1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.570 2009/10/16 09:51:12 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 614: Added timeout setting to transport filter.
95 Patch provided by Dean Brooks
97 TK/05 Add native DKIM support (does not depend on external libraries).
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
121 TK/01 Add preliminary DKIM support. Currently requires a forked version of
122 ALT-N's libdkim that I have put here:
123 http://duncanthrax.net/exim-experimental/
125 Note to Michael Haardt: I had to rename some vars in sieve.c. They
126 were called 'true' and it seems that C99 defines that as a reserved
127 keyword to be used with 'bool' variable types. That means you could
128 not include C99-style headers which use bools without triggering
129 build errors in sieve.c.
131 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
132 as mailq or other aliases. Changed the --help handling significantly
133 to do whats expected. exim_usage() emits usage/help information.
135 SC/01 Added the -bylocaldomain option to eximstats.
137 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
139 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
141 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
147 PH/01 Another patch from the Sieve maintainer.
149 PH/02 When an IPv6 address is converted to a string for single-key lookup
150 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
151 dots are used instead of colons so that keys in lsearch files need not
152 contain colons. This was done some time before quoting was made available
153 in lsearch files. However, iplsearch files do require colons in IPv6 keys
154 (notated using the quote facility) so as to distinguish them from IPv4
155 keys. This meant that lookups for IP addresses in host lists did not work
156 for iplsearch lookups.
158 This has been fixed by arranging for IPv6 addresses to be expressed with
159 colons if the lookup type is iplsearch. This is not incompatible, because
160 previously such lookups could never work.
162 The situation is now rather anomolous, since one *can* have colons in
163 ordinary lsearch keys. However, making the change in all cases is
164 incompatible and would probably break a number of configurations.
166 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
169 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
170 conversion specification without a maximum field width, thereby enabling
171 a rogue spamd server to cause a buffer overflow. While nobody in their
172 right mind would setup Exim to query an untrusted spamd server, an
173 attacker that gains access to a server running spamd could potentially
174 exploit this vulnerability to run arbitrary code as the Exim user.
176 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
177 $primary_hostname instead of what libspf2 thinks the hosts name is.
179 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
180 a directory entry by the name of the lookup key. Previously, if a
181 symlink pointed to a non-existing file or a file in a directory that
182 Exim lacked permissions to read, a lookup for a key matching that
183 symlink would fail. Now it is enough that a matching directory entry
184 exists, symlink or not. (Bugzilla 503.)
186 PH/03 The body_linecount and body_zerocount variables are now exported in the
189 PH/04 Added the $dnslist_matched variable.
191 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
192 This means they are set thereafter only if the connection becomes
195 PH/06 Added the client_condition to authenticators so that some can be skipped
196 by clients under certain conditions.
198 PH/07 The error message for a badly-placed control=no_multiline_responses left
199 "_responses" off the end of the name.
201 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
203 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
204 (without spaces) instead of just copying the configuration text.
206 PH/10 Added the /noupdate option to the ratelimit ACL condition.
208 PH/11 Added $max_received_linelength.
210 PH/12 Added +ignore_defer and +include_defer to host lists.
212 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
213 way in which PCRE > 7.0 is built.
215 PH/14 Implemented queue_only_load_latch.
217 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
218 MAIL command. The effect was to mangle the value on 64-bit systems.
220 PH/16 Another patch from the Sieve maintainer.
222 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
224 PH/18 If a system quota error occurred while trying to create the file for
225 a maildir delivery, the message "Mailbox is full" was not appended to the
226 bounce if the delivery eventually timed out. Change 4.67/27 below applied
227 only to a quota excession during the actual writing of the file.
229 PH/19 It seems that peer DN values may contain newlines (and other non-printing
230 characters?) which causes problems in log lines. The DN values are now
231 passed through string_printing() before being added to log lines.
233 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
234 and InterBase are left for another time.)
236 PH/21 Added message_body_newlines option.
238 PH/22 Guard against possible overflow in moan_check_errorcopy().
240 PH/23 POSIX allows open() to be a macro; guard against that.
242 PH/24 If the recipient of an error message contained an @ in the local part
243 (suitably quoted, of course), incorrect values were put in $domain and
244 $local_part during the evaluation of errors_copy.
250 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
251 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
254 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
255 issue a MAIL command.
257 PH/02 In an ACL statement such as
259 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
261 if a client was not listed at all, or was listed with a value other than
262 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
263 the condition was not true (as it should be), so access was not denied.
264 The bug was that the ! inversion was incorrectly passed on to the second
265 item. This has been fixed.
267 PH/03 Added additional dnslists conditions == and =& which are different from
268 = and & when the dns lookup returns more than one IP address.
270 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
271 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
273 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
274 FSYNC, which compiles an option called disable_fsync that allows for
275 bypassing fsync(). The documentation is heavily laced with warnings.
277 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
279 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
280 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
281 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
282 including adding "make clean"; (3) Added -fPIC when compiling the test
283 dynamically loaded module, to get rid of a warning.
285 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
286 message fails, move_frozen_messages = true and ignore_bounce_errors_after
287 = 0s. The bug is otherwise harmless.
289 PH/07 There was a bug in the dovecot authenticator such that the value of
290 $auth1 could be overwritten, and so not correctly preserved, after a
291 successful authentication. This usually meant that the value preserved by
292 the server_setid option was incorrect.
294 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
296 PH/09 Installed PCRE release 7.0.
298 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
299 run for batched SMTP input. It is now run at the start of every message
300 in the batch. While fixing this I discovered that the process information
301 (output by running exiwhat) was not always getting set for -bs and -bS
302 input. This is fixed, and it now also says "batched" for BSMTP.
304 PH/11 Added control=no_pipelining.
306 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
307 patch, slightly modified), and move the expansion of helo_data till after
308 the connection is made in the smtp transport (so it can use these
311 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
313 PH/14 Added log_selector = +pid.
315 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
317 PH/16 Add ${if forany and ${if forall.
319 PH/17 Added dsn_from option to vary the From: line in DSNs.
321 PH/18 Flush SMTP output before performing a callout, unless control =
322 no_callout_flush is set.
324 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
325 was true (the default) a successful delivery failed to delete the retry
326 item, thus causing premature timeout of the address. The bug is now
329 PH/20 Added hosts_avoid_pipelining to the smtp transport.
331 PH/21 Long custom messages for fakedefer and fakereject are now split up
332 into multiline reponses in the same way that messages for "deny" and
333 other ACL rejections are.
335 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
336 with slight modification.
338 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
339 draft, changing the syntax and factoring some duplicate code".
341 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
342 for deliveries of the second and subsequent messages over the same SMTP
345 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
346 ${reduce, with only minor "tidies".
348 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
350 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
351 expansion side effects.
353 PH/27 When a message times out after an over-quota error from an Exim-imposed
354 quota, the bounce message says "mailbox is full". This message was not
355 being given when it was a system quota that was exceeded. It now should
358 MH/03 Made $recipients available in local_scan(). local_scan() already has
359 better access to the recipient list through recipients_list[], but
360 $recipients can be useful in postmaster-provided expansion strings.
362 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
363 in the case of a MAIL command with additional options following the
364 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
365 were accidentally chopped off.
367 PH/29 SMTP synchronization checks are implemented when a command is read -
368 there is a check that no more input is waiting when there shouldn't be
369 any. However, for some commands, a delay in an ACL can mean that it is
370 some time before the response is written. In this time, more input might
371 arrive, invalidly. So now there are extra checks after an ACL has run for
372 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
373 pipelining has not been advertised.
375 PH/30 MH's patch to allow iscntrl() characters to be list separators.
377 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
378 returned in the SMTP response when smtp_return_error_details was false.
381 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
382 instead of the C I/O that was originally supplied, because problems were
385 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
386 Exim which did not show up earlier: it was assuming that a call to
387 SSL_CTX_set_info_callback() might give an error value. In fact, there is
388 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
389 was a macro that became an assignment, so it seemed to work. This has
390 changed to a proper function call with a void return, hence the compile
391 error. Exim's code has been fixed.
393 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
396 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
398 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
400 PH/37 If a message is not accepted after it has had an id assigned (e.g.
401 because it turns out to be too big or there is a timeout) there is no
402 "Completed" line in the log. When some messages of this type were
403 selected by exigrep, they were listed as "not completed". Others were
404 picked up by some special patterns. I have improved the selection
405 criteria to be more general.
407 PH/38 The host_find_failed option in the manualroute router can now be set
408 to "ignore", to completely ignore a host whose IP address cannot be
409 found. If all hosts are ignored, the behaviour is controlled by the new
410 host_all_ignored option.
412 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
413 homing or because of multiple MX records with /mx) generated more than
414 one IP address, and the following item turned out to be the local host,
415 all the secondary addresses of the first item were incorrectly removed
416 from the list, along with the local host and any following hosts (which
417 is what is supposed to happen).
419 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
420 whoever called Exim into the -H file. In the case of the daemon it was
421 behaving confusingly. When first started, it used values for whoever
422 started the daemon, but after a SIGHUP it used the Exim user (because it
423 calls itself on a restart). I have changed the code so that it now always
426 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
427 message are rejected with the same error (e.g. no authentication or bad
428 sender address), and a DATA command is nevertheless sent (as can happen
429 with PIPELINING or a stupid MUA), the error message that was given to the
430 RCPT commands is included in the rejection of the DATA command. This is
431 intended to be helpful for MUAs that show only the final error to their
434 PH/42 Another patch from the Sieve maintainer.
436 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
437 Eximstats - Fixed some broken HTML links and added missing column headers
439 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
440 columns for Rejects, Temp Rejects, Ham, and Spam rows.
442 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
444 PH/43 Yet another patch from the Sieve maintainer.
446 PH/44 I found a way to check for a TCP/IP connection going away before sending
447 the response to the final '.' that terminates a message, but only in the
448 case where the client has not sent further data following the '.'
449 (unfortunately, this is allowed). However, in many cases there won't be
450 any further data because there won't be any more messages to send. A call
451 to select() can be used: if it shows that the input is "ready", there is
452 either input waiting, or the socket has been closed. An attempt to read
453 the next input character can distinguish the two cases. Previously, Exim
454 would have sent an OK response which the client would never have see.
455 This could lead to message repetition. This fix should cure that, at
456 least in a lot of common cases.
458 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
459 advertised in response to EHLO.
465 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
466 fixed by 4.65/MH/01 (is this a record?) are fixed:
468 (i) An empty string was always treated as zero by the numeric comparison
469 operators. This behaviour has been restored.
471 (ii) It is documented that the numeric comparison operators always treat
472 their arguments as decimal numbers. This was broken in that numbers
473 starting with 0 were being interpreted as octal.
475 While fixing these problems I realized that there was another issue that
476 hadn't been noticed. Values of message_size_limit (both the global option
477 and the transport option) were treated as octal if they started with 0.
478 The documentation was vague. These values are now always treated as
479 decimal, and I will make that clear in the documentation.
485 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
486 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
489 MH/01 Don't check that the operands of numeric comparison operators are
490 integers when their expansion is in "skipping" mode (fixes bug
491 introduced by 4.64-PH/07).
493 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
494 child addresses, Exim now panics and dies. Previously, because the count
495 is held in a short int, deliveries were likely to be lost. As such a
496 large number of recipients for a single message is ridiculous
497 (performance will be very, very poor), I have chosen to impose a limit
498 rather than extend the field.
504 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
505 leftover -K file (the existence of which was triggered by #402).
506 While we were at it, introduced process PID as part of the -K
507 filename. This should rule out race conditions when creating
510 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
511 processing considerably. Previous code took too long for large mails,
512 triggering a timeout which in turn triggers #401.
514 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
515 in the DK code in transports.c. sendfile() is not really portable,
516 hence the _LINUX specificness.
518 TF/01 In the add_headers option to the mail command in an Exim filter,
519 there was a bug that Exim would claim a syntax error in any
520 header after the first one which had an odd number of characters
523 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
524 callout verification, Exim cached a "reject" for the entire domain. This
525 is correct for most verifications, but it is not correct for a recipient
526 verification with use_sender or use_postmaster set, because in that case
527 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
528 case of MAIL FROM:<> rejection from other early rejections (e.g.
529 rejection of HELO). When verifying a recipient using a non-null MAIL
530 address, the cache is ignored if it shows MAIL FROM:<> rejection.
531 Whatever the result of the callout, the value of the domain cache is
532 left unchanged (for any other kind of callout, getting as far as trying
533 RCPT means that the domain itself is ok).
535 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
538 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
539 manifest itself as EPIPE rather than ECONNECT. When tidying away a
540 session, the daemon ignores ECONNECT errors and logs others; it now
541 ignores EPIPE as well.
543 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
544 (quoted-printable decoding).
546 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
547 later the small subsequent patch to fix an introduced bug.
549 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
551 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
553 PH/08 An error is now given if message_size_limit is specified negative.
555 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
556 to be given (somewhat) arbitrary names.
558 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
561 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
562 miscellaneous code fixes
564 PH/10 Added the log_reject_target ACL modifier to specify where to log
567 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
568 hostname. This is wrong, because it relates to the incoming message (and
569 probably the interface on which it is arriving) and not to the outgoing
570 callout (which could be using a different interface). This has been
571 changed to use the value of the helo_data option from the smtp transport
572 instead - this is what is used when a message is actually being sent. If
573 there is no remote transport (possible with a router that sets up host
574 addresses), $smtp_active_hostname is used.
576 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
577 tweaks were necessary in order to get it to work (see also 21 below):
578 (a) The code assumed that strncpy() returns a negative number on buffer
579 overflow, which isn't the case. Replaced with Exim's string_format()
581 (b) There were several signed/unsigned issues. I just did the minimum
582 hacking in of casts. There is scope for a larger refactoring.
583 (c) The code used strcasecmp() which is not a standard C function.
584 Replaced with Exim's strcmpic() function.
585 (d) The code set only $1; it now sets $auth1 as well.
586 (e) A simple test gave the error "authentication client didn't specify
587 service in request". It would seem that Dovecot has changed its
588 interface. Fortunately there's a specification; I followed it and
589 changed what the client sends and it appears to be working now.
591 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
594 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
595 address A is aliased to B and C, where B exists and C does not. Without
596 -v the output is "A verified" because verification stops after a
597 successful redirection if more than one address is generated. However,
598 with -v the child addresses are also verified. Exim was outputting "A
599 failed to verify" and then showing the successful verification for C,
600 with its parentage. It now outputs "B failed to verify", showing B's
601 parentage before showing the successful verification of C.
603 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
604 look up a TXT record in a specific list after matching in a combined
607 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
608 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
609 they consult the DNS. I had assumed they would set it the way they
610 wanted; and indeed my experiments on Linux seem to show that in some
611 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
612 To be on the safe side, however, I have now made the interface to
613 host_find_byname() similar to host_find_bydns(), with an argument
614 containing the DNS resolver options. The host_find_byname() function now
615 sets these options at its start, just as host_find_bydns() does. The smtp
616 transport options dns_qualify_single and dns_search_parents are passed to
617 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
618 of host_find_byname() use the default settings of RES_DEFNAMES
619 (qualify_single) but not RES_DNSRCH (search_parents).
621 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
622 spool_read_header() do less string testing, by means of a preliminary
623 switch on the second character of optional "-foo" lines. (This is
624 overdue, caused by the large number of possibilities that now exist.
625 Originally there were few.) While I was there, I also converted the
626 str(n)cmp tests so they don't re-test the leading "-" and the first
627 character, in the hope this might squeeze out yet more improvement.
629 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
630 flag allowing group syntax was set by the header_syntax check but not
631 turned off, possible causing trouble later; (2) The flag was not being
632 set at all for the header_verify test, causing "group"-style headers to
633 be rejected. I have now set it in this case, and also caused header_
634 verify to ignore an empty address taken from a group. While doing this, I
635 came across some other cases where the code for allowing group syntax
636 while scanning a header line wasn't quite right (mostly, not resetting
637 the flag correctly in the right place). These bugs could have caused
638 trouble for malformed header lines. I hope it is now all correct.
640 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
641 with the "reply" argument non-NULL. The code, however (which originally
642 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
643 but it didn't always do it. This confused somebody who was copying the
644 code for some other use. I have removed all the tests.
646 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
647 feature that was used to support insecure browsers during the U.S. crypto
648 embargo. It requires special client support, and Exim is probably the
649 only MTA that supported it -- and would never use it because real RSA is
650 always available. This code has been removed, because it had the bad
651 effect of slowing Exim down by computing (never used) parameters for the
652 RSA_EXPORT functionality.
654 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
655 authenticator to fail if there's a tab character in the incoming data
656 (there should never be unless someone is messing about, as it's supposed
657 to be base64-encoded). Also added, on Timo's advice, the "secured" option
658 if the connection is using TLS or if the remote IP is the same as the
659 local IP, and the "valid-client-cert option" if a client certificate has
662 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
663 authenticators. This can be used for authorization after authentication
664 succeeds. (In the case of plaintext, it servers for both authentication
667 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
668 if any retry times were supplied.
670 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
671 connection, where there is no client IP address to check. In this
672 situation, the verify now always succeeds.
674 PH/25 Applied John Jetmore's -Mset patch.
676 PH/26 Added -bem to be like -Mset, but loading a message from a file.
678 PH/27 In a string expansion for a processed (not raw) header when multiple
679 headers of the same name were present, leading whitespace was being
680 removed from all of them, but trailing whitespace was being removed only
681 from the last one. Now trailing whitespace is removed from each header
682 before concatenation. Completely empty headers in a concatenation (as
685 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
686 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
688 PH/29 [Removed. This was a change that I later backed out, and forgot to
689 correct the ChangeLog entry (that I had efficiently created) before
690 committing the later change.]
692 PH/30 Exim was sometimes attempting to deliver messages that had suffered
693 address errors (4xx response to RCPT) over the same connection as other
694 messages routed to the same hosts. Such deliveries are always "forced",
695 so retry times are not inspected. This resulted in far too many retries
696 for the affected addresses. The effect occurred only when there were more
697 hosts than the hosts_max_try setting in the smtp transport when it had
698 the 4xx errors. Those hosts that it had tried were not added to the list
699 of hosts for which the message was waiting, so if all were tried, there
700 was no problem. Two fixes have been applied:
702 (i) If there are any address or message errors in an SMTP delivery, none
703 of the hosts (tried or untried) are now added to the list of hosts
704 for which the message is waiting, so the message should not be a
705 candidate for sending over the same connection that was used for a
706 successful delivery of some other message. This seems entirely
707 reasonable: after all the message is NOT "waiting for some host".
708 This is so "obvious" that I'm not sure why it wasn't done
709 previously. Hope I haven't missed anything, but it can't do any
710 harm, as the worst effect is to miss an optimization.
712 (ii) If, despite (i), such a delivery is accidentally attempted, the
713 routing retry time is respected, so at least it doesn't keep
714 hammering the server.
716 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
717 in ${readsocket because some servers need this prod.
719 PH/32 Added some extra debug output when updating a wait-xxx database.
721 PH/33 The hint "could be header name not terminated by colon", which has been
722 given for certain expansion errors for a long time, was not being given
723 for the ${if def:h_colon_omitted{... case.
725 PH/34 The spec says: "With one important exception, whenever a domain list is
726 being scanned, $domain contains the subject domain." There was at least
727 one case where this was not true.
729 PH/35 The error "getsockname() failed: connection reset by peer" was being
730 written to the panic log as well as the main log, but it isn't really
731 panic-worthy as it just means the connection died rather early on. I have
732 removed the panic log writing for the ECONNRESET error when getsockname()
735 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
736 runs only) independently of the message's sender address. This meant
737 that, if the 4xx error was in fact related to the sender, a different
738 message to the same recipient with a different sender could confuse
739 things. In particualar, this can happen when sending to a greylisting
740 server, but other circumstances could also provoke similar problems.
741 I have changed the default so that the retry time for these errors is now
742 based a combination of the sender and recipient addresses. This change
743 can be overridden by setting address_retry_include_sender=false in the
746 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
747 remote server are returned as part of bounce messages. This was not
748 happening for LMTP over a pipe (the lmtp transport), but now it is the
749 same for both kinds of LMTP.
751 PH/38 Despite being documented as not happening, Exim was rewriting addresses
752 in header lines that were in fact CNAMEs. This is no longer the case.
754 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
755 and queue runs started by the daemon processed all messages. This has
756 been fixed so that -R and -S can now usefully be given with -q<time>.
758 PH/40 Import PCRE release 6.7 (fixes some bugs).
760 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
762 PH/42 Give an error if -q is specified more than once.
764 PH/43 Renamed the variables $interface_address and $interface_port as
765 $received_ip_address and $received_port, to make it clear that these
766 values apply to message reception, and not to the outgoing interface when
767 a message is delivered. (The old names remain recognized, of course.)
769 PH/44 There was no timeout on the connect() call when using a Unix domain
770 socket in the ${readsocket expansion. There now is.
772 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
773 be meaningful with "accept".
775 SC/01 Eximstats V1.43
776 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
778 SC/02 Eximstats V1.44
779 Use a glob alias rather than an array ref in the generated
780 parser. This improves both readability and performance.
782 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
783 Collect SpamAssassin and rejection statistics.
784 Don't display local sender or destination tables unless
785 there is data to show.
786 Added average volumes into the top table text output.
788 SC/04 Eximstats V1.46
789 Collect data on the number of addresses (recipients)
790 as well as the number of messages.
792 SC/05 Eximstats V1.47
793 Added 'Message too big' to the list of mail rejection
794 reasons (thanks to Marco Gaiarin).
796 SC/06 Eximstats V1.48
797 Mainlog lines which have GMT offsets and are too short to
798 have a flag are now skipped.
800 SC/07 Eximstats V1.49 (Alain Williams)
801 Added the -emptyok flag.
803 SC/08 Eximstats V1.50
804 Fixes for obtaining the IP address from reject messages.
806 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
807 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
808 whitesspace changes from 4.64-PH/27
810 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
813 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
814 are found, allow negative numbers in numeric criteria)
816 JJ/06 exipick.20061117.2, added new $message_body_missing variable
818 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
819 to match changes made in 4.64-PH/43
821 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
823 PH/47 Put in an explicit test for a DNS lookup of an address record where the
824 "domain" is actually an IP address, and force a failure. This locks out
825 those revolvers/nameservers that support "A-for-A" lookups, in
826 contravention of the specifications.
828 PH/48 When a host name was looked up from an IP address, and the subsequent
829 forward lookup of the name timed out, the host name was left in
830 $sender_host_name, contrary to the specification.
832 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
833 restricted to single-key lookups, Exim was not diagnosing an error if
834 * or *@ was used with a query-style lookup.
836 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
838 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
839 long ago, but noone interested enough thought of it. Let's just say that
840 the "1.1" means that there are some new functions that weren't there at
841 some point in the past.
843 PH/51 Error processing for expansion failure of helo_data from an smtp
844 transport during callout processing was broken.
846 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
847 tested/used via the -bh/-bhc/-bs options.
849 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
850 bug, fixed in subsequent PCRE releases).
852 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
853 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
855 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
861 SC/01 Use a glob alias rather than an array ref in eximstats generated
862 parser. This improves both readability and performance.
864 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
865 Don't display local sender or destination tables in eximstats unless
866 there is data to show.
867 Added average volumes into the eximstats top table text output.
869 SC/03 Collect data on the number of addresses (recipients) as well
870 as the number of messages in eximstats.
872 TF/01 Correct an error in the documentation for the redirect router. Exim
873 does (usually) call initgroups() when daemonizing.
875 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
876 with consistent privilege compared to when running as a daemon.
878 TF/03 Note in the spec that $authenticated_id is not set for local
879 submissions from trusted users.
881 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
882 Thanks to Dean Brooks <dean@iglou.com> for the patch.
884 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
885 by adding some example configuration directives to the default
886 configuration file. A little bit of work is required to uncomment the
887 directives and define how usernames and passwords are checked, but
888 there is now a framework to start from.
890 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
891 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
892 without this. I don't know how relevant this is to other LDAP libraries.
894 PH/02 Add the verb name to the "unknown ACL verb" error.
896 PH/03 Magnus Holmgren's patch for filter_prepend_home.
898 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
900 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
901 directory not expanded when it should be if an expanded home directory
902 was set for the address (which is overridden by the transport).
904 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
907 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
908 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
909 because it is too late at that time, and has no effect.
911 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
912 security issue with \' (bugzilla #107). I could not use the
913 PQescapeStringConn() function, because it needs a PGconn value as one of
916 PH/08 When testing addresses using -bt, indicate those final addresses that
917 are duplicates that would not cause an additional delivery. At least one
918 person was confused, thinking that -bt output corresponded to deliveries.
919 (Suppressing duplicates isn't a good idea as you lose the information
920 about possibly different redirections that led to the duplicates.)
922 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
923 systems where poll() doesn't work, in particular OS X.
925 PH/10 Added more information to debugging output for retry time not reached.
927 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
928 operations in malware.c.
930 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
933 PH/13 If write_rejectlog was set false when logging was sent to syslog with
934 syslog_duplication set false, log lines that would normally be written
935 both the the main log and to the reject log were not written to syslog at
938 PH/14 In the default configuration, change the use of "message" in ACL warn
939 statements to "add_header".
941 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
942 not followed by a command (e.g. "seen endif").
944 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
945 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
948 PH/17 Added extra conditions to the default value of delay_warning_condition
952 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
953 { match{$h_precedence:}{(?i)bulk|list|junk} } \
954 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
957 The Auto-Submitted: and various List- headers are standardised, whereas I
958 don't think Precedence: ever was.
960 PH/18 Refactored debugging code in route_finduser() to show more information,
961 in particular, the error code if getpwnam() issues one.
963 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
964 This is apparently needed in addition to the PH/07 change above to avoid
965 any possible encoding problems.
967 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
968 but not after initializing Perl.
970 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
971 output them only if debugging. By default they are written stderr,
972 apparently, which is not desirable.
974 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
977 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
980 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
982 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
983 authenticated or an ident call has been made. Suppress the default
984 values for $authenticated_id and $authenticated_sender (but permit -oMai
985 and -oMas) when testing with -bh.
987 PH/24 Re-jigged the order of the tests in the default configuration so that the
988 tests for valid domains and recipients precede the DNS black list and CSA
989 tests, on the grounds that those ones are more expensive.
991 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
992 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
993 This bug exists in every version of Exim that I still have, right back to
996 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
997 However, an attempt to turn on case-sensitivity in a regex key by
998 including (?-i) didn't work because the subject string was already
999 lowercased, and the effects were non-intuitive. It turns out that a
1000 one-line patch can be used to allow (?-i) to work as expected.
1006 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1007 other effects) broke the use of negated acl sub-conditions.
1009 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1012 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1013 "Deny" causes Exim to reject the incoming connection with a 554 error.
1014 Unfortunately, if there is a major crisis, such as a disk failure,
1015 tcp-wrappers gives "deny", whereas what one would like would be some
1016 kind of temporary error. A kludge has been added to help with this.
1017 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1018 554 error is used if errno is still zero or contains ENOENT (which occurs
1019 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1022 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1024 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1025 errors. Otherwise a message that provokes a temporary error (when other
1026 messages do not) can cause a whole host to time out.
1028 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1029 addresses were routed directly to files or pipes from a redirect router.
1030 File deliveries just didn't batch; pipe deliveries might have suffered
1033 PH/06 A failure to get a lock for a hints database would erroneously always say
1034 "Failed to get write lock", even when it was really a read lock.
1036 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1037 of 0600. This has been changed to use the value of the lockfile_mode
1038 option (which defaults to 0600).
1040 PH/08 Applied small patch from the Sieve maintainer.
1042 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1043 folder from quota calculations, a direct delivery into this folder messed
1044 up the contents of the maildirsize file. This was because the regex was
1045 used only to exclude .Trash (or whatever) when the size of the mailbox
1046 was calculated. There was no check that a delivery was happening into an
1047 excluded directory. This bug has been fixed by ignoring all quota
1048 processing for deliveries into excluded directories.
1050 PH/10 Added the maildirfolder_create_regex option to appendfile.
1056 PH/01 The code for finding all the local interface addresses on a FreeBSD
1057 system running IPv6 was broken. This may well have applied to all BSD
1058 systems, as well as to others that have similar system calls. The broken
1059 code found IPv4 interfaces correctly, but gave incorrect values for the
1060 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1061 that it would not match correctly against @[] and not recognize the IPv6
1064 PH/02 The ipliteral router was not recognizing addresses of the form user@
1065 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1067 PH/03 Added disable_ipv6.
1069 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1070 decoded form, because it is most often used to construct To: headers
1071 lines in autoreplies, and the decoded form may well be syntactically
1072 invalid. However, $reply_address has leading white space removed, and all
1073 newlines turned into spaces so that the autoreply transport does not
1076 PH/05 If group was specified without a user on a router, and no group or user
1077 was specified on a transport, the group from the router was ignored.
1079 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1080 for visible compile-time settings that can be used to change these
1081 numbers, for those that want even more. Backwards compatibility with old
1082 spool files has been maintained. However, going back to a previous Exim
1083 release will lost any variables that are in spool files.
1085 PH/07 Two small changes when running in the test harness: increase delay when
1086 passing a TCP/IP connection to a new process, in case the original
1087 process has to generate a bounce, and remove special handling of
1088 127.0.0.2 (sic), which is no longer necessary.
1090 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1091 be the same on different OS.
1093 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1096 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1097 whether --show-vars was specified or not
1099 JJ/02 exipick: Added support for new ACL variable spool format introduced
1102 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1103 syntactically invalid From: or Reply-to: line, and a filter used this to
1104 generate an autoreply, and therefore failed to obtain an address for the
1105 autoreply, Exim could try to deliver to a non-existent relative file
1106 name, causing unrelated and misleading errors. What now happens is that
1107 it logs this as a hard delivery error, but does not attempt to create a
1110 PH/11 The exinext utility has a -C option for testing purposes, but although
1111 the given file was scanned by exinext itself; it wasn't being passed on
1112 when Exim was called.
1114 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1115 an end-of-file indication when reading a command response.
1117 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1118 compiled. In many other places in Exim, IPv6 addresses are always
1119 recognized, so I have changed this. It also means that IPv4 domain
1120 literals of the form [IPV4:n.n.n.n] are now always recognized.
1122 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1123 used if the router is not running as root, for example, when verifying at
1124 ACL time, or when using -bh. The debugging output from this situation was
1125 non-existent - all you got was a failure to exec. I have made two
1128 (a) Failures to set uid/gid, the current directory, or a process leader
1129 in a subprocess such as that created by queryprogram now generate
1130 suitable debugging ouput when -d is set.
1132 (b) The queryprogram router detects when it is not running as root,
1133 outputs suitable debugging information if -d is set, and then runs
1134 the subprocess without attempting to change uid/gid.
1136 PH/15 Minor change to Makefile for building test_host (undocumented testing
1139 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1140 additional section of a DNS packet that returns MX or SRV records.
1141 Instead, it always explicitly searches for A/AAAA records. This avoids
1142 major problems that occur when a DNS server includes only records of one
1143 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1144 fixed another bug: if SRV records were looked up and the corresponding
1145 address records were *not* found in the additional section, the port
1146 values from the SRV records were lost.
1148 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1149 using the correct key (the original address) when searching the retry
1150 rules in order to find which one to use for generating the retry hint.
1152 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1153 adding the default one. Similarly, if it contains a Reply-To: header, the
1154 errors_reply_to option, if set, is not used.
1156 PH/19 When calculating a retry time, Exim used to measure the "time since
1157 failure" by looking at the "first failed" field in the retry record. Now
1158 it does not use this if it is later than than the arrival time of the
1159 message. Instead it uses the arrival time. This makes for better
1160 behaviour in cases where some deliveries succeed, thus re-setting the
1161 "first failed" field. An example is a quota failure for a huge message
1162 when small messages continue to be delivered. Without this change, the
1163 "time since failure" will always be short, possible causing more frequent
1164 delivery attempts for the huge message than are intended.
1165 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1167 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1168 $1, $2, $3) because the numerical variables can be reset during some
1169 expansion items (e.g. "match"), thereby losing the authentication data.
1171 PH/21 Make -bV show the size of off_t variables so that the test suite can
1172 decide whether to run tests for quotas > 2G.
1174 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1175 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1176 filecount value is greater than 2G or if a quota value is greater than 2G
1177 on a system where the size of off_t is not greater than 4, a panic error
1180 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1181 never match. The debug and -bh output now contains an explicit error
1182 message indicating a malformed IPv4 address or mask.
1184 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1185 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1186 PH/23 above applies.
1188 PH/25 Do not write to syslog when running in the test harness. The only
1189 occasion when this arises is a failure to open the main or panic logs
1190 (for which there is an explicit test).
1192 PH/26 Added the /no_tell option to "control=freeze".
1194 PH/27 If a host name lookup failed very early in a connection, for example, if
1195 the IP address matched host_lookup and the reverse lookup yielded a name
1196 that did not have a forward lookup, an error message of the form "no IP
1197 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1198 could be logged. Now it outputs the IP address instead of "NULL".
1200 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1201 allows the sender and the authenticated sender to be set when
1202 submitting a message from within Exim. Since child_open_exim() is
1203 documented for local_scan(), the new function should be too.
1205 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1206 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1207 results in an empty string is now treated as unset.
1209 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1211 PH/31 Added sender_verify_fail logging option.
1213 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1214 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1215 tidied the source and removed it altogether.
1217 PH/33 When a queue run was abandoned because the load average was too high, a
1218 log line was always written; now it is written only if the queue_run log
1219 selector is set. In addition, the log line for abandonment now contains
1220 information about the queue run such as the pid. This is always present
1221 in "start" and "stop" lines but was omitted from the "abandon" line.
1223 PH/34 Omit spaces between a header name and the colon in the error message that
1224 is given when verify = headers_syntax fails (if there are lots of them,
1225 the message gets confusing).
1227 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1228 names, as there are now some PTR records that contain slashes. This check
1229 is only to protect against broken name servers that fall over on strange
1230 characters, so the fact that it applies to all lookups doesn't matter.
1232 PH/36 Now that the new test suite is complete, we can remove some of the
1233 special code in Exim that was needed for the old test suite. For example,
1234 sorting DNS records because real resolvers return them in an arbitrary
1235 order. The new test suite's fake resolver always returns records in the
1238 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1239 bounces) except when queue_only is set, to avoid logging races between
1240 the different processes.
1242 PH/38 Panic-die if .include specifies a non-absolute path.
1244 PH/39 A tweak to the "H" retry rule from its user.
1246 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1247 a label. They prevented compilation on older perls.
1249 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1250 a warning to be raised on newish perls.
1252 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1253 on queue. Changes to match documented behaviour of showing count of
1254 messages matching specified criteria.
1256 PH/40 Changed the default ident timeout from 30s to 5s.
1258 PH/41 Added support for the use of login_cap features, on those BSD systems
1259 that have them, for controlling the resources used by pipe deliveries.
1261 PH/42 The content-scanning code uses fopen() to create files in which to put
1262 message data. Previously it was not paying any attention to the mode of
1263 the files. Exim runs with umask(0) because the rest of the code creates
1264 files with open(), and sets the required mode explicitly. Thus, these
1265 files were ending up world-writeable. This was not a big issue, because,
1266 being within the spool directory, they were not world-accessible. I have
1267 created a function called modefopen, which takes an additional mode
1268 argument. It sets umask(777), creates the file, chmods it to the required
1269 mode, then resets the umask. All the relevant calls to fopen() in the
1270 content scanning code have been changed to use this function.
1272 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1273 to 24 hours. This avoids potential overflow problems when processing G
1274 and H retry rules. I suspect nobody ever tinkers with this value.
1276 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1278 PH/45 When the plaintext authenticator is running as a client, the server's
1279 challenges are checked to ensure they are valid base64 strings. By
1280 default, the authentication attempt is cancelled if an invalid string is
1281 received. Setting client_ignore_invalid_base64 true ignores these errors.
1282 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1283 they are received. Thus, the responses can be made to depend on the
1284 challenges. If an invalid string is ignored, an empty string is placed in
1287 PH/46 Messages that are created by the autoreply transport now contains a
1288 References: header, in accordance with RFCs 2822 and 3834.
1290 PH/47 Added authenticated_sender_force to the smtp transport.
1292 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1294 PH/49 Installed latest patch from the Sieve maintainer.
1296 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1297 was also set, the appendfile transport was unnecessarily scanning a
1298 directory of message files (e.g. for maildir delivery) to find the count
1299 of files (along with the size), even though it did not need this
1300 information. It now does the scan only if it needs to find either the
1301 size of the count of files.
1303 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1305 PH/52 Two bugs concerned with error handling when the smtp transport is
1308 (i) Exim was not creating retry information for temporary errors given
1309 for individual recipients after the DATA command when the smtp transport
1310 was used in LMTP mode. This meant that they could be retried too
1311 frequently, and not timed out correctly.
1313 (ii) Exim was setting the flag that allows error details to be returned
1314 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1315 recipients that were returned after the DATA command.
1317 PH/53 This is related to PH/52, but is more general: for any failing address,
1318 when detailed error information was permitted to be returned to the
1319 sender, but the error was temporary, then after the final timeout, only
1320 "retry timeout exceeded" was returned. Now it returns the full error as
1321 well as "retry timeout exceeded".
1323 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1324 do this, and (what is worse) MTAs that accept it.
1326 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1327 will now be deprecated.
1329 PH/56 New os.c-cygwin from the Cygwin maintainer.
1331 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1332 formats (previously only available in exim formats via -bpr, -bpru,
1333 and -bpra. Now also available in native and exiqgrep formats)
1335 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1336 with very large, slow to parse queues
1338 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1340 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1342 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1343 responses to authentication challenges, though it was showing the
1344 challenges; (ii) I've removed the CR characters from the debug output for
1347 PH/58 Allow for the insertion of a newline as well as a space when a string
1348 is turned into more than one encoded-word during RFC 2047 encoding. The
1349 Sieve code now uses this.
1351 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1352 data_4xx, lost_connection, tls_required.
1354 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1355 message was being sent as an SMTP response.
1357 PH/61 Add -l and -k options to exicyclog.
1359 PH/62 When verifying, if an address was redirected to one new address, so that
1360 verification continued, and the new address failed or deferred after
1361 having set something in $address_data, the value of $address_data was not
1362 passed back to the ACL. This was different to the case when no
1363 redirection occurred. The value is now passed back in both cases.
1365 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1366 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1367 use login_cap.h, so on its own it isn't the distinguishing feature. The
1368 new name refers directly to the setclassresources() function.
1370 PH/65 Added configuration files for NetBSD3.
1372 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1374 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1375 is preferred over IPv4.
1377 PH/68 The bounce_return_message and bounce_return_body options were not being
1378 honoured for bounces generated during the reception of non-SMTP messages.
1379 In particular, this applied to messages rejected by the ACL. This bug has
1380 been fixed. However, if bounce_return_message is true and bounce_return_
1381 body is false, the headers that are returned for a non-SMTP message
1382 include only those that have been read before the error was detected.
1383 (In the case of an ACL rejection, they have all been read.)
1385 PH/69 The HTML version of the specification is now built in a directory called
1386 spec_html instead of spec.html, because the latter looks like a path with
1387 a MIME-type, and this confuses some software.
1389 PH/70 Catch two compiler warnings in sieve.c.
1391 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1392 function verify_get_ident() calls ip_connect() to connect a socket, but
1393 if the "connect()" function timed out, ip_connect() used to close the
1394 socket. However, verify_get_ident() also closes the socket later, and in
1395 between Exim writes to the log, which may get opened at this point. When
1396 the socket was closed in ip_connect(), the log could get the same file
1397 descriptor number as the socket. This naturally causes chaos. The fix is
1398 not to close the socket in ip_connect(); the socket should be closed by
1399 the function that creates it. There was only one place in the code where
1400 this was missing, in the iplookup router, which I don't think anybody now
1401 uses, but I've fixed it anyway.
1403 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1404 well as to direct DNS lookups. Otherwise the handling of names in host
1405 lists is inconsistent and therefore confusing.
1411 PH/01 Two changes to the default runtime configuration:
1413 (1) Move the checks for relay_from_hosts and authenticated clients from
1414 after to before the (commented out) DNS black list checks.
1416 (2) Add control=submission to the relay_from_hosts and authenticated
1417 clients checks, on the grounds that messages accepted by these
1418 statements are most likely to be submissions.
1420 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1422 (1) Generate an error if the third argument for the ${prvs expansion is
1425 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1428 (3) Reset the variables that are obtained from the first argument of
1429 ${prvscheck and used in the second argument before leaving the code,
1430 because their memory is reclaimed, so using them afterwards may do
1433 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1434 one (it's much easier than Tom thought :-).
1436 (5) Because of (4), we can now allow for the use of $prvscheck_result
1437 inside the third argument.
1439 PH/03 For some reason, the default setting of PATH when running a command from
1440 a pipe transport was just "/usr/bin". I have changed it to
1443 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1444 anything to be listed in the output from -bV.
1446 PH/05 When a filter generated an autoreply, the entire To: header line was
1447 quoted in the delivery log line, like this:
1449 => >A.N.Other <ano@some.domain> <original@ddress> ...
1451 This has been changed so that it extracts the operative address. There
1452 may be more than one such address. If so, they are comma-separated, like
1455 => >ano@some.domain,ona@other.domain <original@ddress> ...
1457 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1458 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1459 not being looked up in the rDNS to get a host name, Exim was showing the
1460 IP address twice in Received: lines, even though the IP addresses were
1461 identical. For example:
1463 Received: from [1.2.3.4] (helo=[1.2.3.4])
1465 However, if the real host name was known, it was omitting the HELO data
1466 if it matched the actual IP address. This has been tidied up so that it
1467 doesn't show the same IP address twice.
1469 PH/07 When both +timestamp and +memory debugging was on, the value given by
1470 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1471 called by the debug printing, thereby overwriting the timestamp buffer.
1472 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1474 PH/08 When the original message was included in an autoreply transport, it
1475 always said "this is a copy of the message, including all the headers",
1476 even if body_only or headers_only was set. It now gives an appropriate
1479 PH/09 Applied a patch from the Sieve maintainer which:
1481 o fixes some comments
1482 o adds the (disabled) notify extension core
1483 o adds some debug output for the result of if/elsif tests
1484 o points to the current vacation draft in the documentation
1485 and documents the missing references header update
1489 o fixes a bug in processing the envelope test (when testing
1490 multiple envelope elements, the last element determinted the
1493 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1494 Electronic Mail") by including:
1496 Auto-submitted: auto-generated
1498 in the messages that it generates (bounce messages and others, such as
1499 warnings). In the case of bounce messages for non-SMTP mesages, there was
1500 also a typo: it was using "Auto_submitted" (underscore instead of
1501 hyphen). Since every message generated by Exim is necessarily in response
1502 to another message, thes have all been changed to:
1504 Auto-Submitted: auto-replied
1506 in accordance with these statements in the RFC:
1508 The auto-replied keyword:
1510 - SHOULD be used on messages sent in direct response to another
1511 message by an automatic process,
1513 - MUST NOT be used on manually-generated messages,
1515 - MAY be used on Delivery Status Notifications (DSNs) and Message
1516 Disposition Notifications (MDNs),
1518 - MUST NOT be used on messages generated by automatic or periodic
1519 processes, except for messages which are automatic responses to
1522 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1523 to the default Received: header definition.
1525 PH/12 Added log selector acl_warn_skipped (default on).
1527 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1528 variables because (a) they are in the wrong storage pool and (b) even if
1529 they were copied, it wouldn't work properly because of the caching.
1531 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1532 checking when decoding. Apparently there are clients that generate
1533 overlong encoded strings. Why am I not surprised?
1535 PH/15 If the first argument of "${if match_address" was not empty, but did not
1536 contain an "@" character, Exim crashed. Now it writes a panic log message
1537 and treats the condition as false.
1539 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1541 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1542 extension "envelope-auth". The code is finished and in agreement with
1543 other implementations, but there is no documentation so far and in fact,
1544 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1545 not changing the active code.
1547 Print executed "if" and "elsif" statements when debugging is used. This
1548 helps a great deal to understand what a filter does.
1550 Document more things not specified clearly in RFC3028. I had all this
1551 sorted out, when out of a sudden new issues came to my mind. Oops."
1553 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1556 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1557 canonical form (as documented). However, after a host name lookup from
1558 the IP address, check_host() was doing a simple string comparison with
1559 addresses acquired from the DNS when checking that the found name did
1560 have the original IP as one of its addresses. Since any found IPv6
1561 addresses are likely to be in abbreviated form, the comparison could
1562 fail. Luckily, there already exists a function for doing the comparison
1563 by converting both addresses to binary, so now that is used instead of
1564 the text comparison.
1566 PH/20 There was another similar case to PH/19, when a complete host name was
1567 given in a host list; looking up its IP address could give an abbreviated
1568 form, whereas the current host's name might or might not be abbreviated.
1569 The same fix has been applied.
1575 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1576 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1579 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1580 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1582 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1584 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1585 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1586 Also, for "Auto-Submitted", treat anything other than "no" as
1587 non-personal, in accordance with RFC 3834. (Previously it treated
1588 anything starting "auto-" as non-personal.)
1590 TF/01 The control=submission/name=... option had a problem with syntax
1591 errors if the name included a slash character. The /name= option
1592 now slurps the rest of the string, so it can include any characters
1593 but it must come last in the list of options (after /sender_retain
1596 PH/05 Some modifications to the interface to the fake nameserver for the new
1604 TK/01 Added the "success_on_redirect" address verification option. See
1605 NewStuff for rationale and an example.
1607 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1609 PH/02 Patch to exigrep to allow it to work on syslog lines.
1611 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1612 fread() to skip over the body file's header line, because in Cygwin the
1613 header line is locked and is inaccessible.
1615 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1616 co-exist for some time) to make it clear that it is the Exim ID that is
1617 referenced, not the Message-ID: header line.
1619 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1620 string_format() function, because snprintf() does not exist on all
1623 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1624 ${stat: expansion item.
1626 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1627 protocol synchronization error", to keep the pedants happy.
1629 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1630 well as for IRIX systems, when gcc is being used. See the host.c source
1633 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1635 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1638 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1639 transport and to the smtp transport in LMTP mode.
1641 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1643 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1644 run a filter in a subprocess. This could lead to confusion in subsequent
1645 lookups in the parent process. There should also be a search_tidyup() at
1646 the end of the subprocess.
1648 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1649 only if the host matched helo_try_verify_hosts, which caused the
1650 verification to occur when the EHLO/HELO command was issued. The ACL just
1651 tested the remembered result. Now, if a previous verification attempt has
1652 not happened, "verify = helo" does it there and then.
1654 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1656 TK/03 Fix log output including CR from clamd.
1658 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1659 exist provoked a memory error which could cause a segfault.
1661 PH/15 Installed PCRE 6.2
1663 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1665 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1666 of the problem. Specifically, suggested +O2 rather than +O1 for the
1669 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1671 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1672 record, the port was not correctly logged when the outgoing_port log
1673 selector was set (it logged the transort's default port).
1675 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1676 fallback_hosts, and "hosts" in the smtp transport.
1678 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1679 host errors such as "Connection refused".
1681 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1682 authentication with radiusclient 0.4.9:
1684 - Error returned from rc_read_config was caught wrongly
1685 - Username/password not passed on to radius server due to wrong length.
1687 The presumption is that some radiusclient API changes for 4.51/PH/17
1688 were not taken care of correctly. The code is still untested by me (my
1689 Linux distribution still has 0.3.2 of radiusclient), but it was
1690 contributed by a Radius user.
1692 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1693 expanding the "port" option of the smtp transport.
1695 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1696 while reading a MIME header. Thanks to Tom Hughes for a patch.
1698 PH/24 Include config.h inside local_scan.h so that configuration settings are
1701 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1702 that in an ACL for RCPT (for example), you can examine exactly what was
1705 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1706 commands, but it was not correctly comparing the address with the actual
1707 client host address. Thus, it would show the EHLO address in Received:
1708 header lines when this was not necessary.
1710 PH/27 Added the % operator to ${eval:}.
1712 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1713 it should be ignoring failures (because with -C, for example, it has lost
1714 privilege). It wasn't ignoring creation failures other than "already
1717 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1720 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1721 because an input line was too long, either on its own, or by virtue of
1722 too many continuations, the temporary file was not being removed, and the
1723 return code was incorrect.
1725 PH/31 Missing "BOOL" in function definition in filtertest.c.
1727 PH/32 Applied Sieve patches from the maintainer.
1729 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1731 PH/33 Added "verify = not_blind".
1733 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1734 Local/Makefile (with some defaults set). These are used in built scripts
1735 such as exicyclog, but they have never been used in the exim_install
1736 script (though there are many overriding facilities there). I have
1737 arranged that the exim_install script now takes note of these two
1740 PH/35 Installed configuration files for Dragonfly.
1742 PH/36 When a locally submitted message by a trusted user did not contain a
1743 From: header, and the sender address was obtained from -f or from an SMTP
1744 MAIL command, and the trusted user did not use -F to supply a sender
1745 name, $originator_name was incorrectly used when constructing a From:
1746 header. Furthermore, $originator_name was used for submission mode
1747 messages from external hosts without From: headers in a similar way,
1748 which is clearly wrong.
1750 PH/37 Added control=suppress_local_fixups.
1752 PH/38 When log_selector = +received_sender was set, and the addition of the
1753 sender made the log line's construction buffer exactly full, or one byte
1754 less than full, an overflow happened when the terminating "\n" was
1757 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1758 when the result of a list match is failure because a DNS lookup failed.
1760 PH/40 RM_COMMAND is now used in the building process.
1762 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1763 the "build-* directories that it finds.
1765 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1766 address in a domain literal was a prefix of an interface address.
1768 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1769 when verifying a sender address, unless rewrite_headers is false.
1771 PH/44 Wrote a long comment about why errors_to addresses are verified as
1772 recipients, not senders.
1774 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1775 the ratelimit ACL was added.
1777 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1779 PH/46 Added extra information about PostgreSQL errors to the error string.
1781 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1782 avoiding the need to install special zones in a real server. This is
1783 backwards compatible; if it can't find the fake resolver, it drops back.
1784 Thus, both old and new test suites can be run.
1786 TF/02 Added util/ratelimit.pl
1788 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1789 clock is set back in time.
1791 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1792 Candler <B.Candler@pobox.com>.
1794 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1795 OK for addresses that are the result of redirections.
1797 PH/48 A number of further additions for the benefit of the new test suite,
1798 including a fake gethostbyname() that interfaces to the fake DNS resolver
1801 TF/06 The fix for widen_domains has also been applied to qualify_single and
1802 search_parents which are the other dnslookup options that can cause
1805 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1808 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1810 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1811 service" if it encountered an MX record with an empty target hostname.
1812 The message is now "an MX or SRV record indicated no SMTP service".
1814 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1815 if the DNS of the sending site is misconfigured. This is quite a
1816 common situation. This change restores the behaviour of treating a
1817 helo verification defer as a failure.
1819 PH/51 If self=fail was set on a router, the bounce message did not include the
1820 actual error message.
1826 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1828 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1829 command itself ended in error, the underlying message about the transport
1830 filter timeout was being overwritten with the pipe command error. Now the
1831 underlying error message should be appended to the second error message.
1833 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1835 PH/02 Reduce the amount of output that "make" produces by default. Full output
1836 can still be requested.
1838 PH/03 The warning log line about a condition test deferring for a "warn" verb
1839 was being output only once per connection, rather than after each
1840 occurrence (because it was using the same function as for successful
1841 "warn" verbs). This seems wrong, so I have changed it.
1843 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1844 it should not have, which might have caused a crash in the right
1845 circumstances, but probably never did.
1847 PH/04 Installed a modified version of Tony Finch's patch to make submission
1848 mode fix the return path as well as the Sender: header line, and to
1849 add a /name= option so that you can make the user's friendly name appear
1852 TF/03 Added the control = fakedefer ACL modifier.
1854 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1855 Mark Lowes for thorough testing.
1857 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1859 TK/03 Merged latest SRS patch from Miles Wilton.
1861 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1862 with the definition in sysexits.h (which is #included earlier).
1863 Fortunately, Exim does not actually use EX_OK. The code used to try to
1864 preserve the sysexits.h value, by assumimg that macro definitions were
1865 scanned for macro replacements. I have been disabused of this notion,
1866 so now the code just undefines EX_OK before #including unistd.h.
1868 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1869 in the smtp transport. When a block could not be written in a single
1870 write() function, the timeout was being re-applied to each part-write.
1871 This seems wrong - if the receiver was accepting one byte at a time it
1872 would take for ever. The timeout is now adjusted when this happens. It
1873 doesn't have to be particularly precise.
1875 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1876 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1878 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1879 without a domain if the check to <postmaster@domain> fails.
1881 SC/01 Eximstats: added -xls and the ability to specify output files
1882 (patch written by Frank Heydlauf).
1884 SC/02 Eximstats: use FileHandles for outputing results.
1886 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1888 SC/04 Eximstats: fixed display of large numbers with -nvr option
1890 SC/05 Eximstats: fixed merging of reports with empty tables.
1892 SC/06 Eximstats: added the -include_original_destination flag
1894 SC/07 Eximstats: removed tabs and trailing whitespace.
1896 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1898 TK/06 MBOX spool code: Add real "From " MBOX separator line
1899 so the .eml file is really in mbox format (even though
1900 most programs do not really care). Patch from Alex Miller.
1902 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1903 The latter is generated from $received_to and is only set if the
1904 message has one envelope recipient. SA can use these headers,
1905 obviously out-of-the-box. Patch from Alex Miller.
1907 PH/08 The ${def test on a variable was returning false if the variable's
1908 value was "0", contrary to what the specification has always said!
1909 The result should be true unless the variable is empty.
1911 PH/09 The syntax error of a character other than { following "${if
1912 def:variable_name" (after optional whitespace) was not being diagnosed.
1913 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1914 accidental colon was present, for example, could give incorrect results.
1916 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1917 result is used (not including appendfile, where other changes are about
1920 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1921 This involved changing a lot of size variables from int to off_t. It
1922 should work with maildirs and everything.
1924 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1925 spamd dying while we are connected to it.
1927 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1930 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1931 function for BDB 4.3.
1933 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1935 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1936 the sources for PCRE have been re-arranged and more files are now
1939 PH/15 The code I had for printing potentially long long variables in PH/11
1940 above was not the best (it lost precision). The length of off_t variables
1941 is now inspected at build time, and an appropriate printing format (%ld
1942 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1943 to be "long long int" or "long int". This is needed for the internal
1944 formatting function string_vformat().
1946 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1947 the configuration file to be ":syslog", then the script "guesses" where
1948 the logs files are, rather than using the compiled in default. In our
1949 case the guess is not the same as the compiled default, so the script
1950 suddenly stopped working when I started to use syslog. The patch checks
1951 to see if log_file_path is "". If so, it attempts to read it from exim
1952 with no configuration file to get the compiled in version, before it
1953 falls back to the previous guessing code."
1955 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1956 implementing BATV in an Exim configuration. See NewStuff for the gory
1959 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1960 Makefile that are specific to HP-UX.
1962 PH/18 If the "use_postmaster" option was set for a recipient callout together
1963 with the "random" option, the postmaster address was used as the MAIL
1964 FROM address for the random test, but not for the subsequent recipient
1965 test. It is now used for both.
1967 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1968 patch removes a few documentation additions to RFC 3028, because the
1969 latest draft now contains them. It adds the new en;ascii-case comparator
1970 and a new error check for 8bit text in MIME parts. Comparator and
1971 require names are now matched exactly. I enabled the subaddress
1972 extension, but it is not well tested yet (read: it works for me)."
1974 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1975 rework some of the code of TK/09 above to avoid the hardwired use of
1976 "%lld" and "long long". Replaced the call to snprintf() with a call to
1979 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1980 records point to non-existent hosts", "retry timeout exceeded", and
1981 "retry time not reached for any host after a long failure period".
1983 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1984 experimental DomainKeys support:
1986 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1987 (2) On an error such as an illegally used "control", the wrong name for
1988 the control was given.
1990 These problems did NOT occur unless DomainKeys support was compiled.
1992 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1994 PH/24 Added ${if match_ip condition.
1996 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1997 databases so that it will be absolutely obvious if a crash occurs in the
1998 DB library. This is a regular occurrence (often caused by mis-matched
2001 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2002 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2003 picked up on a user's system that detects such things. There doesn't seem
2004 to be a gcc warning option for this - only an attribute that has to be
2005 put on the function's prototype. It seems that in Fedora Core 4 they have
2006 set this on a number of new functions. No doubt there will be more in due
2009 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2010 specify a transport. However, if an address that was verified by such a
2011 router was the subject of a callout, Exim crashed because it tried to
2012 read the rcpt_include_affixes from the non-existent transport. Now it
2013 just assumes that the setting of that option is false. This bug was
2014 introduced by 4.51/PH/31.
2016 PH/28 Changed -d+all to exclude +memory, because that information is very
2017 rarely of interest, but it makes the output a lot bigger. People tend to
2018 do -d+all out of habit.
2020 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2021 code in os-type was giving problems when libc.so lives in lib64, like on
2024 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2025 aren't the modern standard, and it seems that some systems' include files
2026 don't always have them. Exim was already checking for some of the newer
2027 ones like T_AAAA, and defining it itself. I've added checks for all the
2028 record types that Exim uses.
2030 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2031 not automatically generating a new one, as it is supposed to. This
2032 prevented TLS from working. If the file did exist, but contained invalid
2033 data, a new version was generated, as expected. It was only the case of a
2034 non-existent file that was broken.
2036 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2037 with a change in libdomainkeys > 0.64.
2039 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2040 from DNS. If the selector record carries the flag, it now has
2041 precedence over the domain-wide flag.
2043 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2045 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2046 the use of a port name that isn't defined in /etc/services) occurred, the
2047 message was deferred as in a normal delivery, and thus remained on the
2048 spool, instead of being failed because of the mua_wrapper setting. This
2049 is now fixed, and I tidied up some of the mua_wrapper messages at the
2052 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2053 the messages in a hash of arrays rather than using individual hashes.
2054 This is a bit cleaner and results in dramatic memory savings, albeit
2055 at a slight CPU cost.
2057 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2058 as requested by Marc Sherman.
2060 SC/10 Eximstats: added histograms for user specified patterns as requested
2063 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2065 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2066 fopen() in the content-scanning modules that did not already have it.
2072 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2073 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2075 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2077 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2079 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2080 to test Sieve filters that use "vacation".
2082 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2083 that changes the way the GnuTLS parameters are stored in the cache file.
2084 The new format can be generated externally. For backward compatibility,
2085 if the data in the cache doesn't make sense, Exim assumes it has read an
2086 old-format file, and it generates new data and writes a new file. This
2087 means that you can't go back to an older release without removing the
2090 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2091 work if there are any delivery delays because "one_time" forces the
2092 parent to be marked "delivered", so its unseen clone is never tried
2093 again. For this reason, Exim now forbids the simultaneous setting of
2096 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2097 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2098 entry if you want to know the details. The fix, however, neglected to
2099 consider the case when local delivery batching is involved. The test for
2100 "previously delivered" was not happening when checking to see if an
2101 address could be batched with a previous (undelivered) one; under
2102 certain circumstances this could lead to multiple deliveries to the same
2105 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2106 in its include files, and this causes problems building Exim.
2108 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2109 header_syntax) but Exim was just ignoring anything given after a slash.
2110 In particular, this caused confusion with an attempt to use "verify =
2111 reverse_host_lookup/defer_ok". An error is now given when options are
2112 supplied for verify items that do not have them. (Maybe reverse_host_
2113 lookup should have a defer_ok option, but that's a different point.)
2115 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2116 defined by RFC 821) to 2048, because there were problems with some AUTH
2117 commands, and RFC 1869 says the size should be increased for extended
2118 SMTP commands that take arguments.
2120 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2123 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2124 "unknown" error; now it says that the functionality isn't in the binary.
2126 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2127 an address' error message when a string expansion fails (syntax or
2128 whatever). Otherwise the password may appear in the log. Following change
2129 PH/42 below, there is no longer a chance of it appearing in a bounce
2132 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2134 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2135 of its addresses was ever tried. (Bugzilla bug #2.)
2137 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2138 the result incorrectly in the debug output. (It correctly added a newline
2139 to what was transported.)
2141 TF/01 Added $received_time.
2143 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2144 commented out examples of how to interface to a virus scanner and to
2145 SpamAssassin. Also added commented examples of av_scanner and
2146 spamd_address settings.
2148 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2149 and controls are allowed in which ACLs. There were a couple of minor
2150 errors. Some of the entries in the conditions table (which is a table of
2151 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2152 negation of where the condition IS allowed.
2154 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2156 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2157 header file does not have a version number, so I've had to invent a new
2158 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2159 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2160 radiusclient), but it was contributed by a Radius user.
2162 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2163 files or directories, for OpenSSL.
2165 PH/19 When an Exim process that is running as root has to create an Exim log
2166 file, it does so in a subprocess that runs as exim:exim so as to get the
2167 ownership right at creation (otherwise, other Exim processes might see
2168 the file with the wrong ownership). There was no test for failure of this
2169 fork() call, which would lead to the process getting stuck as it waited
2170 for a non-existent subprocess. Forks do occasionally fail when resources
2171 run out. I reviewed all the other calls to fork(); they all seem to check
2174 PH/20 When checking for unexpected SMTP input at connect time (before writing
2175 the banner), Exim was not dealing correctly with a non-positive return
2176 from the read() function. If the client had disconnected by this time,
2177 the result was a log entry for a synchronization error with an empty
2178 string after "input=" when read() returned zero. If read() returned -1
2179 (an event I could not check), uninitialized data bytes were printed.
2180 There were reports of junk text (parts of files, etc) appearing after
2183 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2185 PH/22 Added support for macro redefinition, and (re)definition in between
2186 driver and ACL definitions.
2188 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2189 forgetting to use the resulting value; it was using the unexpanded value.
2191 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2192 hadn't been configured. The fix is from Juergen Kreileder, who
2193 understands it better than I do:
2195 "Here's what I see happening with three configured cyrus_sasl
2196 authenticators configured (plain, login, cram-md5):
2198 On startup auth_cyrus_sasl_init() gets called for each of these.
2199 This means three calls to sasl_listmech() without a specified mech_list.
2200 => SASL tests which mechs of all available mechs actually work
2201 => three warnings about OTP not working
2202 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2204 With the patch, sasl_listmech() also gets called three times. But now
2205 SASL's mech_list option is set to the server_mech specified in the the
2206 authenticator. Or in other words, the answer from sasl_listmech()
2207 gets limited to just the mech you're testing for (which is different
2209 => the return list contains just 'plain' or 'login', 'cram-md5' or
2210 nothing depending on the value of ob->server_mech.
2212 I've just tested the patch: Authentication still works fine,
2213 unavailable mechs specified in the exim configuration are still
2214 caught, and the auth.log warnings about OTP are gone."
2216 PH/25 When debugging is enabled, the contents of the command line are added
2217 to the debugging output, even when log_selector=+arguments is not
2220 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2221 answer is "GNU", and only if the return is "GNU/something" is the answer
2224 PH/27 $acl_verify_message is now set immediately after the failure of a
2225 verification in an ACL, and so is available in subsequent modifiers. In
2226 particular, the message can be preserved by coding like this:
2228 warn !verify = sender
2229 set acl_m0 = $acl_verify_message
2231 Previously, $acl_verify_message was set only while expanding "message"
2232 and "log_message" when a very denied access.
2234 PH/28 Modified OS/os.c-Linux with
2236 -#ifndef OS_LOAD_AVERAGE
2237 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2239 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2240 nomenclature these days.)
2242 PH/29 Installed patch from the Sieve maintainer that adds the options
2243 sieve_useraddress and sieve_subaddress to the redirect router.
2245 PH/30 In these circumstances:
2246 . Two addresses routed to the same list of hosts;
2247 . First host does not offer TLS;
2248 . First host accepts first address;
2249 . First host gives temporary error to second address;
2250 . Second host offers TLS and a TLS session is established;
2251 . Second host accepts second address.
2252 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2253 and peerdn, if requested) that were in fact used only for the second
2256 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2257 attention to any local part prefix or suffix that was matched by the
2258 router that accepted the address. It now behaves in the same way as it
2259 does for delivery: the affixes are removed from the local part unless
2260 rcpt_include_affixes is set on the transport.
2262 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2263 timeout during the DATA phase of an incoming message.
2265 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2266 applied a patch sanctioned by the Sieve maintainer.
2268 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2269 the uid or gid is negative. A case of a negative gid caused this to be
2270 noticed. The fix allows for either to be negative.
2272 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2273 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2276 PH/36 The change PH/12 above was broken. Fixed it.
2278 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2279 the grounds that routing the same address twice would always produce the
2280 same answer. This might have been true once, but it is certainly no
2281 longer true now. Routing a child address may depend on the previous
2282 routing that produced that child. Some complicated redirection strategies
2283 went wrong when messages had multiple recipients, and made Exim's
2284 behaviour dependent on the order in which the addresses were given.
2286 I have moved the duplicate checking until after the routing is complete.
2287 Exim scans the addresses that are assigned to local and remote
2288 transports, and removes any duplicates. This means that more work will be
2289 done, as duplicates will always all be routed, but duplicates are
2290 presumably rare, so I don't expect this is of any significance.
2292 For deliveries to pipes, files, and autoreplies, the duplicate checking
2293 still happens during the routing process, since they are not going to be
2296 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2297 It corrects a timeout issue with spamd. This is Ian's comment: "The
2298 background is that sometimes spamd either never reads data from a
2299 connection it has accepted, or it never writes response data. The exiscan
2300 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2301 blindly assumes that writes won't block so it may never time out."
2303 PH/39 Allow G after quota size as well as K and M.
2305 PH/40 The value set for $authenticated_id in an authenticator may not contain
2306 binary zeroes or newlines because the value is written to log lines and
2307 to spool files. There was no check on this. Now the value is run through
2308 the string_printing() function so that such characters are converted to
2309 printable escape sequences.
2311 PH/41 $message_linecount is a new variable that contains the total number of
2312 lines in the message. Compare $body_linecount, which is the count for the
2315 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2316 bounce and delay warning messages, except in certain special cases, which
2319 (a) An SMTP error message from a remote host;
2320 (b) A message specified in a :fail: redirection;
2321 (c) A message specified in a "fail" command in a system filter;
2322 (d) A message specified in a FAIL return from the queryprogram router;
2323 (e) A message specified by the cannot_route_message router option.
2325 In these cases only, Exim does include the error details in bounce and
2326 warning messages. There are also a few cases where bland messages such
2327 as "unrouteable address" or "local delivery error" are given.
2329 PH/43 $value is now also set for the "else" part of a ${run expansion.
2331 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2332 being worked on, but at least Exim now implements the latest version to
2335 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2336 process to complete was treated as a delivery failure, a timeout while
2337 writing the message to the pipe was logged, but erroneously treated as a
2338 successful delivery. Such timeouts include transport filter timeouts. For
2339 consistency with the overall process timeout, these timeouts are now
2340 treated as errors, giving rise to delivery failures by default. However,
2341 there is now a new Boolean option for the pipe transport called
2342 timeout_defer, which, if set TRUE, converts the failures into defers for
2343 both kinds of timeout. A transport filter timeout is now identified in
2346 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2347 systems where "make" and "gmake" are different, calling "gmake" at top
2348 level broke things. I've arranged for the value of $(MAKE) to be passed
2349 from the Makefile to this script so that it can call the same version of
2353 A note about Exim versions 4.44 and 4.50
2354 ----------------------------------------
2356 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2357 changes of various kinds. As a consequence, a big documentation update was
2358 needed. This delayed the release for rather longer than seemed good, especially
2359 in the light of a couple of (minor) security issues. Therefore, the changes
2360 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2361 release. So 4.44 and 4.50 are in effect two different branches that both start
2364 I have left the 4.50 change log unchanged; it contains all the changes since
2365 4.43. The change log for 4.44 is below; many of its items are identical to
2366 those for 4.50. This seems to be the most sensible way to preserve the
2367 historical information.
2373 1. Minor wording change to the doc/README.SIEVE file.
2375 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2376 computation of the current number of files was incorrect.
2378 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2379 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2382 4. Give more explanation in the error message when the command for a transport
2383 filter fails to execute.
2385 5. There are several places where Exim runs a non-Exim command in a
2386 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2387 was being done only for the command run by the queryprogram router. It is
2388 now done for all such subprocesses. The other cases are: ${run, transport
2389 filters, and the commands run by the lmtp and pipe transports.
2391 6. Added CONFIGURE_GROUP build-time option.
2393 7. Some older OS have a limit of 256 on the maximum number of file
2394 descriptors. Exim was using setrlimit() to set 1000 as a large value
2395 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2396 systems. I've change it so that if it can't get 1000, it tries for 256.
2398 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2399 was an oversight, and furthermore, ever since the addition of extra
2400 controls (e.g. 4.43/32), the checks on when to allow different forms of
2401 "control" were broken. There should now be diagnostics for all cases when a
2402 control that does not make sense is encountered.
2404 9. Added the /retain_sender option to "control=submission".
2406 10. $recipients is now available in the predata ACL (oversight).
2408 11. Tidy the search cache before the fork to do a delivery from a message
2409 received from the command line. Otherwise the child will trigger a lookup
2410 failure and thereby defer the delivery if it tries to use (for example) a
2411 cached ldap connection that the parent has called unbind on.
2413 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2414 of $address_data from the recipient verification was clobbered by the
2415 sender verification.
2417 13. The value of address_data from a sender verification is now available in
2418 $sender_address_data in subsequent conditions in the ACL statement.
2420 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2422 15. Added a new option "connect=<time>" to callout options, to set a different
2425 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2426 was its contents. (It was OK if the option was not defined at all.)
2428 17. A "Completed" log line is now written for messages that are removed from
2429 the spool by the -Mrm option.
2431 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2432 information about exactly what failed.
2434 19. Added -dd to debug only the daemon process.
2436 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2437 handles timeouts, both on the server side and network timeouts. Renamed the
2438 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2440 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2441 It is now set to "smtps".
2443 22. $host_address is now set to the target address during the checking of
2444 ignore_target_hosts.
2446 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2447 being passed; this would have caused $sender_host_name to have been used if
2448 matching the list had actually called for a host name (not very likely,
2449 since this list is usually IP addresses). A host name is now passed as
2452 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2453 code that specifies a non-restarting handler (typically sigaction() in
2454 modern systems) in an attempt to fix a rare and obscure crash bug.
2456 25. Narrowed the window for a race in the daemon that could cause it to ignore
2457 SIGCHLD signals. This is not a major problem, because they are used only to
2458 wake it up if nothing else does.
2460 26. A malformed maildirsize file could cause Exim to calculate negative values
2461 for the mailbox size or file count. Odd effects could occur as a result.
2462 The maildirsize information is now recalculated if the size or filecount
2465 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2466 support for a long time. Removed HAVE_SYS_VFS_H.
2468 28. Installed the latest version of exipick from John Jetmore.
2470 29. In an address list, if the pattern was not a regular expression, an empty
2471 subject address (from a bounce message) matched only if the pattern was an
2472 empty string. Non-empty patterns were not even tested. This was the wrong
2473 because it is perfectly reasonable to use an empty address as part of a
2474 database query. An empty address is now tested by patterns that are
2475 lookups. However, all the other forms of pattern expect the subject to
2476 contain a local part and a domain, and therefore, for them, an empty
2477 address still always fails if the pattern is not itself empty.
2479 30. Exim went into a mad DNS loop when attempting to do a callout where the
2480 host was specified on an smtp transport, and looking it up yielded more
2481 than one IP address.
2483 31. Re-factored the code for checking spool and log partition space into a
2484 function that finds that data and another that does the check. The former
2485 is then used to implement four new variables: $spool_space, $log_space,
2486 $spool_inodes, and $log_inodes.
2488 32. The RFC2047 encoding function was originally intended for short strings
2489 such as real names; it was not keeping to the 75-character limit for
2490 encoded words that the RFC imposes. It now respects the limit, and
2491 generates multiple encoded words if necessary. To be on the safe side, I
2492 have increased the buffer size for the ${rfc2047: expansion operator from
2495 33. It is now permitted to omit both strings after an "if" condition; if the
2496 condition is true, the result is "true". As before, when the second string
2497 is omitted, a false condition yields an empty string. This makes it less
2498 cumbersome to write custom ACL and router conditions.
2500 34. Failure to deliver a bounce message always caused it to be frozen, even if
2501 there was an errors_to setting on the router. The errors_to setting is now
2504 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2505 canonical form (fully expanded) before being placed in
2506 $sender_host_address.
2508 36. The table in the code that translates DNS record types into text (T_A to
2509 "A" for instance) was missing entries for NS and CNAME. It is just possible
2510 that this could have caused confusion if both these types were looked up
2511 for the same domain, because the text type is used as part of Exim's
2512 per-process caching. But the chance of anyone hitting this buglet seems
2515 37. The dnsdb lookup has been extended in a number of ways.
2517 (1) There is a new type, "zns", which walks up the domain tree until it
2518 finds some nameserver records. It should be used with care.
2520 (2) There is a new type, "mxh", which is like "mx" except that it returns
2521 just the host names, not the priorities.
2523 (3) It is now possible to give a list of domains (or IP addresses) to be
2524 looked up. The behaviour when one of the lookups defers can be
2525 controlled by a keyword.
2527 (4) It is now possible to specify the separator character for use when
2528 multiple records are returned.
2530 38. The dnslists ACL condition has been extended: it is now possible to supply
2531 a list of IP addresses and/or domains to be looked up in a particular DNS
2534 39. Added log_selector=+queue_time_overall.
2536 40. When running the queue in the test harness, wait just a tad after forking a
2537 delivery process, to get repeatability of debugging output.
2539 41. Include certificate and key file names in error message when GnuTLS fails
2540 to set them up, because the GnuTLS error message doesn't include the name
2541 of the failing file when there is a problem reading it.
2543 42. Allow both -bf and -bF in the same test run.
2545 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2547 44. The "Exiscan patch" is now merged into the mainline Exim source.
2549 45. Sometimes the final signoff response after QUIT could fail to get
2550 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2551 < 0 before doing a fflush(). This bug looks as though it goes back to the
2552 introduction of TLS in release 3.20, but "sometimes" must have been rare
2553 because the tests only now provoked it.
2555 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2556 (this can affect the format of dates).
2558 47. exim_tidydb, when checking for the continued existence of a message for
2559 which it has found a message-specific retry record, was not finding
2560 messages that were in split spool directories. Consequently, it was
2561 deleting retry records that should have stayed in existence.
2563 48. Steve fixed some bugs in eximstats.
2565 49. The SPA authentication driver was not abandoning authentication and moving
2566 on to the next authenticator when an expansion was forced to fail,
2567 contradicting the general specification for all authenticators. Instead it
2568 was generating a temporary error. It now behaves as specified.
2570 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2571 (the order specifies the preference for clients). The order is now AES256,
2572 AES128, 3DES, ARCFOUR128.
2574 51. Small patch to Sieve code - explicitly set From: when generating an
2577 52. Exim crashed if a remote delivery caused a very long error message to be
2578 recorded - for instance if somebody sent an entire SpamAssassin report back
2579 as a large number of 550 error lines. This bug was coincidentally fixed by
2580 increasing the size of one of Exim's internal buffers (big_buffer) that
2581 happened as part of the Exiscan merge. However, to be on the safe side, I
2582 have made the code more robust (and fixed the comments that describe what
2585 53. Now that there can be additional text after "Completed" in log lines (if
2586 the queue_time_overall log selector is set), a one-byte patch to exigrep
2587 was needed to allow it to recognize "Completed" as not the last thing in
2590 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2591 patch that reportedly fixes this has been added. I am not expert enough to
2592 create a test for it. This is what the patch creator wrote:
2594 "I found a little strange behaviour of ldap code when working with
2595 Windows 2003 AD Domain, where users was placed in more than one
2596 Organization Units. When I tried to give exim partial DN, the exit code
2597 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2598 But simultaneously result of request was absolutely normal ldap result,
2599 so I produce this patch..."
2601 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2602 REFERENCE, so I have modified the code to exclude the patch when that macro
2605 55. Some experimental protocols are using DNS PTR records for new purposes. The
2606 keys for these records are domain names, not reversed IP addresses. The
2607 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2608 leaves it alone. Component reversal etc. now happens only for IP addresses.
2611 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2613 57. Double the size of the debug message buffer (to 2048) so that more of very
2614 long debug lines gets shown.
2616 58. The exicyclog utility now does better if the number of log files to keep
2617 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2619 59. Two changes related to the smtp_active_hostname option:
2621 (1) $smtp_active_hostname is now available as a variable.
2622 (2) The default for smtp_banner uses $smtp_active_hostname instead
2623 of $primary_hostname.
2625 60. The host_aton() function is supposed to be passed a string that is known
2626 to be a valid IP address. However, in the case of IPv6 addresses, it was
2627 not checking this. This is a hostage to fortune. Exim now panics and dies
2628 if the condition is not met. A case was found where this could be provoked
2629 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2630 components; fortuitously, this particular loophole had already been fixed
2631 by change 4.50/55 above.
2633 If there are any other similar loopholes, the new check in host_aton()
2634 itself should stop them being exploited. The report I received stated that
2635 data on the command line could provoke the exploit when Exim was running as
2636 exim, but did not say which command line option was involved. All I could
2637 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2638 running as the user.
2641 61. There was a buffer overflow vulnerability in the SPA authentication code
2642 (which came originally from the Samba project). I have added a test to the
2643 spa_base64_to_bits() function which I hope fixes it.
2646 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2647 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2649 63. The daemon start-up calls getloadavg() while still root for those OS that
2650 need the first call to be done as root, but it missed one case: when
2651 deliver_queue_load_max is set with deliver_drop_privilege. This is
2652 necessary for the benefit of the queue runner, because there is no re-exec
2653 when deliver_drop_privilege is set.
2655 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2656 This has been fixed.
2658 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2659 was in use, was not putting the data itself into the right store pool;
2660 consequently, it could be overwritten for a subsequent message in the same
2661 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2664 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2666 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2667 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2668 the function were treating the return as a boolean value, which happened to
2669 work because 0=false and not-0=true, but is not correct code.
2671 68. The host_aton() function was not handling scoped IPv6 addresses (those
2672 with, for example, "%eth0" on the end) correctly.
2674 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2675 negated items (that is, ~something) in unsigned ints. Some compilers
2676 apparently mutter when there is no cast.
2678 70. If an address verification called from an ACL failed, and did not produce a
2679 user-specific message (i.e. there was only a "system" message), nothing was
2680 put in $acl_verify_message. In this situation, it now puts the system
2683 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2684 session; change 4.31/43 added the unwanted input to the log line - except
2685 that it did not do this in the start of session case. It now does.
2687 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2688 This is wrong and can cause the other end to generate a synchronization
2689 error if it is another Exim or anything else that does the synchronization
2690 check. A QUIT command is no longer sent after a timeout.
2692 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2693 during host lookups.
2695 74. The defer_ok option of callout verification was not working if it was used
2696 when verifying addresses in header lines, that is, for this case:
2698 verify = header_sender/callout=defer_ok
2700 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2701 those file descriptors could be used for SMTP connections. If anything
2702 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2703 could be sent to the SMTP client, causing chaos. The daemon now opens
2704 stdin, stdout, and stderr to /dev/null when it puts itself into the
2707 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2708 log by default. The user can override this with suitable Perl magic.
2710 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2711 the log message when discard triggers, was not working for the DATA ACL or
2712 for the non-SMTP ACL.
2714 78. Error message wording change in sieve.c.
2716 79. If smtp_accept_max_per_host was set, the number of connections could be
2717 restricted to fewer than expected, because the daemon was trying to set up
2718 a new connection before checking whether the processes handling previous
2719 connections had finished. The check for completed processes is now done
2720 earlier. On busy systems, this bug wouldn't be noticed because something
2721 else would have woken the daemon, and it would have reaped the completed
2724 80. If a message was submitted locally by a user whose login name contained one
2725 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2726 It caused a spool format error. I have fixed the spool reading code. A
2727 related problem was that the "from" clause in the Received: line became
2728 illegal because of the space(s). It is now covered by ${quote_local_part.
2730 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2733 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2735 83. Patch from Sieve maintainer to fix unterminated string problem in
2736 "vacation" handling.
2738 84. Some minor changes to the Linux configuration files to help with other
2739 OS variants using glibc.
2741 85. One more patch for Sieve to update vacation handling to latest spec.
2744 ----------------------------------------------------
2745 See the note above about the 4.44 and 4.50 releases.
2746 ----------------------------------------------------
2752 1. Change 4.43/35 introduced a bug that caused file counts to be
2753 incorrectly computed when quota_filecount was set in an appendfile
2756 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2757 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2760 3. Give more explanation in the error message when the command for a transport
2761 filter fails to execute.
2763 4. There are several places where Exim runs a non-Exim command in a
2764 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2765 was being done only for the command run by the queryprogram router. It is
2766 now done for all such subprocesses. The other cases are: ${run, transport
2767 filters, and the commands run by the lmtp and pipe transports.
2769 5. Some older OS have a limit of 256 on the maximum number of file
2770 descriptors. Exim was using setrlimit() to set 1000 as a large value
2771 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2772 systems. I've change it so that if it can't get 1000, it tries for 256.
2774 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2775 was an oversight, and furthermore, ever since the addition of extra
2776 controls (e.g. 4.43/32), the checks on when to allow different forms of
2777 "control" were broken. There should now be diagnostics for all cases when a
2778 control that does not make sense is encountered.
2780 7. $recipients is now available in the predata ACL (oversight).
2782 8. Tidy the search cache before the fork to do a delivery from a message
2783 received from the command line. Otherwise the child will trigger a lookup
2784 failure and thereby defer the delivery if it tries to use (for example) a
2785 cached ldap connection that the parent has called unbind on.
2787 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2788 of $address_data from the recipient verification was clobbered by the
2789 sender verification.
2791 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2792 was its contents. (It was OK if the option was not defined at all.)
2794 11. A "Completed" log line is now written for messages that are removed from
2795 the spool by the -Mrm option.
2797 12. $host_address is now set to the target address during the checking of
2798 ignore_target_hosts.
2800 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2801 being passed; this would have caused $sender_host_name to have been used if
2802 matching the list had actually called for a host name (not very likely,
2803 since this list is usually IP addresses). A host name is now passed as
2806 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2807 code that specifies a non-restarting handler (typically sigaction() in
2808 modern systems) in an attempt to fix a rare and obscure crash bug.
2810 15. Narrowed the window for a race in the daemon that could cause it to ignore
2811 SIGCHLD signals. This is not a major problem, because they are used only to
2812 wake it up if nothing else does.
2814 16. A malformed maildirsize file could cause Exim to calculate negative values
2815 for the mailbox size or file count. Odd effects could occur as a result.
2816 The maildirsize information is now recalculated if the size or filecount
2819 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2820 support for a long time. Removed HAVE_SYS_VFS_H.
2822 18. Updated exipick to current release from John Jetmore.
2824 19. Allow an empty sender to be matched against a lookup in an address list.
2825 Previously the only cases considered were a regular expression, or an
2828 20. Exim went into a mad DNS lookup loop when doing a callout where the
2829 host was specified on the transport, if the DNS lookup yielded more than
2832 21. The RFC2047 encoding function was originally intended for short strings
2833 such as real names; it was not keeping to the 75-character limit for
2834 encoded words that the RFC imposes. It now respects the limit, and
2835 generates multiple encoded words if necessary. To be on the safe side, I
2836 have increased the buffer size for the ${rfc2047: expansion operator from
2839 22. Failure to deliver a bounce message always caused it to be frozen, even if
2840 there was an errors_to setting on the router. The errors_to setting is now
2843 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2844 canonical form (fully expanded) before being placed in
2845 $sender_host_address.
2847 24. Updated eximstats to version 1.33
2849 25. Include certificate and key file names in error message when GnuTLS fails
2850 to set them up, because the GnuTLS error message doesn't include the name
2851 of the failing file when there is a problem reading it.
2853 26. Expand error message when OpenSSL has problems setting up cert/key files.
2856 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2857 (this can affect the format of dates).
2859 28. exim_tidydb, when checking for the continued existence of a message for
2860 which it has found a message-specific retry record, was not finding
2861 messages that were in split spool directories. Consequently, it was
2862 deleting retry records that should have stayed in existence.
2864 29. eximstats updated to version 1.35
2865 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2866 1.35 - bugfix such that pie charts by volume are generated correctly
2868 30. The SPA authentication driver was not abandoning authentication and moving
2869 on to the next authenticator when an expansion was forced to fail,
2870 contradicting the general specification for all authenticators. Instead it
2871 was generating a temporary error. It now behaves as specified.
2873 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2874 (the order specifies the preference for clients). The order is now AES256,
2875 AES128, 3DES, ARCFOUR128.
2877 31. Small patch to Sieve code - explicitly set From: when generating an
2880 32. Exim crashed if a remote delivery caused a very long error message to be
2881 recorded - for instance if somebody sent an entire SpamAssassin report back
2882 as a large number of 550 error lines. This bug was coincidentally fixed by
2883 increasing the size of one of Exim's internal buffers (big_buffer) that
2884 happened as part of the Exiscan merge. However, to be on the safe side, I
2885 have made the code more robust (and fixed the comments that describe what
2888 33. Some experimental protocols are using DNS PTR records for new purposes. The
2889 keys for these records are domain names, not reversed IP addresses. The
2890 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2891 leaves it alone. Component reversal etc. now happens only for IP addresses.
2894 34. The host_aton() function is supposed to be passed a string that is known
2895 to be a valid IP address. However, in the case of IPv6 addresses, it was
2896 not checking this. This is a hostage to fortune. Exim now panics and dies
2897 if the condition is not met. A case was found where this could be provoked
2898 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2899 components; fortuitously, this particular loophole had already been fixed
2900 by change 4.50/55 or 4.44/33 above.
2902 If there are any other similar loopholes, the new check in host_aton()
2903 itself should stop them being exploited. The report I received stated that
2904 data on the command line could provoke the exploit when Exim was running as
2905 exim, but did not say which command line option was involved. All I could
2906 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2907 running as the user.
2910 35. There was a buffer overflow vulnerability in the SPA authentication code
2911 (which came originally from the Samba project). I have added a test to the
2912 spa_base64_to_bits() function which I hope fixes it.
2915 36. The daemon start-up calls getloadavg() while still root for those OS that
2916 need the first call to be done as root, but it missed one case: when
2917 deliver_queue_load_max is set with deliver_drop_privilege. This is
2918 necessary for the benefit of the queue runner, because there is no re-exec
2919 when deliver_drop_privilege is set.
2921 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2922 was in use, was not putting the data itself into the right store pool;
2923 consequently, it could be overwritten for a subsequent message in the same
2924 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2927 38. Sometimes the final signoff response after QUIT could fail to get
2928 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2929 < 0 before doing a fflush(). This bug looks as though it goes back to the
2930 introduction of TLS in release 3.20, but "sometimes" must have been rare
2931 because the tests only now provoked it.
2937 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2938 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2939 Now it returns an integer. A number of calls were still expecting a T/F
2940 return. Fortuitously, in all cases, the tests worked in OK situations,
2941 which is the norm. However, things would have gone wrong on any write
2942 failures on the smtp file descriptor. This function is used when sending
2943 messages over SMTP and also when doing verify callouts.
2945 2. When Exim is called to do synchronous delivery of a locally submitted
2946 message (the -odf or -odi options), it no longer closes stderr before doing
2949 3. Implemented the mua_wrapper option.
2951 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2953 5. Implemented the functions header_remove(), header_testname(),
2954 header_add_at_position(), and receive_remove_recipient(), and exported them
2957 6. If an ACL "warn" statement specified the addition of headers, Exim already
2958 inserted X-ACL-Warn: at the start if there was no header name. However, it
2959 was not making this test for the second and subsequent header lines if
2960 there were newlines in the string. This meant that an invalid header could
2961 be inserted if Exim was badly configured.
2963 7. Allow an ACL "warn" statement to add header lines at the start or after all
2964 the Received: headers, as well as at the end.
2966 8. Added the rcpt_4xx retry error code.
2968 9. Added postmaster_mailfrom=xxx to callout verification option.
2970 10. Added mailfrom=xxxx to the callout verification option, for verify=
2973 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2974 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2976 12. Inserted some casts to stop certain compilers warning when using pointer
2977 differences as field lengths or precisions in printf-type calls (mostly
2978 affecting debugging statements).
2980 13. Added optional readline() support for -be (dynamically loaded).
2982 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2983 same clock tick as a message's arrival, so that its received time was the
2984 same as the "first fail" time on the retry record, and that message
2985 remained on the queue past the ultimate address timeout, every queue runner
2986 would try a delivery (because it was past the ultimate address timeout) but
2987 after another failure, the ultimate address timeout, which should have then
2988 bounced the address, did not kick in. This was a "< instead of <=" error;
2989 in most cases the first failure would have been in the next clock tick
2990 after the received time, and all would be well.
2992 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2993 being recognized when the domain list was tested by the match_domain
2994 condition in an expansion string.
2996 16. Added the ${str2b64: operator.
2998 17. Exim was always calling setrlimit() to set a large limit for the number of
2999 processes, without checking whether the existing limit was already
3000 adequate. (It did check for the limit on file descriptors.) Furthermore,
3001 errors from getrlimit() and setrlimit() were being ignored. Now they are
3002 logged to the main and panic logs, but Exim does carry on, to try to do its
3003 job under whatever limits there are.
3005 18. Imported PCRE 5.0.
3007 19. Trivial typo in log message " temporarily refused connection" (the leading
3010 20. If the log selector return_path_on_delivery was set and an address was
3011 redirected to /dev/null, the delivery process crashed because it assumed
3012 that a return path would always be set for a "successful" delivery. In this
3013 case, the whole delivery is bypassed as an optimization, and therefore no
3016 21. Internal re-arrangement: the function for sending a challenge and reading
3017 a response while authentication was assuming a zero-terminated challenge
3018 string. It's now changed to take a pointer and a length, to allow for
3019 binary data in such strings.
3021 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3023 23. Exim was not respecting finduser_retries when seeking the login of the
3024 uid under which it was called; it was always trying 10 times. (The default
3025 setting of finduser_retries is zero.) Also, it was sleeping after the final
3026 failure, which is pointless.
3028 24. Implemented tls_on_connect_ports.
3030 25. Implemented acl_smtp_predata.
3032 26. If the domain in control=submission is set empty, Exim assumes that the
3033 authenticated id is a complete email address when it generates From: or
3034 Sender: header lines.
3036 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3037 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3038 chown and chgrp in /bin and hostname in /usr/bin.
3040 28. Exim was keeping the "process log" file open after each use, just as it
3041 does for the main log. This opens the possibility of it remaining open for
3042 long periods when the USR1 signal hits a daemon. Occasional processlog
3043 errors were reported, that could have been caused by this. Anyway, it seems
3044 much more sensible not to leave this file open at all, so that is what now
3047 29. The long-running daemon process does not normally write to the log once it
3048 has entered its main loop, and it closes the log before doing so. This is
3049 so that log files can straightforwardly be renamed and moved. However,
3050 there are a couple of unusual error situations where the daemon does write
3051 log entries, and I had neglected to close the log afterwards.
3053 30. The text of an SMTP error response that was received during a remote
3054 delivery was being truncated at 512 bytes. This is too short for some of
3055 the long messages that one sometimes sees. I've increased the limit to
3058 31. It is now possible to make retry rules that apply only when a message has a
3059 specific sender, in particular, an empty sender.
3061 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3062 it possible to be selective about when SMTP synchronization is enforced.
3064 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3066 32. Implemented hosts_connection_nolog.
3068 33. Added an ACL for QUIT.
3070 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3073 35. Added mailbox_size and mailbox_filecount to appendfile.
3075 36. Added control = no_multiline_responses to ACLs.
3077 37. There was a bug in the logic of the code that waits for the clock to tick
3078 in the case where the clock went backwards by a substantial amount such
3079 that the microsecond fraction of "now" was more than the microsecond
3080 fraction of "then" (but the whole seconds number was less).
3082 38. Added support for the libradius Radius client library this is found on
3083 FreeBSD (previously only the radiusclient library was supported).
3089 1. When certain lookups returned multiple values in the form name=value, the
3090 quoting of the values was not always being done properly. Specifically:
3091 (a) If the value started with a double quote, but contained no whitespace,
3093 (b) If the value contained whitespace other than a space character (i.e.
3094 tabs or newlines or carriage returns) it was not quoted.
3095 This fix has been applied to the mysql and pgsql lookups by writing a
3096 separate quoting function and calling it from the lookup code. The fix
3097 should probably also be applied to nisplus, ibase and oracle lookups, but
3098 since I cannot test any of those, I have not disturbed their existing code.
3100 2. A hit in the callout cache for a specific address caused a log line with no
3101 reason for rejecting RCPT. Now it says "Previous (cached) callout
3102 verification failure".
3104 3. There was an off-by-one bug in the queryprogram router. An over-long
3105 return line was truncated at 256 instead of 255 characters, thereby
3106 overflowing its buffer with the terminating zero. As well as fixing this, I
3107 have increased the buffer size to 1024 (and made a note to document this).
3109 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3110 when Exim is waiting for an SMTP response from a remote server, Exim
3111 restarts its select() call on the socket, thereby resetting its timeout.
3112 This is not a problem when such interrupts are rare. Somebody set up a cron
3113 job to run exiwhat every 2 minutes, which is less than the normal select()
3114 timeout (5 or 10 minutes). This meant that the select() timeout never
3115 kicked in because it was always reset. I have fixed this by comparing the
3116 time when an interrupt arrives with the time at the start of the first call
3117 to select(). If more time than the timeout has elapsed, the interrupt is
3118 treated as a timeout.
3120 5. Some internal re-factoring in preparation for the addition of Sieve
3121 extensions (by MH). In particular, the "personal" test is moved to a
3122 separate function, and given an option for scanning Cc: and Bcc: (which is
3123 not set for Exim filters).
3125 6. When Exim created an email address using the login of the caller as the
3126 local part (e.g. when creating a From: or Sender: header line), it was not
3127 quoting the local part when it contained special characters such as @.
3129 7. Installed new OpenBSD configuration files.
3131 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3132 try to make them clearer.
3134 9. Callout options, other than the timeout value, were being ignored when
3135 verifying sender addresses in header lines. For example, when using
3137 verify = header_sender/callout=no_cache
3139 the cache was (incorrectly) being used.
3141 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3142 only the Cygwin environment.
3144 11. When return_path_on_delivery was set as a log selector, if different remote
3145 addresses in the same message used different return paths and parallel
3146 remote delivery occurred, the wrong values would sometimes be logged.
3147 (Whenever a remote delivery process finished, the return path value from
3148 the most recently started remote delivery process was logged.)
3150 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3151 lines when AUTH and/or TLS are in use. This is the "received protocol"
3152 field. Exim used to use "asmtp" for authenticated SMTP, without any
3153 indication (in the protocol name) for TLS use. Now it follows the RFC and
3154 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3155 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3156 names appear in log lines as well as in Received: header lines.
3158 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3159 extensions, and comparison tests, and to fix some bugs.
3161 14. Changes to the "personal" filter test:
3163 (1) The test was buggy in that it was just doing the equivalent of
3164 "contains" tests on header lines. For example, if a user's address was
3165 anne@some.where, the "personal" test would incorrectly be true for
3167 To: susanne@some.where
3169 This test is now done by extracting each address from the header in turn,
3170 and checking the entire address. Other tests that are part of "personal"
3171 are now done using regular expressions (for example, to check local parts
3172 of addresses in From: header lines).
3174 (2) The list of non-personal local parts in From: addresses has been
3175 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3176 taken from the Sieve specification recommendations.
3178 (3) If the message contains any header line starting with "List-" it is
3179 treated as non-personal.
3181 (4) The test for "circular" in the Subject: header line has been removed
3182 because it now seems ill-conceived.
3184 15. Minor typos in src/EDITME comments corrected.
3186 16. Installed latest exipick from John Jetmore.
3188 17. If headers_add on a router specified a text string that was too long for
3189 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3190 of string_sprintf() is now avoided.
3192 18. $message_body_size was not set (it was always zero) when running the DATA
3193 ACL and the local_scan() function.
3195 19. For the "mail" command in an Exim filter, no default was being set for
3196 the once_repeat time, causing a random time value to be used if "once" was
3197 specified. (If the value happened to be <= 0, no repeat happened.) The
3198 default is now 0s, meaning "never repeat". The "vacation" command was OK
3199 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3200 (I found it when inspecting the code).
3202 20. There is now an overall timeout for performing a callout verification. It
3203 defaults to 4 times the callout timeout, which applies to individual SMTP
3204 commands during the callout. The overall timeout applies when there is more
3205 than one host that can be tried. The timeout is checked before trying the
3206 next host. This prevents very long delays if there are a large number of
3207 hosts and all are timing out (e.g. when the network connections are timing
3208 out). The value of the overall timeout can be changed by specifying an
3209 additional sub-option for "callout", called "maxwait". For example:
3211 verify = sender/callout=5s,maxwait=20s
3213 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3214 to the end before writing, but this should make it even safer).
3216 22. Exim was forgetting that it had advertised PIPELINING for the second and
3217 subsequent messages on an SMTP connection. It was also not resetting its
3218 memory on STARTTLS and an internal HELO.
3220 23. When Exim logs an SMTP synchronization error within a session, it now
3221 records whether PIPELINING has been advertised or not.
3223 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3224 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3225 rather than long int.
3227 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3229 26. Added the never_mail option to autoreply.
3235 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3236 crash if the getsockname() call failed; this can happen if a connection is
3237 closed very soon after it is established. The problem was simply in the
3238 order in which certain operations were done, causing Exim to try to write
3239 to the SMTP stream before it had set up the file descriptor. The bug has
3240 been fixed by making things happen in the correct order.
3246 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3247 before the connection was closed, thus losing the rejection response.
3249 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3250 some early Solaris releases, but causes trouble in current releases where
3251 socklen_t is defined.
3253 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3256 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3259 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3260 that releasing the top of it at the end releases what was used for sub-
3261 expansions (unless the block got too big). However, discard this block if
3262 the first thing is a variable or header, so that we can use its block when
3263 it is dynamic (useful for very large $message_headers, for example).
3265 6. Lookups now cache *every* query, not just the most recent. A new, separate
3266 store pool is used for this. It can be recovered when all lookup caches are
3267 flushed. Lookups now release memory at the end of their result strings.
3268 This has involved some general refactoring of the lookup sources.
3270 7. Some code has been added to the store_xxx() functions to reduce the amount
3271 of flapping under certain conditions.
3273 8. log_incoming_interface used to affect only the <= reception log lines. Now
3274 it causes the local interface and port to be added to several more SMTP log
3275 lines, for example "SMTP connection from", and rejection lines.
3277 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3279 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3281 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3282 could be overwritten at the end of the current message (or the start of a
3283 new message if it was set in a HELO ACL). The value is now preserved for
3284 the duration of the SMTP connection.
3286 12. If a transport had a headers_rewrite setting, and a matching header line
3287 contained an unqualified address, that address was qualified, even if it
3288 did not match any rewriting rules. The underlying bug was that the values
3289 of the flags that permit the existence of unqualified sender and recipient
3290 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3291 non-local messages, and by -bnq for local messages) were not being
3292 preserved with the message after it was received.
3294 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3295 "next input=" as part of the text comprising the host identity instead of
3296 the correct text. The code was using the same buffer for two different
3297 strings. However, depending on which order the printing function evaluated
3298 its arguments, the bug did not always show up. Under Linux, for example, my
3299 test suite worked just fine.
3301 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3302 doesn't work with some older versions of Perl. It has been changed to "my",
3303 which in any case is probably the better facility to use.
3305 15. A really picky compiler found some instances of statements for creating
3306 error messages that either had too many or two few arguments for the format
3309 16. The size of the buffer for calls to the DNS resolver has been increased
3310 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3311 for addresses that have a lot of PTR records. This alleviates a problem; it
3312 does not fully solve it.
3314 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3315 buffer now truncates the list and logs the incident, which is the same
3316 action as happens when Exim is looking up a host name and its aliases.
3317 Previously in this situation something unpredictable would happen;
3318 sometimes it was "internal error: store_reset failed".
3320 18. If a server dropped the connection unexpectedly when an Exim client was
3321 using GnuTLS and trying to read a response, the client delivery process
3322 crashed while trying to generate an error log message.
3324 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3325 string, for example:
3327 warn message = H1: something\nH2: something
3329 the text was added as a single header line from Exim's point of view
3330 though it ended up OK in the delivered message. However, searching for the
3331 second and subsequent header lines using $h_h2: did not work. This has been
3332 fixed. Similarly, if a system filter added multiple headers in this way,
3333 the routers could not see them.
3335 20. Expanded the error message when iplsearch is called with an invalid key to
3336 suggest using net-iplsearch in a host list.
3338 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3339 ACLs are no longer actually imposed (and a message to that effect is
3342 22. If a "gecos" field in a passwd entry contained escaped characters, in
3343 particular, if it contained a \" sequence, Exim got it wrong when building
3344 a From: or a Sender: header from that name. A second bug also caused
3345 incorrect handling when an unquoted " was present following a character
3346 that needed quoting.
3348 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3349 was not being matched caselessly.
3351 24. Arranged for all hyphens in the exim.8 source to be escaped with
3354 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3355 properly. Recipient callout cache records were still being keyed to include
3356 the sender, even when use_sender was set false. This led to far more
3357 callouts that were necessary. The sender is no longer included in the key
3358 when use_sender is false.
3360 26. Added "control = submission" modifier to ACLs.
3362 27. Added the ${base62d: operator to decode base 62 numbers.
3364 28. dnsdb lookups can now access SRV records.
3366 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3367 the configuration file.
3369 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3370 (-v) mode. This makes the output for a verbose queue run more intelligible.
3372 31. Added a use_postmaster feature to recipient callouts.
3374 32. Added the $body_zerocount variable, containing the number of binary zero
3375 bytes in the message body.
3377 33. The time of last modification of the "new" subdirectory is now used as the
3378 "mailbox time last read" when there is a quota error for a maildir
3381 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3383 35. Added +ignore_unknown as a special item in host lists.
3385 36. Code for decoding IPv6 addresses in host lists is now included, even if
3386 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3387 address was recognized as an IP address, but was then not correctly decoded
3388 into binary, causing unexpected and incorrect effects when compared with
3395 1. Very minor rewording of debugging text in manualroute to say "list of
3396 hosts" instead of "hostlist".
3398 2. If verify=header_syntax was set, and a header line with an unqualified
3399 address (no domain) and a large number of spaces between the end of the
3400 name and the colon was received, the reception process suffered a buffer
3401 overflow, and (when I tested it) crashed. This was caused by some obsolete
3402 code that should have been removed. The fix is to remove it!
3404 3. When running in the test harness, delay a bit after writing a bounce
3405 message to get a bit more predictability in the log output.
3407 4. Added a call to search_tidyup() just before forking a reception process. In
3408 theory, someone could use a lookup in the expansion of smtp_accept_max_
3409 per_host which, without the tidyup, could leave open a database connection.
3411 5. Added the variables $recipient_data and $sender_data which get set from a
3412 lookup success in an ACL "recipients" or "senders" condition, or a router
3413 "senders" option, similar to $domain_data and $local_part_data.
3415 6. Moved the writing of debug_print from before to after the "senders" test
3418 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3419 problems for message scanning, either using a data ACL, or using
3420 local_scan() because the Received: header was not generated till after they
3421 were called (in order to set the time as the time of reception completion).
3422 I have revised the way this works. The header is now generated after the
3423 body is received, but before the ACL or local_scan() are called. After they
3424 are run, the timestamp in the header is updated.
3430 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3431 before starting a queue runner without re-exec. This happened only when
3432 deliver_drop_privilege was set or when the Exim user was set to root. The
3433 effect of the bug was that timeouts during subsequent deliveries caused
3434 crashes instead of being properly handled. The handler is now left at its
3435 default (and expected) setting.
3437 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3438 message, again when deliver_drop_privilege is set or Exim is run as root.
3439 The bug described in (1) was not present in this case, but the tidying up
3440 of the other signals was missing. I have made the two cases consistent.
3442 3. The ignore_target_hosts setting on a manualroute router was being ignored
3443 for hosts that were looked up using the /MX notation.
3445 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3448 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3449 operated on the sender address. After changing the $sender_address to <>
3450 for the sender address verify, Exim was re-instated it as the original
3451 (before rewriting) address, but remembering that it had rewritten it, so it
3452 wasn't rewriting it again. This bug also had the effect of breaking the
3453 sender address verification caching when the sender address was rewritten.
3455 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3456 This has been changed so that if the ip literal address matches
3457 ignore_target_hosts, the router declines.
3459 7. Added expansion conditions match_domain, match_address, and match_local_
3460 part (NOT match_host).
3462 8. The placeholder for the Received: header didn't have a length field set.
3464 9. Added code to Exim itself and to exim_lock to test for a specific race
3465 condition that could lead to file corruption when using MBX delivery. The
3466 issue is with the lockfile that is created in /tmp. If this file is removed
3467 after a process has opened it but before that process has acquired a lock,
3468 there is the potential for a second process to recreate the file and also
3469 acquire a lock. This could lead to two Exim processes writing to the file
3470 at the same time. The added code performs the same test as UW imapd; it
3471 checks after acquiring the lock that its file descriptor still refers to
3472 the same named file.
3474 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3475 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3478 11. Added the smtp_active_hostname option. If used, this will typically be made
3479 to depend on the incoming interface address. Because $interface_address is
3480 not set up until the daemon has forked a reception process, error responses
3481 that can happen earlier (such as "too many connections") no longer contain
3484 12. If an expansion in a condition on a "warn" statement fails because a lookup
3485 defers, the "warn" statement is abandoned, and the next ACL statement is
3486 processed. Previously this caused the whole ACL to be aborted.
3488 13. Added the iplsearch lookup type.
3490 14. Added ident_timeout as a log selector.
3492 15. Added tls_certificate_verified as a log selector.
3494 16. Added a global option tls_require_ciphers (compare the smtp transport
3495 option of the same name). This controls incoming TLS connections.
3497 17. I finally figured out how to make tls_require_ciphers do a similar thing
3498 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3499 before starting the TLS session.
3501 18. Tabs are now shown as \t in -bP output.
3503 19. If the log selector return_path_on_delivery was set, Exim crashed when
3504 bouncing a message because it had too many Received: header lines.
3506 20. If two routers both had headers_remove settings, and the first one included
3507 a superfluous trailing colon, the final name in the first list and the
3508 first name in the second list were incorrectly joined into one item (with a
3509 colon in the middle).
3515 1. Added -C and -D options to the exinext utility, mainly to make it easier
3516 to include in the automated testing, but these could be helpful when
3517 multiple configurations are in use.
3519 2. The exinext utility was not formatting the output nicely when there was
3520 an alternate port involved in the retry record key, nor when there was a
3521 message id as well (for retries that were specific to a specific message
3522 and a specific host). It was also confused by IPv6 addresses, because of
3523 the additional colons they contain. I have fixed the IPv4 problem, and
3524 patched it up to do a reasonable job for IPv6.
3526 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3527 delivery, the log line now contains "pipelined" if PIPELINING was used.
3529 4. An SMTP transport process used to panic and die if the bind() call to set
3530 an explicit outgoing interface failed. This has been changed; it is now
3531 treated in the same way as a connect() failure.
3533 5. A reference to $sender_host_name in the part of a conditional expansion
3534 that was being skipped was still causing a DNS lookup. This no longer
3537 6. The def: expansion condition was not recognizing references to header lines
3538 that used bh_ and bheader_.
3540 7. Added the _cache feature to named lists.
3542 8. The code for checking quota_filecount in the appendfile transport was
3543 allowing one more file than it should have been.
3545 9. For compatibility with Sendmail, the command line option
3553 and sets the incoming protocol and host name (for trusted callers). The
3554 host name and its colon can be omitted when only the protocol is to be set.
3555 Note the Exim already has two private options, -pd and -ps, that refer to
3556 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3557 "s", but I don't think that's a major issue.
3559 10. A number of refactoring changes to the code, none of which should affect
3562 (a) The number of logging options was getting close to filling up the
3563 32-bit word that was used as a bit map. I have split them into two classes:
3564 those that are passed in the argument to log_write(), and those that are
3565 only ever tested independently outside of that function. These are now in
3566 separate 32-bit words, so there is plenty of room for expansion again.
3567 There is no change in the user interface or the logging behaviour.
3569 (b) When building, for example, log lines, the code previously used a
3570 macro that called string_cat() twice, in order to add two strings. This is
3571 not really sufficiently general. Furthermore, there was one instance where
3572 it was actually wrong because one of the argument was used twice, and in
3573 one call a function was used. (As it happened, calling the function twice
3574 did not affect the overall behaviour.) The macro has been replaced by a
3575 function that can join an arbitrary number of extra strings onto a growing
3578 (c) The code for expansion conditions now uses a table and a binary chop
3579 instead of a serial search (which was left over from when there were very
3580 few conditions). Also, it now recognizes conditions like "pam" even when
3581 the relevant support is not compiled in: a suitably worded error message is
3582 given if an attempt is made to use such a condition.
3584 11. Added ${time_interval:xxxxx}.
3586 12. A bug was causing one of the ddress fields not to be passed back correctly
3587 from remote delivery subprocesses. The field in question was not being
3588 subsequently used, so this caused to problems in practice.
3590 13. Added new log selectors queue_time and deliver_time.
3592 14. Might have fixed a bug in maildirsizefile handling that threw up
3593 "unexpected character" debug warnings, and recalculated the data
3594 unnecessarily. In any case, I expanded the warning message to give more
3597 15. Added the message "Restricted characters in address" to the statements in
3598 the default ACL that block characters like @ and % in local parts.
3600 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3601 Three changes have been made:
3603 (a) There was a serious bug; a negative response to MAIL caused the whole
3604 recipient domain to be cached as invalid, thereby blocking all messages
3605 to all local parts at the same domain, from all senders. This bug has
3606 been fixed. The domain is no longer cached after a negative response to
3607 MAIL if the sender used is not empty.
3609 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3612 (c) A new callout option, "use_sender" has been added for people who want
3613 the modified behaviour.
3619 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3622 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3623 indeed breaks things for older releases.
3625 3. Added additional logging to the case where there is a problem reading data
3626 from a filter that is running in a subprocess using a pipe, in order to
3627 try to track down a specific problem.
3629 4. Testing facility fudge: when running in the test harness and attempting
3630 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3631 getting "No route to host". Convert this to a timeout.
3633 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3636 6. Some OS don't have socklen_t but use size_t instead. This affects the
3637 fifth argument of getsockopt() amongst other things. This is now
3638 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3639 can be set for individual OS. I have set it for SunOS5, OSF1, and
3640 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3641 some earlier ones do not.
3643 7. Change 4.30/15 was not doing the test caselessly.
3645 8. The standard form for an IPv6 address literal was being rejected by address
3646 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3647 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3648 this, as well as the form without the "IPv6" on the front (but only when
3649 address literals are enabled, of course).
3651 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3653 10. Exim crashed if a message with an empty sender address specified by -f
3654 encountered a router with an errors_to setting. This could be provoked only
3655 by a command such as
3659 where an empty string was supplied; "<>" did not hit this bug.
3661 11. Installed PCRE release 4.5.
3663 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3664 remained set. It is now erased.
3666 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3667 times from message ids (which are base 36 rather than the normal 62).
3669 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3670 were being counted as actual protocol errors, and logged if the log
3671 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3672 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3673 and DATA following a set of rejected RCPTs do not count as protocol errors.
3674 In other words, Exim assumes they were pipelined, though this may not
3675 actually be the case. Of course, in all cases the client gets an
3676 appropriate error code.
3678 15. If a lookup fails in an ACL condition, a message about the failure may
3679 be available; it is used if testing the ACL cannot continue, because most
3680 such messages specify what the cause of the deferral is. However, some
3681 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3682 that caused an old message to be retained and used if a later statement
3683 caused a defer, replacing the real cause of the deferral.
3685 16. If an IP address had so many PTR records that the DNS lookup buffer
3686 was not large enough to hold them, Exim could crash while trying to process
3687 the truncated data. It now detects and logs this case.
3689 17. Further to 4.21/58, another change has been made: if (and only if) the
3690 first line of a message (the first header line) ends with CRLF, a bare LF
3691 in a subsequent header line has a space inserted after it, so as not to
3692 terminate the header.
3694 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3695 unnecessarily, used atoi() instead of strtol(), and didn't check the
3696 termination when getting file sizes from file names by regex.
3698 19. Completely re-implemented the support for maildirsize files, in the light
3699 of a number of problems with the previous contributed implementation
3700 (4.30/29). In particular:
3702 . If the quota is zero, the maildirsize file is maintained, but no quota is
3705 . If the maildir directory does not exist, it is created before any attempt
3706 to write a maildirsize file.
3708 . The quota value in the file is just a cache; if the quota is changed in
3709 the transport, the new value overrides.
3711 . A regular expression is available for excluding directories from the
3714 20. The autoreply transport checks the characters in options that define the
3715 message's headers; it allows continued headers, but it was checking with
3716 isspace() after an embedded newline instead of explicitly looking for a
3719 21. If all the "regular" hosts to which an address was routed had passed their
3720 expiry times, and had not reached their retry times, the address was
3721 bounced, even if fallback hosts were defined. Now Exim should go on to try
3724 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3725 equivalent code in the SMTP transport. Some hosts send humungous responses
3726 to HELO/EHLO, more than 1024 it seems.
3728 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3729 gives compiler warnings in some environments. I've now done it "properly",
3732 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3733 (because of problems with the built-in one) was declared to return uschar *
3734 instead of char *, causing compiler failure.
3736 25. Fixed a file descriptor leak when processing alias/forward files.
3738 26. Fixed a minor format string issue in dbfn.c.
3740 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3742 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3743 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3746 29. When a sender address is verified, it is cached, to save repeating the test
3747 when there is more than one recipient in a message. However, when the
3748 verification involves a callout, it is possible for different callout
3749 options to be set for different recipients. It is too complicated to keep
3750 track of this in the cache, so now Exim always runs a verification when a
3751 callout is required, relying on the callout cache for the optimization.
3752 The overhead is duplication of the address routing, but this should not be
3755 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3756 to be verified with callout=postmaster, and the main callout worked but the
3757 postmaster check failed, the verification correctly failed. However, if a
3758 subsequent RCPT command asked for sender verification *without* the
3759 postmaster check, incorrect caching caused this verification also to fail,
3762 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3763 it was not caching the DNS options (qualify_single, search_parents) that
3764 were used when the lookup failed. A subsequent lookup with different
3765 options therefore always gave the same answer, though there were cases
3766 where it should not have. (Example: a "domains = !$mx_any" option on a
3767 dnslookup router: the "domains" option is always processed without any
3768 widening, but the router might have qualify_single set.) Now Exim uses the
3769 cached value only when the same options are set.
3771 32. Added John Jetmore's "exipick" utility to the distribution.
3773 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3774 than a timeout (e.g. a certificate is required, and is not provided), an
3775 Exim server now closes the connection immediately. Previously it waited for
3776 the client to close - but if the client is SSL, it seems that they each
3777 wait for each other, leading to a delay before one of them times out.
3779 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3780 maintained 0.8.x compatibility because I don't think many are using it, and
3781 it is clearly obsolete.
3783 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3786 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3787 client certificate was expired. A simple patch fixes this, though I don't
3788 understand the full logic of why the verify callback is called multiple
3791 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3792 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3793 which causes problems with some clients (such as the Certicom SSL Plus
3794 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3795 disables the coutermeasure allowing Eudora to connect."
3797 38. Exim was not checking that a write() to a log file succeeded. This could
3798 lead to Bad Things if a log got too big, in particular if it hit a file
3799 size limit. Exim now panics and dies if it cannot write to a log file, just
3800 as it does if it cannot open a log file.
3802 39. Modified OS/Makefile-Linux so that it now contains
3804 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3806 The two -D definitions ensure that Exim is compiled with large file
3807 support, which makes it possible to handle log files that are bigger than
3810 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3811 instance) a domain was checked against a named list that involved a lookup,
3812 causing $domain_data to be set, then another domain was checked against the
3813 same list, then the first domain was re-checked, the value of $domain_data
3814 after the final check could be wrong. In particular, if the second check
3815 failed, it could be set empty. This bug probably also applied to
3818 41. The strip_trailing_dot option was not being applied to the address given
3819 with the -f command-line option.
3821 42. The code for reading a message's header from the spool was incrementing
3822 $received_count, but never initializing it. This meant that the value was
3823 incorrect (doubled) while delivering a message in the same process in which
3824 it was received. In the most common configuration of Exim, this never
3825 happens - a fresh exec is done - but it can happen when
3826 deliver_drop_privilege is set.
3828 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3829 it now includes up to 150 characters of the unexpected data in the log
3832 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3833 and building data strings. The size of both of these buffers was 10 000
3834 bytes - far larger than anybody would *ever* want, thought I. Needless to
3835 say, somebody hit the limit. I have increased the maximum line length to
3836 20 000 and the maximum data length of concatenated lines to 100 000. I have
3837 also fixed two bugs, because there was no checking on these buffers. Tsk,
3838 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3839 buffer is too small.
3841 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3842 lsearch lookups. Now it does.
3844 46. When parsing a route_list item in a manualroute router, a fixed-length
3845 buffer was used for the list of hosts. I made this 1024 bytes long,
3846 thinking that nobody would ever have a list of hosts that long. Wrong.
3847 Somebody had a whole pile of complicated expansion conditions, and the
3848 string was silently truncated, leading to an expansion error. It turns out
3849 that it is easier to change to an unlimited length (owing to other changes
3850 that have happened since this code was originally written) than to build
3851 structure for giving a limitation error. The length of the item that
3852 expands into the list of hosts is now unlimited.
3854 47. The lsearch lookup could not handle data where the length of text line was
3855 more than 4095 characters. Such lines were truncated, leading to shortened
3856 data being returned. It should now handle lines of any length.
3858 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3859 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3862 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3863 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3864 no longer happens in comment lines. A list of replacements is now placed
3865 at the head of all of the source files, except those whose only change is
3866 to replace PERL_COMMAND in the very first #! line.
3868 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3869 messages on the queue, with a bottom-up merge sort, using code contributed
3870 by Michael Haardt. This should make operations like -bp somewhat faster on
3871 large queues. It won't affect queue runners, except when queue_run_in_order
3874 51. Installed eximstats 1.31 in the distribution.
3876 52. Added support for SRV lookups to the dnslookup router.
3878 53. If an ACL referred to $message_body or $message_body_end, the value was not
3879 reset for any messages that followed in the same SMTP session.
3881 54. The store-handling optimization for building very long strings was not
3882 differentiating between the different store pools. I don't think this
3883 actually made any difference in practice, but I've tidied it.
3885 55. While running the routers to verify a sender address, $sender_address
3886 was still set to the sender address. This is wrong, because when routing to
3887 send a bounce to the sender, it would be empty. Therefore, I have changed
3888 it so that, while verifying a sender address, $sender_address is set to <>.
3889 (There is no change to what happens when verifying a recipient address.)
3891 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3892 target A or AAAA records (if not already returned) without resetting the
3893 qualify_single or search_parents options of the DNS resolver. These are
3894 inappropriate in this case because the targets of MX and SRV records must
3895 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3896 target that, when qualified, matched something in the local domain. These
3897 two options are now turned off when doing these lookups.
3899 57. It seems that at least some releases of Reiserfs (which does not have the
3900 concept of a fixed number of inodes) returns zero and not -1 for the
3901 number of available inodes. This interacted badly with check_spool_inodes,
3902 which assumed that -1 was the "no such thing" setting. What I have done is
3903 to check that the total number of inodes is greater than zero before doing
3904 the test of how many are available.
3906 58. When a "warn" ACL statement has a log_message modifier, the message is
3907 remembered, and not repeated. This is to avoid a lot of repetition when a
3908 message has many recipients that cause the same warning to be written.
3909 Howewer, Exim was preserving the list of already written lines for an
3910 entire SMTP session, which doesn't seem right. The memory is now reset if a
3911 new message is started.
3913 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3914 debugging output unless log_rewrite was also set.
3916 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3917 of (int)(handle) when we know that handle contains (void *)(-1).
3919 61. The Exim daemon panic-logs an error return when it closes the incoming
3920 connection. However "connection reset by peer" seems to be common, and
3921 isn't really an error worthy of noting specially, so that particular error
3924 62. When Exim is trying to find all the local interfaces, it used to panic and
3925 die if the ioctl to get the interface flags failed. However, it seems that
3926 on at least one OS (Solaris 9) it is possible to have an interface that is
3927 included in the list of interfaces, but for which you get a failure error
3928 for this call. This happens when the interface is not "plumbed" into a
3929 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3930 failure of the "get flags" call assumes that the interface is down.
3932 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3933 makes life easier for people who are doing arithmetic on fields extracted
3934 from dates, where you often get leading zeros that should not be
3935 interpreted as octal.
3937 64. Added qualify_domain to the redirect router, to override the global
3940 65. If a pathologically long header line contained very many addresses (the
3941 report of this problem mentioned 10 000) and each of them was rewritten,
3942 Exim could use up a very large amount of memory. (It kept on making new
3943 copies of the header line as it rewrote, and never released the old ones.)
3944 At the expense of a bit more processing, the header rewriting function has
3945 been changed so that it no longer eats memory in this way.
3947 66. The generation of the Received: header has been moved from the time that a
3948 message starts to be received, to the time that it finishes. The timestamp
3949 in the Received: header should now be very close to that of the <= log
3950 line. There are two side-effects of this change:
3952 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3953 logged header lines no longer include the local Received: line, because
3954 it has not yet been created. The same applies to a copy of the message
3955 that is returned to a non-SMTP sender when a message is rejected.
3957 (b) When a filter file is tested using -bf, no additional Received: header
3958 is added to the test message. After some thought, I decided that this
3961 This change does not affect the value of $received_for. It is still set
3962 after address rewriting, but before local_scan() is called.
3964 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3966 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3967 gave an unhelpful panic error message, and a defer error. I have managed to
3968 change this behaviour so that it now rejects any supplied certificate,
3969 which seems right, as the list of acceptable certificates is empty.
3971 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3972 gave an unhelpful defer error. I have not managed to make this reject any
3973 supplied certificates, but the error message it gives is "no certificate
3974 supplied", which is not helpful.
3976 70. exigrep's output now also includes lines that are not associated with any
3977 message, but which match the given pattern. Implemented by a patch from
3978 Martin Sluka, which also tidied up the Perl a bit.
3980 71. Recipient callout verification, like sender verification, was using <> in
3981 the MAIL FROM command. This isn't really the right thing, since the actual
3982 sender may affect whether the remote host accepts the recipient or not. I
3983 have changed it to use the actual sender in the callout; this means that
3984 the cache record is now keyed on a recipient/sender pair, not just the
3985 recipient address. There doesn't seem to be a real danger of callout loops,
3986 since a callout by the remote host to check the sender would use <>.
3987 [SEE ABOVE: changed after hitting problems.]
3989 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3990 temporary errors. However, in the case of such a code being given after
3991 the end of a data transmission (i.e. after ".") Exim was failing to write
3992 a retry record for the message. (Yes, there was some broken host that was
3993 actually sending 8xx at this point.)
3995 73. An unknown lookup type in a host list could cause Exim to panic-die when
3996 the list was checked. (An example that provoked this was putting <; in the
3997 middle of a list instead of at the start.) If this happened during a DATA
3998 ACL check, a -D file could be left lying around. This kind of configuration
3999 error no longer causes Exim to die; instead it causes a defer errror. The
4000 incident is still logged to the main and panic logs.
4002 74. Buglet left over from Exim 3 conversion. The message "too many messages
4003 in one connection" was written to the rejectlog but not the mainlog, except
4004 when address rewriting (yes!) was being logged.
4006 75. Added write_rejectlog option.
4008 76. When a system filter was run not as root (that is, when system_filter_user
4009 was set), the values of the $n variables were not being returned to the
4010 main process; thus, they were not subsequently available in the $sn
4013 77. Added +return_path_on_delivery log selector.
4015 78. A connection timeout was being treated differently from recipients deferred
4016 when testing hosts_max_try with a message that was older than the host's
4017 retry timeout. (The host should not be counted, thus allowing all hosts to
4018 be tried at least once before bouncing.) This may have been the cause of an
4019 occasionally reported bug whereby a message would remain on the queue
4020 longer than the retry timeout, but would be bounced if a delivery was
4021 forced. I say "may" because I never totally pinned down the problem;
4022 setting up timeout/retry tests is difficult. See also the next item.
4024 79. The ultimate address timeout was not being applied to errors that involved
4025 a combination of host plus message (for example, a timeout on a MAIL
4026 command). When an address resolved to a number of possible hosts, and they
4027 were not all tried for each delivery (e.g. because of hosts_max_try), a
4028 message could remain on the queue longer than the retry timeout.
4030 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4033 81. Fixed an obscure SMTP outgoing bug which required at least the following
4034 conditions: (a) there was another message waiting for the same server;
4035 (b) the server returned 5xx to all RCPT commands in the first message so
4036 that the message was not completed; (c) the server dropped the connection
4037 or gave a negative response to the RSET that Exim sends to abort the
4038 transaction. The observed case was a dropped connection after DATA that had
4039 been sent in pipelining mode. That is, the server had advertised PIPELINING
4040 but was not implementing it correctly. The effect of the bug was incorrect
4041 behaviour, such as trying another host, and this could lead to a crash.
4047 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4048 and daemon.c were passed as pointers to ints; they should have been
4049 pointers to socklen_t variables (which are typically unsigned ints).
4051 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4054 3. Fixed a really odd bug that affected only the testing scheme; patching a
4055 certain fixed string in the binary changed the value of another string that
4056 happened to be identical to the end of the original first string.
4058 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4059 name", it returns that address as the IP address. On some operating
4060 systems (e.g. Solaris), it also passes back the IP address string as the
4061 "host name". However, on others (e.g. Linux), it passes back an empty
4062 string. Exim wasn't checking for this, and was changing the host name to an
4063 empty string, assuming it had been canonicized.
4065 5. Although rare, it is permitted to have more than one PTR record for a given
4066 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4067 all the names associated with an address, because they do in Solaris.
4068 However, it seems that they do not in Linux for data that comes from the
4069 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4070 I found this out when I moved to a new Linux workstation and tried to run
4071 the Exim test suite.
4073 To get round this problem I have changed the code so that it now does its
4074 own call to the DNS to look up PTR records when searching for a host name.
4075 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4076 addresses that are only in /etc/hosts are still found.
4078 This behaviour is, however, controlled by an option called host_lookup_
4079 order, which defaults to "bydns:byaddr". If people want to use the other
4080 order, or indeed, just use one or the other means of lookup, they can
4081 specify it in this variable.
4083 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4084 some operating systems, this comes back from gethostbyaddr() as an empty
4085 string, and this is what Exim used to test for. However, it seems that in
4086 other systems, "." is yielded. Exim now tests for this case too.
4088 7. The values of check_spool_space and check_log_space are now held internally
4089 as a number of kilobytes instead of an absolute number of bytes. If a
4090 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4091 kilobyte. This means that much larger values can be stored.
4093 8. Exim monitor: an attempt to get the action menu when not actually pointing
4094 at a message produces an empty menu entitled "No message selected". This
4095 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4096 no entries in it ("Shell widget menu has zero width and/or height"). So I
4097 have added a single, blank menu entry in this case.
4099 9. Added ${quote_local_part.
4101 10. MIME decoding is now applied to the contents of Subject: header lines when
4104 11. Now that a reference to $sender_host_address automatically causes a reverse
4105 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4106 host lookup before query-style lookups in lists that might use this
4107 variable. This has therefore been abolished, and the "net-" prefix is no
4108 longer necessary for query-style lookups.
4110 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4111 have been a typo for LFLAGS, so it has been changed.
4113 13. The install script calls Exim with "-C /dev/null" in order to find the
4114 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4115 to be output. Howeve, since Exim outputs its version number before the
4116 error, it didn't break the script. It just looked ugly. I fixed this by
4117 always allowing "-C /dev/null" if the caller is root.
4119 14. Ignore overlarge ACL variable number when reading spool file - insurance
4120 against a later release with more variables having written the file.
4122 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4123 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4124 this, as well as the form without the "IPv6" on the front.
4126 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4127 OS/Makefile-Darwin file.
4129 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4130 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4132 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4133 variable. However, some OS use ints for inodes. I've added cast to long int
4134 to get rid of the compiler warning.
4136 19. I had forgotten to lock out "/../" in configuration file names when
4137 ALT_CONFIG_PREFIX was set.
4139 20. Routers used for verification do not need to specify transports. However,
4140 if such a router generated a host list, and callout was configured, Exim
4141 crashed, because it could not find a port number from the (non-existent)
4142 transport. It now assumes port 25 in this circumstance.
4144 21. Added the -t option to exigrep.
4146 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4147 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4148 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4149 them is set, LOOKUP_LSEARCH is forced.
4151 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4153 24. Added sender and host information to the "rejected by local_scan()" log
4154 line; previously there was no indication of these.
4156 25. Added .include_if_exists.
4158 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4159 Linux. It turns out that not all file systems support this. Apparently some
4160 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4161 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4162 not supported on the file descriptor, is now ignored when Exim is trying to
4163 sync a directory. This applies only to Linux.
4165 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4167 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4168 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4171 29. Added support for maildirsize files from supplied patch (modified a bit).
4173 30. The use of :fail: followed by an empty string could lead Exim to respond to
4174 sender verification failures with (e.g.):
4176 550 Verification failed for <xxx>
4177 550 Sender verify failed
4179 where the first response line was missing the '-' that indicates it is not
4180 the final line of the response.
4182 31. The loop for finding the name of the user that called Exim had a hardwired
4183 limit of 10; it now uses the value of finduser_retries, which is used for
4184 all other user lookups.
4186 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4189 33. Exim was neglecting to zero errno before one call of strtol() when
4190 expanding a string and expecting an integer value. On some systems this
4191 resulted in spurious "integer overflow" errors. Also, it was casting the
4192 result into an int without checking.
4194 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4195 did not work. The code looks as if it has *never* worked, though it appears
4196 to have been documented since at least releast 1.62. I have made it work.
4198 35. The "timeout_DNS" error in retry rules, also documented since at least
4199 1.62, also never worked. As it isn't clear exactly what this means, and
4200 clearly it isn't a major issue, I have abolished the feature by treating it
4201 as "timeout", and writing a warning to the main and panic logs.
4203 36. The display of retry rules for -brt wasn't always showing the error code
4206 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4207 timeout_connect_A, timeout_connect_MX.
4209 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4210 to the empty sender.
4212 39. The daemon was not analysing the content of -oX till after it had closed
4213 stderr and disconnected from the controlling terminal. This meant that any
4214 syntax errors were only noted on the panic log, and the return code from
4215 the command was 0. By re-arranging the code a little, I've made the
4216 decoding happen first, so such errors now appear on stderr, and the return
4217 code is 1. However, the actual setting up of the sockets still happens in
4218 the disconnected process, so errors there are still only recorded on the
4221 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4222 connections (as happens on some IP stacks) was logged at start up time as
4223 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4224 it from "IPv6 and IPv4", which means that two separate sockets are being
4227 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4228 says whether AF_INET or AF_INET6 was passed as an argument.
4230 42. Exiwhat output was messed up when time zones were included in log
4233 43. Exiwhat now gives more information about the daemon's listening ports,
4234 and whether -tls-on-connect was used.
4236 44. The "port" option of the smtp transport is now expanded.
4238 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4239 silently ignored. Now an error message is written to the main and panic
4242 46. There's a new ACL modifier called "logwrite" which writes to a log file
4243 as soon as it is encountered.
4245 47. Added $local_user_uid and $local_user_gid at routing time.
4247 48. Exim crashed when trying to verify a sender address that was being
4250 49. Exim was recognizing only a space character after ".include". It now also
4251 recognizes a tab character.
4253 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4254 extracting the relevant information from the specification. The man page no
4255 longer contains scrambled data for the -d option, and I've added a section
4256 at the front about calling Exim under different names.
4258 51. Added "extra_headers" argument to the "mail" command in filter files.
4260 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4263 53. Installed eximstats 1.29.
4265 54. Added transport_filter_timeout as a generic transport option.
4267 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4268 Cc: header lines. This was required by RFC 822, but it not required by RFC
4271 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4272 incoming messages that did not have them. Now it does so only if the
4273 message originates locally, that is, if there is no associated remote host
4274 address. When Resent- header lines are present, this applies to the Resent-
4275 lines rather than the non-Resent- lines.
4277 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4278 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4280 58. Messages for configuration errors now include the name of the main
4281 configuration file - useful now that there may be more than one file in a
4282 list (.included file names were always shown).
4284 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4285 for those rare installations that do not start the daemon as root or run it
4286 setuid root. I've cut out the call to initgroups() if the daemon is not
4289 60. The Exim user and group can now be bound into the binary as text strings
4290 that are looked up at the start of Exim's processing.
4292 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4294 62. Added $mailstore_basename variable.
4296 63. Installed patch to sieve.c from Michael Haardt.
4298 64. When Exim failed to open the panic log after failing to open the main log,
4299 the original message it was trying to log was written to stderr and debug
4300 output, but if they were not available (the usual case in production), it
4301 was lost. Now it is written to syslog before the two lines that record the
4302 failures to open the logs.
4304 65. Users' Exim filters run in subprocesses under the user's uid. It is
4305 possible for a "deliver" command or an alias in a "personal" command to
4306 provoke an address rewrite. If logging of address rewriting is configured,
4307 this fails because the process is not running as root or exim. There may be
4308 a better way of dealing with this, but for the moment (because 4.30 needs
4309 to be released), I have disabled address rewrite logging when running a
4310 filter in a non-root, non-exim process.
4316 1. The buildconfig auxiliary program wasn't quoting the value set for
4317 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4318 not defined. This bug was masked in 4.22 by the effect that was fixed in
4321 2. Some messages that were rejected after a message id was allocated were
4322 shown as "incomplete" by exigrep. It no longer does this for messages that
4323 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4325 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4326 have allow_domain_literals set, the ID did not get logged in the <= line.
4327 Domain literals are now always recognized in Message-ID: header lines.
4329 4. The first argument for a ${extract expansion item is the key name or field
4330 number. Leading and trailing spaces in this item were not being ignored,
4331 causing some misleading effects.
4333 5. When deliver_drop_privilege was set, single queue runner processes started
4334 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4335 same command in the process it spins off) were not dropping privilege.
4337 6. When the daemon running as "exim" started a queue runner, it always
4338 re-executed Exim in the spun-off process. This is a waste of effort when
4339 deliver_drop_privilege is set. The new process now just calls the
4340 queue-runner function directly.
4346 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4349 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4350 ignored. Though the use of -oP was forcing the writing of a pid file, it
4351 was always written to the default place.
4353 3. If the message "no IP address found for host xxxx" is generated during
4354 incoming verification, it is now followed by identification of the incoming
4355 connection (so you can more easily find what provoked it).
4357 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4359 5. Added some features to "harden" Exim a bit more against certain attacks:
4361 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4362 be put in Local/Makefile. This is like the never_users runtime option,
4363 but it cannot be overridden. The default setting is "root".
4365 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4366 prefix string with which any file named in a -C command line option
4369 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4370 is retained for -C and -D only if the caller of Exim is root. Without
4371 it, the exim user may also use -C and -D and retain privilege.
4373 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4374 command line option is disabled.
4376 6. Macro names set by the -D option must start with an upper case letter, just
4377 like macro names defined in the configuration file.
4379 7. Added "dereference=" facility to LDAP.
4381 8. Two instances of the typo "uknown" in the source files are fixed.
4383 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4384 the Configure-Makefile script screwed up while processing it.
4386 10. Incorporated PCRE 4.4.
4388 11. The SMTP synchronization check was not operating right at the start of an
4389 SMTP session. For example, it could not catch a HELO sent before the client
4390 waited for the greeting. There is now a check for outstanding input at the
4391 point when the greeting is written. Because of the duplex, asynchronous
4392 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4393 way, but not yet received, when the check is performed.
4395 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4396 on TCP/IP sockets, because this apparently causes some broken clients to
4399 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4400 unchanged) from the Cygwin maintainer.
4402 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4403 was supported instead of testing for mbx. Effectively a typo.
4405 15. The spa authenticator server code was not checking that the input it
4406 received was valid base64.
4408 16. The debug output line for the "set" modifier in ACLs was not showing the
4409 name of the variable that was being set.
4411 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4413 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4414 lookup on its own. Something else was needed to trigger the lookup. For
4415 example, a match in host_lookup or the need for a host name in a host list.
4416 Now, if $sender_host_name is referenced and the host name has not yet been
4417 looked up, a lookup is performed. If the lookup fails, the variable remains
4418 empty, and $host_lookup_failed is set to "1".
4420 19. Added "eqi" as a case-independent comparison operator.
4422 20. The saslauthd authentication condition could segfault if neither service
4423 nor realm was specified.
4425 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4426 error message that was logged was misleading, and incoming SMTP
4427 connections were dropped. The message is now more accurate, and temporary
4428 errors are given to SMTP connections.
4430 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4431 (except RSET) with a 421 error, until QUIT is received. However, it was
4432 failing to send a response to QUIT.
4434 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4435 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4436 ACL is now run after the helo_try_verify_hosts code.
4438 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4439 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4440 case-independent, so other case variants are also recognized). Apparently
4441 some systems use these upper case variants.
4443 25. If more than two messages were waiting for the same host, and a transport
4444 filter was specified for the transport, Exim sent two messages over the
4445 same TCP/IP connection, and then failed with "socket operation on non-
4446 socket" when it tried to send the third.
4448 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4450 27. The extern definition of crypt16() in expand.c was not being excluded when
4451 the OS had its own crypt16() function.
4453 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4454 as a preferred synonym for return_size_limit, both as an option and as an
4457 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4459 30. Changed the default configuration ACL to relax the local part checking rule
4460 for addresses that are not in any local domains. For these addresses,
4461 slashes and pipe symbols are allowed within local parts, but the sequence
4462 /../ is explicitly forbidden.
4464 31. SPA server authentication was not clearing the challenge buffer before
4467 32. log_message in a "warn" ACL statement was writing to the reject log as
4468 well as to the main log, which contradicts the documentation and doesn't
4469 seem right (because no rejection is happening). So I have stopped it.
4471 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4472 However, I am unable to do any testing of this.
4474 34. Fixed an infelicity in the appendfile transport. When checking directories
4475 for a mailbox, to see if any needed to be created, it was accidentally
4476 using path names with one or more superfluous leading slashes; tracing
4477 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4479 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4480 contents are added to the log line that is written for every discarded
4481 recipient. (Previously a log_message setting was ignored.)
4483 36. The ${quote: operator now quotes the string if it is empty.
4485 37. The install script runs exim in order to find its version number. If for
4486 some reason other than non-existence or emptiness, which it checks, it
4487 could not run './exim', it was installing it with an empty version number,
4488 i.e. as "exim-". This error state is now caught, and the installation is
4491 38. An argument was missing from the function that creates an error message
4492 when Exim fails to connect to the socket for saslauthd authentication.
4493 This could cause Exim to crash, or give a corrupted message.
4495 39. Added isip, isip4, and isip6 to ${if conditions.
4497 40. The ACL variables $acl_xx are now saved with the message, and can be
4498 accessed later in routers, transports, and filters.
4500 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4501 strings in the file are not string-expanded.
4503 42. If a MAIL command specified a SIZE value that was too large to fit into an
4504 int variable, the check against message_size_limit failed. Such values are
4505 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4506 day this will have to be increased, but I don't think I want to be around
4507 when emails are that large.
4514 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4515 iconv() is not standard in FreeBSD.
4517 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4518 IPv6 enabled. The observed symptom was a segmentation fault on return from
4519 the function os_common_find_running_interfaces() in src/os.c.
4521 3. In the check_special_case() function in daemon.c I had used "errno" as an
4522 argument name, which causes warnings on some systems. This was basically a
4523 typo, since it was named "eno" in the comments!
4525 4. The code that waits for the clock to tick (at a resolution of some fraction
4526 of a second) so as to ensure message-id uniqueness was always waiting for
4527 at least one whole tick, when it could have waited for less. [This is
4528 almost certainly not relevant at current processor speeds, where it is
4529 unlikely to ever wait at all. But we try to future-proof.]
4531 5. The function that sleeps for a time interval that includes fractions of a
4532 second contained a race. It did not block SIGALRM between setting the
4533 timer, and suspending (a couple of lines later). If the interval was short
4534 and the sigsuspend() was delayed until after it had expired, the suspension
4535 never ended. On busy systems this could lead to processes getting stuck for
4538 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4539 process, before it forks any delivery processes. The open lookup caching
4540 mechanism meant that the open file or database connection was passed into
4541 the delivery process. The problem was that delivery processes always tidy
4542 up cached lookup data. This could cause a problem for the next delivery
4543 process started by the queue runner, because the external queue runner
4544 process does not know about the closure. So the next delivery process
4545 still has data in the lookup cache. In the case of a file lookup, there was
4546 no problem because closing a file descriptor in a subprocess doesn't affect
4547 the parent. However, if the lookup was caching a connection to a database,
4548 the connection was closed, and the second delivery process was likely to
4549 see errors such as "PGSQL: query failed: server closed the connection
4550 unexpectedly". The problem has been fixed by closing all cached lookups
4551 in a queue runner before running a delivery process.
4553 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4554 seem to have the "const" qualifier which it has on other OS. I've
4557 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4558 *of the same type* (client or server) with the same public name that an
4559 error should be diagnosed.
4561 9. When Exim looked up a host name for an IP address, but failed to find the
4562 original IP address when looking up the host name (a safety check), it
4563 output the message "<ip address> does not match any IP for NULL", which was
4564 confusing, to say the least. The bug was that the host name should have
4565 appeared instead of "NULL".
4567 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4568 user that is built into the binary, and the -C or -D options is used, root
4569 privilege is dropped before the configuration file is read. In addition,
4570 logging is switched to stderr instead of the normal log files. If the
4571 configuration then re-defines the Exim user, the unprivileged environment
4572 is probably not what is expected, so Exim logs a panic warning message (but
4575 However, if deliver_drop_privilege is set, the unprivileged state may well
4576 be exactly what is intended, so the warning has been cut out in that case,
4577 and Exim is allowed to try to write to its normal log files.
4583 1. smtp_return_error_details was not giving details for temporary sender
4584 or receiver verification errors.
4586 2. Diagnose a configuration error if two authenticators have the same public
4589 3. Exim used not to create the message log file for a message until the first
4590 delivery attempt. This could be confusing when incoming messages were held
4591 for policy or load reasons. The message log file is now created at the time
4592 the message is received, and an initial "Received" line is written to it.
4594 4. The automatically generated man page for command line options had a minor
4595 bug that caused no ill effects; however, a more serious problem was that
4596 the procedure for building the man page automatically didn't always
4597 operate. Consequently, release 4.20 contains an out-of-date version. This
4598 shouldn't happen again.
4600 5. When building Exim with embedded Perl support, the script that builds the
4601 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4602 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4604 6. The freeze_tell option was not being used for messages that were frozen on
4605 arrival, either by an ACL or by local_scan().
4607 7. Added the smtp_incomplete_transaction log selector.
4609 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4610 was accepting AUTH without a new EHLO.
4612 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4613 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4616 10. Logging of TCP/IP connections (when configured) now happens in the main
4617 daemon process instead of the child process, so that the TCP/IP connection
4618 count is more accurate (but it can never be perfect).
4620 11. The use of "drop" in a nested ACL was not being handled correctly in the
4621 outer ACL. Now, if condition failure induced by the nested "drop" causes
4622 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4623 or "require"), the connection is dropped.
4625 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4626 that yield "discard" can now be used with an "accept" or a "discard" verb,
4627 but an error is generated for any others (because I can't see a useful way
4628 to define what should happen).
4630 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4631 are now processed in the same way as lines in the Exim configuration file.
4632 In particular, continuation lines are supported.
4634 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4636 15. Added -ti meaning -t -i.
4638 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4639 domains, and warn by logging if others are found.
4641 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4642 returned from GIFCONF when Exim is trying to find the list of interfaces on
4643 a host. The code in os.c has been modified to copy each ifreq to an aligned
4644 structure in all cases.
4646 Also, in some cases, the returned ifreq's were being copied to a 'struct
4647 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4648 means the last couple of bytes of an IPv6 address could be chopped if the
4649 ifreq contained only a normal sockaddr (14 bytes storage).
4651 18. Named domain lists were not supported in the hosts_treat_as_local option.
4652 An entry such as +xxxx was not recognized, and was treated as a literal
4655 19. Ensure that header lines added by a DATA ACL are included in the reject log
4656 if the ACL subsequently rejects the message.
4658 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4659 MD5 (which is deprecated).
4661 21. When testing a filter file using -bf, Exim was writing a message when it
4662 took the sender from a "From " line in the message, but it was not doing so
4663 when it took $return_path from a Return-Path: header line. It now does.
4665 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4666 with a valid header line field name (a series of printing characters
4667 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4669 23. Changed "disc" in the source to "disk" to conform to the documentation and
4670 the book and for uniformity.
4672 24. Ignore Sendmail's -Ooption=value command line item.
4674 25. When execve() failed while trying to run a command in a pipe transport,
4675 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4676 could be confused with a return value of 69 from the command itself. This
4677 has been changed to 127, the value the shell returns if it is asked to run
4678 a non-existent command. The wording for the related log line suggests a
4679 non-existent command as the problem.
4681 26. If received_header_text expands to an empty string, do not add a Received:
4682 header line to the message. (Well, it adds a token one on the spool, but
4683 marks it "old" so that it doesn't get used or transmitted.)
4685 27. Installed eximstats 1.28 (addition of -nt option).
4687 28. There was no check for failure on the call to getsockname() in the daemon
4688 code. This can fail if there is a shortage of resources on the system, with
4689 ENOMEM, for example. A temporary error is now given on failure.
4691 29. Contrary to the C standard, it seems that in some environments, the
4692 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4693 program. Exim now does this explicitly; it affects the formatting of
4694 timestamps using strftime().
4696 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4697 complaints. I've now initialized all the variables, to avoid this.
4699 32. Header lines added by a system filter were not being "seen" during
4700 transport-time rewrites.
4702 33. The info_callback() function passed to OpenSSL is set up with type void
4703 (*)(SSL *, int, int), as described somewhere. However, when calling the
4704 function (actually a macro) that sets it up, the type void(*)() is
4705 expected. I've put in a cast to prevent warnings from picky compilers.
4707 34. If a DNS black list lookup found a CNAME record, but there were no A
4708 records associated with the domain it pointed at, Exim crashed.
4710 35. If a DNS black list lookup returned more than one A record, Exim ignored
4711 all but the first. It now scans all returned addresses if a particular IP
4712 value is being sought. In this situation, the contents of the
4713 $dnslist_value variable are a list of all the addresses, separated by a
4716 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4717 to accept a host name and all its aliases if the forward lookup for any of
4718 them yielded the IP address of the incoming connection. Now it accepts only
4719 those names whose forward lookup yields the correct IP address. Any other
4720 names are discarded. This closes a loophole whereby a rogue DNS
4721 administrator could create reverse DNS records to break through a
4722 wildcarded host restriction in an ACL.
4724 37. If a user filter or a system filter that ran in a subprocess used any of
4725 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4726 the wrong values were passed to the pipe command ($thisaddress had the
4727 value of $0, $0 had the value of $1, etc). This bug was introduced by
4728 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4730 38. Improved the line breaking for long SMTP error messages from ACLs.
4731 Previously, if there was no break point between 40 and 75 characters, Exim
4732 left the rest of the message alone. Two changes have been made: (a) I've
4733 reduced the minimum length to 35 characters; (b) if it can't find a break
4734 point between 35 and 75 characters, it looks ahead and uses the first one
4735 that it finds. This may give the occasional overlong line, but at least the
4736 remaining text gets split now.
4738 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4739 file descriptors might be low, and that setting 1000 would always raise it.
4740 It turns out that in some environments, the limit is already over 1000 and
4741 that lowering it causes trouble. So now Exim takes care not to decrease it.
4743 40. When delivering a message, the value of $return_path is set to $sender_
4744 address at the start of routing (routers may change the value). By an
4745 oversight, this default was not being set up when an address was tested by
4746 -bt or -bv, which affected the outcome if any router or filter referred to
4749 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4750 log only when "message" or "log_message" are set. However, if one of the
4751 conditions was an address verification, or a call to a nested ACL, the
4752 messages generated by the underlying test were being passed through. This
4753 no longer happens. The underlying message is available in $acl_verify_
4754 message for both "message" and "log_message" expansions, so it can be
4755 passed through if needed.
4757 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4758 new expansion $bh_ to give the encoded byte string without charset
4759 translation. Translation happens only if iconv() is available; HAVE_ICONV
4760 indicates this at build time. HEADERS_CHARSET gives the charset to
4761 translate to; headers_charset can change it in the configuration, and
4762 "headers charset" can change it in an individual filter file.
4764 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4765 that creates RFC 2047 encoded "words" labels them as that charset instead
4766 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4767 expansion operator; (ii) when Exim creates a From: line for a local
4768 message; (iii) when a header line is rewritten to include a "phrase" part.
4770 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4771 buggy, causing it to skip the first lines of messages whose message ID
4772 ended in 'D'. This would not have bitten before Exim release 4.14, because
4773 message IDs were unlikely to end in 'D' before then. The effect was to have
4774 incorrect size information for certain domains.
4776 45. #include "config.h" was missing at the start of the crypt16.c module. This
4777 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4780 46. If there was a timeout during a "random" callout check, Exim treated it as
4781 a failure of the random address, and carried on sending RSET and the real
4782 address. If the delay was just some slowness somewhere, the response to the
4783 original RCPT would be taken as a response to RSET and so on, causing
4784 mayhem of various kinds.
4786 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4787 when I implemented it. It didn't allow for the fact that some option values
4788 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4789 the right test for positive values.
4791 48. Domain names in DNS records are case-independent. Exim always looks them up
4792 in lower case. Some resolvers return domain names in exactly the case they
4793 appear in the zone file, that is, they may contain uppercase letters. Not
4794 all resolvers do this - some return always lower case. Exim was treating a
4795 change of case by a resolver as a change of domain, similar to a widening
4796 of a domain abbreviation. This triggered its re-routing code and so it was
4797 trying to route what was effectively the same domain again. This normally
4798 caused routing to fail (because the router wouldn't handle the domain
4799 twice). Now Exim checks for this case specially, and just changes the
4800 casing of the domain that it ultimately uses when it transmits the message
4803 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4806 50. If a filter generated a file delivery with a non-absolute name (possible if
4807 no home directory exists for the router), the forbid_file option was not
4810 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4811 the existing equality matching.
4813 52. Exim was using ints instead of ino_t variables in some places where it was
4814 dealing with inode numbers.
4816 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4817 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4818 called TMPDIR, and if it finds it is different, it changes its value.
4820 54. The smtp_printf() function is now made available to local_scan() so
4821 additional output lines can be written before returning. There is also an
4822 smtp_fflush() function to enable the detection of a dropped connection.
4823 The variables smtp_input and smtp_batched_input are exported to
4826 55. Changed the default runtime configuration: the message "Unknown user"
4827 has been removed from the ACL, and instead placed on the localuser router,
4828 using the cannot_route_message feature. This means that any verification
4829 failures that generate their own messages won't get overridden. Similarly,
4830 the "Unrouteable address" message that was in the ACL for unverifiable
4831 relay addresses has also been removed.
4833 56. Added hosts_avoid_esmtp to the smtp transport.
4835 57. The exicyclog script was not checking for the esoteric option
4836 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4837 will work only if exicyclog is run under the appropriate euid.
4839 58. Following a discussion on the list, the rules by which Exim recognises line
4840 endings on incoming messages have been changed. The -dropcr and drop_cr
4841 options are now no-ops, retained only for backwards compatibility. The
4842 following line terminators are recognized: LF CRLF CR. However, special
4843 processing applies to CR:
4845 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4846 nor a local message in the state where . is a terminator.
4848 (ii) If a bare CR is encountered in a header line, an extra space is added
4849 after the line terminator so as not to end the header. The reasoning
4850 behind this is that bare CRs in header lines are most likely either
4851 to be mistakes, or people trying to play silly games.
4853 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4854 was being incorrectly given as 18 bytes larger than it should have been.
4855 This is a VOB (very old bug).
4857 60. This may never have affected anything current, but just in case it has:
4858 When the local host is found other than at the start of a list of hosts,
4859 the local host, those with the same MX, and any that follow, are discarded.
4860 When the list in question was part of a longer list of hosts, the following
4861 hosts (not currently being processed) were also being discarded. This no
4862 longer happens. I'm not sure if this situation could ever has previously
4865 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4868 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4869 header. This is something that is recommended in a new Internet Draft, and
4870 is something that is documented as being done by Sendmail. There are two
4871 possible values. For messages generated by the autoreply transport, Exim
4874 Auto-Submitted: auto-replied
4876 whereas for all other generated messages (e.g. bounces) it adds
4878 Auto-Submitted: auto-generated
4880 63. The "personal" condition in filters now includes a test for the
4881 Auto-Submitted: header. If it contains the string "auto-" the message it
4882 not considered personal.
4884 64. Added rcpt_include_affixes as a generic transport option.
4886 65. Added queue_only_override (default true).
4888 66. Added the syslog_duplication option.
4890 67. If what should have been the first header line of a message consisted of
4891 a space followed by a colon, Exim was mis-interpreting it as a header line.
4892 It isn't of course - it is syntactically invalid and should therefore be
4893 treated as the start of the message body. The misbehaviour could have
4894 caused a number of strange effects, including loss of data in subsequent
4895 header lines, and spool format errors.
4897 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4898 client host had authenticated. This control can now be exercised by an ACL
4899 for more flexibility.
4901 69. By default, callouts do not happen when testing with -bh. There is now a
4902 variant, -bhc, which does actually run the callout code, including
4903 consulting and updating the callout cache.
4905 70. Added support for saslauthd authentication, courtesy of Alexander
4908 71. If statvfs() failed on the spool or log directories while checking their
4909 size for availability, Exim confusingly gave the error "space shortage".
4910 Furthermore, in debugging mode it crashed with a floating point exception.
4911 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4912 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4913 serious problem, Exim now writes to the main and panic logs when this
4914 happens, with details of the failure. It then refuses to accept the
4915 incoming message, giving the message "spool directory problem" or "log
4916 directory problem" with a 421 code for SMTP messages.
4918 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4919 0, 1, and 2 exist, because some OS complain for execs without them (see
4920 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4921 descriptors. However, the code omitted to check that the open succeeded,
4922 causing mysterious errors if for some reason the permissions on /dev/null
4923 got screwed. Now Exim writes a message to the main and panic logs, and
4924 bombs out if it can't open /dev/null.
4926 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4927 interact so that it is all more flexible. It is supposed to remain
4928 backwards compatible. Also added extra_local_interfaces.
4930 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4931 to bomb out with an assertion failure - to the client this appears as a
4932 connection drop. This problem occurs in the part of the code that was taken
4933 from the Samba project. Fortunately, the assertion is in a very simple
4934 function, so I have fixed this by reproducing the function inline in the
4935 one place where it is called, and arranging for authentication to fail
4936 instead of killing the process with assert().
4938 75. The SPA client code was not working when the server requested OEM rather
4939 than Unicode encoding.
4941 76. Added code to make require_files with a specific uid setting more usable in
4942 the case where statting the file as root fails - usually a non-root-mounted
4943 NFS file system. When this happens and the failure is EACCES, Exim now
4944 forks a subprocess and does the per-uid checking as the relevant uid.
4946 77. Added process_log_path.
4948 78. If log_file_path was not explicitly set, a setting of check_log_space or
4949 check_log_inodes was ignored.
4951 79. If a space check for the spool or log partitions fails, the incident is now
4952 logged. Of course, in the latter case the data may get lost...
4954 80. Added the %p formatting code to string_format() so that it can be used to
4955 print addresses in debug_print(). Adjusted all the address printing in the
4956 debugging in store.c to use %p rather than %d.
4958 81. There was a concern that a line of code in smtp_in.c could overflow a
4959 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4960 initially expressed, the concern was not well-founded, because trailing
4961 spaces are removed early. However, if the trailing spaces were followed by
4962 a NULL, they did not get removed, so the overflow was possible. Two fixes
4965 (a) I re-wrote the offending code in a cleaner fashion.
4966 (b) If an incoming SMTP command contains a NULL character, it is rejected
4969 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4970 runs initgroups(), so that if the Exim user is in any additional groups,
4971 they will be used during message reception.
4977 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob