finds. This may give the occasional overlong line, but at least the
4767 remaining text gets split now.
4769 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4770 file descriptors might be low, and that setting 1000 would always raise it.
4771 It turns out that in some environments, the limit is already over 1000 and
4772 that lowering it causes trouble. So now Exim takes care not to decrease it.
4774 40. When delivering a message, the value of $return_path is set to $sender_
4775 address at the start of routing (routers may change the value). By an
4776 oversight, this default was not being set up when an address was tested by
4777 -bt or -bv, which affected the outcome if any router or filter referred to
4780 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4781 log only when "message" or "log_message" are set. However, if one of the
4782 conditions was an address verification, or a call to a nested ACL, the
4783 messages generated by the underlying test were being passed through. This
4784 no longer happens. The underlying message is available in $acl_verify_
4785 message for both "message" and "log_message" expansions, so it can be
4786 passed through if needed.
4788 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4789 new expansion $bh_ to give the encoded byte string without charset
4790 translation. Translation happens only if iconv() is available; HAVE_ICONV
4791 indicates this at build time. HEADERS_CHARSET gives the charset to
4792 translate to; headers_charset can change it in the configuration, and
4793 "headers charset" can change it in an individual filter file.
4795 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4796 that creates RFC 2047 encoded "words" labels them as that charset instead
4797 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4798 expansion operator; (ii) when Exim creates a From: line for a local
4799 message; (iii) when a header line is rewritten to include a "phrase" part.
4801 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4802 buggy, causing it to skip the first lines of messages whose message ID
4803 ended in 'D'. This would not have bitten before Exim release 4.14, because
4804 message IDs were unlikely to end in 'D' before then. The effect was to have
4805 incorrect size information for certain domains.
4807 45. #include "config.h" was missing at the start of the crypt16.c module. This
4808 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4811 46. If there was a timeout during a "random" callout check, Exim treated it as
4812 a failure of the random address, and carried on sending RSET and the real
4813 address. If the delay was just some slowness somewhere, the response to the
4814 original RCPT would be taken as a response to RSET and so on, causing
4815 mayhem of various kinds.
4817 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4818 when I implemented it. It didn't allow for the fact that some option values
4819 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4820 the right test for positive values.
4822 48. Domain names in DNS records are case-independent. Exim always looks them up
4823 in lower case. Some resolvers return domain names in exactly the case they
4824 appear in the zone file, that is, they may contain uppercase letters. Not
4825 all resolvers do this - some return always lower case. Exim was treating a
4826 change of case by a resolver as a change of domain, similar to a widening
4827 of a domain abbreviation. This triggered its re-routing code and so it was
4828 trying to route what was effectively the same domain again. This normally
4829 caused routing to fail (because the router wouldn't handle the domain
4830 twice). Now Exim checks for this case specially, and just changes the
4831 casing of the domain that it ultimately uses when it transmits the message
4834 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4837 50. If a filter generated a file delivery with a non-absolute name (possible if
4838 no home directory exists for the router), the forbid_file option was not
4841 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4842 the existing equality matching.
4844 52. Exim was using ints instead of ino_t variables in some places where it was
4845 dealing with inode numbers.
4847 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4848 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4849 called TMPDIR, and if it finds it is different, it changes its value.
4851 54. The smtp_printf() function is now made available to local_scan() so
4852 additional output lines can be written before returning. There is also an
4853 smtp_fflush() function to enable the detection of a dropped connection.
4854 The variables smtp_input and smtp_batched_input are exported to
4857 55. Changed the default runtime configuration: the message "Unknown user"
4858 has been removed from the ACL, and instead placed on the localuser router,
4859 using the cannot_route_message feature. This means that any verification
4860 failures that generate their own messages won't get overridden. Similarly,
4861 the "Unrouteable address" message that was in the ACL for unverifiable
4862 relay addresses has also been removed.
4864 56. Added hosts_avoid_esmtp to the smtp transport.
4866 57. The exicyclog script was not checking for the esoteric option
4867 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4868 will work only if exicyclog is run under the appropriate euid.
4870 58. Following a discussion on the list, the rules by which Exim recognises line
4871 endings on incoming messages have been changed. The -dropcr and drop_cr
4872 options are now no-ops, retained only for backwards compatibility. The
4873 following line terminators are recognized: LF CRLF CR. However, special
4874 processing applies to CR:
4876 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4877 nor a local message in the state where . is a terminator.
4879 (ii) If a bare CR is encountered in a header line, an extra space is added
4880 after the line terminator so as not to end the header. The reasoning
4881 behind this is that bare CRs in header lines are most likely either
4882 to be mistakes, or people trying to play silly games.
4884 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4885 was being incorrectly given as 18 bytes larger than it should have been.
4886 This is a VOB (very old bug).
4888 60. This may never have affected anything current, but just in case it has:
4889 When the local host is found other than at the start of a list of hosts,
4890 the local host, those with the same MX, and any that follow, are discarded.
4891 When the list in question was part of a longer list of hosts, the following
4892 hosts (not currently being processed) were also being discarded. This no
4893 longer happens. I'm not sure if this situation could ever has previously
4896 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4899 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4900 header. This is something that is recommended in a new Internet Draft, and
4901 is something that is documented as being done by Sendmail. There are two
4902 possible values. For messages generated by the autoreply transport, Exim
4905 Auto-Submitted: auto-replied
4907 whereas for all other generated messages (e.g. bounces) it adds
4909 Auto-Submitted: auto-generated
4911 63. The "personal" condition in filters now includes a test for the
4912 Auto-Submitted: header. If it contains the string "auto-" the message it
4913 not considered personal.
4915 64. Added rcpt_include_affixes as a generic transport option.
4917 65. Added queue_only_override (default true).
4919 66. Added the syslog_duplication option.
4921 67. If what should have been the first header line of a message consisted of
4922 a space followed by a colon, Exim was mis-interpreting it as a header line.
4923 It isn't of course - it is syntactically invalid and should therefore be
4924 treated as the start of the message body. The misbehaviour could have
4925 caused a number of strange effects, including loss of data in subsequent
4926 header lines, and spool format errors.
4928 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4929 client host had authenticated. This control can now be exercised by an ACL
4930 for more flexibility.
4932 69. By default, callouts do not happen when testing with -bh. There is now a
4933 variant, -bhc, which does actually run the callout code, including
4934 consulting and updating the callout cache.
4936 70. Added support for saslauthd authentication, courtesy of Alexander
4939 71. If statvfs() failed on the spool or log directories while checking their
4940 size for availability, Exim confusingly gave the error "space shortage".
4941 Furthermore, in debugging mode it crashed with a floating point exception.
4942 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4943 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4944 serious problem, Exim now writes to the main and panic logs when this
4945 happens, with details of the failure. It then refuses to accept the
4946 incoming message, giving the message "spool directory problem" or "log
4947 directory problem" with a 421 code for SMTP messages.
4949 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4950 0, 1, and 2 exist, because some OS complain for execs without them (see
4951 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4952 descriptors. However, the code omitted to check that the open succeeded,
4953 causing mysterious errors if for some reason the permissions on /dev/null
4954 got screwed. Now Exim writes a message to the main and panic logs, and
4955 bombs out if it can't open /dev/null.
4957 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4958 interact so that it is all more flexible. It is supposed to remain
4959 backwards compatible. Also added extra_local_interfaces.
4961 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4962 to bomb out with an assertion failure - to the client this appears as a
4963 connection drop. This problem occurs in the part of the code that was taken
4964 from the Samba project. Fortunately, the assertion is in a very simple
4965 function, so I have fixed this by reproducing the function inline in the
4966 one place where it is called, and arranging for authentication to fail
4967 instead of killing the process with assert().
4969 75. The SPA client code was not working when the server requested OEM rather
4970 than Unicode encoding.
4972 76. Added code to make require_files with a specific uid setting more usable in
4973 the case where statting the file as root fails - usually a non-root-mounted
4974 NFS file system. When this happens and the failure is EACCES, Exim now
4975 forks a subprocess and does the per-uid checking as the relevant uid.
4977 77. Added process_log_path.
4979 78. If log_file_path was not explicitly set, a setting of check_log_space or
4980 check_log_inodes was ignored.
4982 79. If a space check for the spool or log partitions fails, the incident is now
4983 logged. Of course, in the latter case the data may get lost...
4985 80. Added the %p formatting code to string_format() so that it can be used to
4986 print addresses in debug_print(). Adjusted all the address printing in the
4987 debugging in store.c to use %p rather than %d.
4989 81. There was a concern that a line of code in smtp_in.c could overflow a
4990 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4991 initially expressed, the concern was not well-founded, because trailing
4992 spaces are removed early. However, if the trailing spaces were followed by
4993 a NULL, they did not get removed, so the overflow was possible. Two fixes
4996 (a) I re-wrote the offending code in a cleaner fashion.
4997 (b) If an incoming SMTP command contains a NULL character, it is rejected
5000 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5001 runs initgroups(), so that if the Exim user is in any additional groups,
5002 they will be used during message reception.
5008 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob