1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.574 2009/10/19 11:41:27 nm4 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TK/01 Added patch by Johannes Berg that expands the main option
10 "spamd_servers" if it starts with a dollar sign.
12 TK/02 Write list of recipients to X-Envelope-Sender header when building
13 the mbox-format spool file for content scanning (suggested by Jakob
16 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
17 (http://www.dcc-servers.net/) support via dccifd. Activated by
18 setting EXPERIMENTAL_DCC=yes in Local/Makefile. Check out
19 experimental_spec.txt for more documentation.
21 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
22 by Mark Daniel Reidel <mr@df.eu>.
24 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
25 When building exim an external PCRE library is now needed -
26 PCRE is a system library on the majority of modern systems.
27 See entry on PCRE_LIBS in EDITME file.
29 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
30 conversation. Added nologin parameter to request.
31 Patch contributed by Kirill Miazine
33 TF/01 Do not log submission mode rewrites if they do not change the address.
35 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
37 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
38 log files in place. Contributed by Roberto Lima
40 NM/04 Bugzilla 667: close socket used by dovecot authenticator
42 TF/03 Bugzilla 615: When checking the local_parts router precondition
43 after a local_part_suffix or local_part_prefix option, Exim now
44 does not use the address's named list lookup cache, since this
45 contains cached lookups for the whole local part.
47 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
48 Robert Millan. Documentation is in experimental-spec.txt
50 TF/04 Bugzilla 668: Fix parallel build (make -j).
52 NM/05 Bugzilla 437: Prevent Maildix aux files being created with mode 000
54 NM/05 Bugzilla 598: Improvement to Dovecot authenticator handling.
55 Patch provided by Jan Srzednicki
57 TF/05 Leading white space used to be stripped from $spam_report which
58 wrecked the formatting. Now it is preserved.
60 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
61 that they are available at delivery time.
63 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
65 TF/08 TLS error reporting now respects the incoming_interface and
66 incoming_port log selectors.
68 TF/09 Produce a more useful error message if an SMTP transport's hosts
69 setting expands to an empty string.
71 NM/06 Bugzilla 744: EXPN did not work under TLS.
72 Patch provided by Phil Pennock
74 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
75 Patch provided by Richard Godbee
77 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
78 acl_smtp_notquit, added index entry.
80 NM/09 Bugzilla 787: Potential buffer overflow in string_format
81 Patch provided by Eugene Bujak
83 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to accept()
84 Patch provided by Maxim Dounin
86 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
87 Patch provided by Phil Pennock
89 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
91 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
92 Patch provided by Brad "anomie" Jorsch
94 NM/14 Bugzilla 622: Added timeout setting to transport filter.
95 Patch provided by Dean Brooks
97 TK/05 Add native DKIM support (does not depend on external libraries).
99 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
100 Patch provided by Graeme Fowler
102 NM/16 Bugzilla 851: Documentation example syntax fix.
104 NM/17 Changed NOTICE file to remove references to embedded PCRE.
106 NM/18 Bugzilla 894: Fix issue with very long lines including comments in lsearch
108 NM/19 Bugzilla 745: TLS version reporting
109 Patch provided by Phil Pennock
111 NM/20 Bugzilla 167: bool: condition support
112 Patch provided by Phil Pennock
114 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken clients
115 Patch provided by Phil Pennock
117 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date
118 Patch provided by Brad "anomie" Jorsch
120 NM/23 Bugzilla 687: Fix misparses in eximstats
121 Patch provided by Heiko Schlittermann
127 TK/01 Add preliminary DKIM support. Currently requires a forked version of
128 ALT-N's libdkim that I have put here:
129 http://duncanthrax.net/exim-experimental/
131 Note to Michael Haardt: I had to rename some vars in sieve.c. They
132 were called 'true' and it seems that C99 defines that as a reserved
133 keyword to be used with 'bool' variable types. That means you could
134 not include C99-style headers which use bools without triggering
135 build errors in sieve.c.
137 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
138 as mailq or other aliases. Changed the --help handling significantly
139 to do whats expected. exim_usage() emits usage/help information.
141 SC/01 Added the -bylocaldomain option to eximstats.
143 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
145 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
147 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
153 PH/01 Another patch from the Sieve maintainer.
155 PH/02 When an IPv6 address is converted to a string for single-key lookup
156 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
157 dots are used instead of colons so that keys in lsearch files need not
158 contain colons. This was done some time before quoting was made available
159 in lsearch files. However, iplsearch files do require colons in IPv6 keys
160 (notated using the quote facility) so as to distinguish them from IPv4
161 keys. This meant that lookups for IP addresses in host lists did not work
162 for iplsearch lookups.
164 This has been fixed by arranging for IPv6 addresses to be expressed with
165 colons if the lookup type is iplsearch. This is not incompatible, because
166 previously such lookups could never work.
168 The situation is now rather anomolous, since one *can* have colons in
169 ordinary lsearch keys. However, making the change in all cases is
170 incompatible and would probably break a number of configurations.
172 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
175 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
176 conversion specification without a maximum field width, thereby enabling
177 a rogue spamd server to cause a buffer overflow. While nobody in their
178 right mind would setup Exim to query an untrusted spamd server, an
179 attacker that gains access to a server running spamd could potentially
180 exploit this vulnerability to run arbitrary code as the Exim user.
182 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
183 $primary_hostname instead of what libspf2 thinks the hosts name is.
185 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
186 a directory entry by the name of the lookup key. Previously, if a
187 symlink pointed to a non-existing file or a file in a directory that
188 Exim lacked permissions to read, a lookup for a key matching that
189 symlink would fail. Now it is enough that a matching directory entry
190 exists, symlink or not. (Bugzilla 503.)
192 PH/03 The body_linecount and body_zerocount variables are now exported in the
195 PH/04 Added the $dnslist_matched variable.
197 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
198 This means they are set thereafter only if the connection becomes
201 PH/06 Added the client_condition to authenticators so that some can be skipped
202 by clients under certain conditions.
204 PH/07 The error message for a badly-placed control=no_multiline_responses left
205 "_responses" off the end of the name.
207 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
209 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
210 (without spaces) instead of just copying the configuration text.
212 PH/10 Added the /noupdate option to the ratelimit ACL condition.
214 PH/11 Added $max_received_linelength.
216 PH/12 Added +ignore_defer and +include_defer to host lists.
218 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
219 way in which PCRE > 7.0 is built.
221 PH/14 Implemented queue_only_load_latch.
223 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
224 MAIL command. The effect was to mangle the value on 64-bit systems.
226 PH/16 Another patch from the Sieve maintainer.
228 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
230 PH/18 If a system quota error occurred while trying to create the file for
231 a maildir delivery, the message "Mailbox is full" was not appended to the
232 bounce if the delivery eventually timed out. Change 4.67/27 below applied
233 only to a quota excession during the actual writing of the file.
235 PH/19 It seems that peer DN values may contain newlines (and other non-printing
236 characters?) which causes problems in log lines. The DN values are now
237 passed through string_printing() before being added to log lines.
239 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
240 and InterBase are left for another time.)
242 PH/21 Added message_body_newlines option.
244 PH/22 Guard against possible overflow in moan_check_errorcopy().
246 PH/23 POSIX allows open() to be a macro; guard against that.
248 PH/24 If the recipient of an error message contained an @ in the local part
249 (suitably quoted, of course), incorrect values were put in $domain and
250 $local_part during the evaluation of errors_copy.
256 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
257 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
260 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
261 issue a MAIL command.
263 PH/02 In an ACL statement such as
265 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
267 if a client was not listed at all, or was listed with a value other than
268 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
269 the condition was not true (as it should be), so access was not denied.
270 The bug was that the ! inversion was incorrectly passed on to the second
271 item. This has been fixed.
273 PH/03 Added additional dnslists conditions == and =& which are different from
274 = and & when the dns lookup returns more than one IP address.
276 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
277 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
279 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
280 FSYNC, which compiles an option called disable_fsync that allows for
281 bypassing fsync(). The documentation is heavily laced with warnings.
283 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
285 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
286 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
287 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
288 including adding "make clean"; (3) Added -fPIC when compiling the test
289 dynamically loaded module, to get rid of a warning.
291 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
292 message fails, move_frozen_messages = true and ignore_bounce_errors_after
293 = 0s. The bug is otherwise harmless.
295 PH/07 There was a bug in the dovecot authenticator such that the value of
296 $auth1 could be overwritten, and so not correctly preserved, after a
297 successful authentication. This usually meant that the value preserved by
298 the server_setid option was incorrect.
300 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
302 PH/09 Installed PCRE release 7.0.
304 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
305 run for batched SMTP input. It is now run at the start of every message
306 in the batch. While fixing this I discovered that the process information
307 (output by running exiwhat) was not always getting set for -bs and -bS
308 input. This is fixed, and it now also says "batched" for BSMTP.
310 PH/11 Added control=no_pipelining.
312 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
313 patch, slightly modified), and move the expansion of helo_data till after
314 the connection is made in the smtp transport (so it can use these
317 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
319 PH/14 Added log_selector = +pid.
321 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
323 PH/16 Add ${if forany and ${if forall.
325 PH/17 Added dsn_from option to vary the From: line in DSNs.
327 PH/18 Flush SMTP output before performing a callout, unless control =
328 no_callout_flush is set.
330 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
331 was true (the default) a successful delivery failed to delete the retry
332 item, thus causing premature timeout of the address. The bug is now
335 PH/20 Added hosts_avoid_pipelining to the smtp transport.
337 PH/21 Long custom messages for fakedefer and fakereject are now split up
338 into multiline reponses in the same way that messages for "deny" and
339 other ACL rejections are.
341 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
342 with slight modification.
344 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
345 draft, changing the syntax and factoring some duplicate code".
347 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
348 for deliveries of the second and subsequent messages over the same SMTP
351 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
352 ${reduce, with only minor "tidies".
354 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
356 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
357 expansion side effects.
359 PH/27 When a message times out after an over-quota error from an Exim-imposed
360 quota, the bounce message says "mailbox is full". This message was not
361 being given when it was a system quota that was exceeded. It now should
364 MH/03 Made $recipients available in local_scan(). local_scan() already has
365 better access to the recipient list through recipients_list[], but
366 $recipients can be useful in postmaster-provided expansion strings.
368 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
369 in the case of a MAIL command with additional options following the
370 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
371 were accidentally chopped off.
373 PH/29 SMTP synchronization checks are implemented when a command is read -
374 there is a check that no more input is waiting when there shouldn't be
375 any. However, for some commands, a delay in an ACL can mean that it is
376 some time before the response is written. In this time, more input might
377 arrive, invalidly. So now there are extra checks after an ACL has run for
378 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
379 pipelining has not been advertised.
381 PH/30 MH's patch to allow iscntrl() characters to be list separators.
383 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
384 returned in the SMTP response when smtp_return_error_details was false.
387 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
388 instead of the C I/O that was originally supplied, because problems were
391 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
392 Exim which did not show up earlier: it was assuming that a call to
393 SSL_CTX_set_info_callback() might give an error value. In fact, there is
394 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
395 was a macro that became an assignment, so it seemed to work. This has
396 changed to a proper function call with a void return, hence the compile
397 error. Exim's code has been fixed.
399 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
402 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
404 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
406 PH/37 If a message is not accepted after it has had an id assigned (e.g.
407 because it turns out to be too big or there is a timeout) there is no
408 "Completed" line in the log. When some messages of this type were
409 selected by exigrep, they were listed as "not completed". Others were
410 picked up by some special patterns. I have improved the selection
411 criteria to be more general.
413 PH/38 The host_find_failed option in the manualroute router can now be set
414 to "ignore", to completely ignore a host whose IP address cannot be
415 found. If all hosts are ignored, the behaviour is controlled by the new
416 host_all_ignored option.
418 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
419 homing or because of multiple MX records with /mx) generated more than
420 one IP address, and the following item turned out to be the local host,
421 all the secondary addresses of the first item were incorrectly removed
422 from the list, along with the local host and any following hosts (which
423 is what is supposed to happen).
425 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
426 whoever called Exim into the -H file. In the case of the daemon it was
427 behaving confusingly. When first started, it used values for whoever
428 started the daemon, but after a SIGHUP it used the Exim user (because it
429 calls itself on a restart). I have changed the code so that it now always
432 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
433 message are rejected with the same error (e.g. no authentication or bad
434 sender address), and a DATA command is nevertheless sent (as can happen
435 with PIPELINING or a stupid MUA), the error message that was given to the
436 RCPT commands is included in the rejection of the DATA command. This is
437 intended to be helpful for MUAs that show only the final error to their
440 PH/42 Another patch from the Sieve maintainer.
442 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
443 Eximstats - Fixed some broken HTML links and added missing column headers
445 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
446 columns for Rejects, Temp Rejects, Ham, and Spam rows.
448 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
450 PH/43 Yet another patch from the Sieve maintainer.
452 PH/44 I found a way to check for a TCP/IP connection going away before sending
453 the response to the final '.' that terminates a message, but only in the
454 case where the client has not sent further data following the '.'
455 (unfortunately, this is allowed). However, in many cases there won't be
456 any further data because there won't be any more messages to send. A call
457 to select() can be used: if it shows that the input is "ready", there is
458 either input waiting, or the socket has been closed. An attempt to read
459 the next input character can distinguish the two cases. Previously, Exim
460 would have sent an OK response which the client would never have see.
461 This could lead to message repetition. This fix should cure that, at
462 least in a lot of common cases.
464 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
465 advertised in response to EHLO.
471 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
472 fixed by 4.65/MH/01 (is this a record?) are fixed:
474 (i) An empty string was always treated as zero by the numeric comparison
475 operators. This behaviour has been restored.
477 (ii) It is documented that the numeric comparison operators always treat
478 their arguments as decimal numbers. This was broken in that numbers
479 starting with 0 were being interpreted as octal.
481 While fixing these problems I realized that there was another issue that
482 hadn't been noticed. Values of message_size_limit (both the global option
483 and the transport option) were treated as octal if they started with 0.
484 The documentation was vague. These values are now always treated as
485 decimal, and I will make that clear in the documentation.
491 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
492 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
495 MH/01 Don't check that the operands of numeric comparison operators are
496 integers when their expansion is in "skipping" mode (fixes bug
497 introduced by 4.64-PH/07).
499 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
500 child addresses, Exim now panics and dies. Previously, because the count
501 is held in a short int, deliveries were likely to be lost. As such a
502 large number of recipients for a single message is ridiculous
503 (performance will be very, very poor), I have chosen to impose a limit
504 rather than extend the field.
510 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
511 leftover -K file (the existence of which was triggered by #402).
512 While we were at it, introduced process PID as part of the -K
513 filename. This should rule out race conditions when creating
516 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
517 processing considerably. Previous code took too long for large mails,
518 triggering a timeout which in turn triggers #401.
520 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
521 in the DK code in transports.c. sendfile() is not really portable,
522 hence the _LINUX specificness.
524 TF/01 In the add_headers option to the mail command in an Exim filter,
525 there was a bug that Exim would claim a syntax error in any
526 header after the first one which had an odd number of characters
529 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
530 callout verification, Exim cached a "reject" for the entire domain. This
531 is correct for most verifications, but it is not correct for a recipient
532 verification with use_sender or use_postmaster set, because in that case
533 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
534 case of MAIL FROM:<> rejection from other early rejections (e.g.
535 rejection of HELO). When verifying a recipient using a non-null MAIL
536 address, the cache is ignored if it shows MAIL FROM:<> rejection.
537 Whatever the result of the callout, the value of the domain cache is
538 left unchanged (for any other kind of callout, getting as far as trying
539 RCPT means that the domain itself is ok).
541 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
544 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
545 manifest itself as EPIPE rather than ECONNECT. When tidying away a
546 session, the daemon ignores ECONNECT errors and logs others; it now
547 ignores EPIPE as well.
549 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
550 (quoted-printable decoding).
552 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
553 later the small subsequent patch to fix an introduced bug.
555 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
557 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
559 PH/08 An error is now given if message_size_limit is specified negative.
561 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
562 to be given (somewhat) arbitrary names.
564 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
567 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
568 miscellaneous code fixes
570 PH/10 Added the log_reject_target ACL modifier to specify where to log
573 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
574 hostname. This is wrong, because it relates to the incoming message (and
575 probably the interface on which it is arriving) and not to the outgoing
576 callout (which could be using a different interface). This has been
577 changed to use the value of the helo_data option from the smtp transport
578 instead - this is what is used when a message is actually being sent. If
579 there is no remote transport (possible with a router that sets up host
580 addresses), $smtp_active_hostname is used.
582 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
583 tweaks were necessary in order to get it to work (see also 21 below):
584 (a) The code assumed that strncpy() returns a negative number on buffer
585 overflow, which isn't the case. Replaced with Exim's string_format()
587 (b) There were several signed/unsigned issues. I just did the minimum
588 hacking in of casts. There is scope for a larger refactoring.
589 (c) The code used strcasecmp() which is not a standard C function.
590 Replaced with Exim's strcmpic() function.
591 (d) The code set only $1; it now sets $auth1 as well.
592 (e) A simple test gave the error "authentication client didn't specify
593 service in request". It would seem that Dovecot has changed its
594 interface. Fortunately there's a specification; I followed it and
595 changed what the client sends and it appears to be working now.
597 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
600 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
601 address A is aliased to B and C, where B exists and C does not. Without
602 -v the output is "A verified" because verification stops after a
603 successful redirection if more than one address is generated. However,
604 with -v the child addresses are also verified. Exim was outputting "A
605 failed to verify" and then showing the successful verification for C,
606 with its parentage. It now outputs "B failed to verify", showing B's
607 parentage before showing the successful verification of C.
609 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
610 look up a TXT record in a specific list after matching in a combined
613 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
614 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
615 they consult the DNS. I had assumed they would set it the way they
616 wanted; and indeed my experiments on Linux seem to show that in some
617 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
618 To be on the safe side, however, I have now made the interface to
619 host_find_byname() similar to host_find_bydns(), with an argument
620 containing the DNS resolver options. The host_find_byname() function now
621 sets these options at its start, just as host_find_bydns() does. The smtp
622 transport options dns_qualify_single and dns_search_parents are passed to
623 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
624 of host_find_byname() use the default settings of RES_DEFNAMES
625 (qualify_single) but not RES_DNSRCH (search_parents).
627 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
628 spool_read_header() do less string testing, by means of a preliminary
629 switch on the second character of optional "-foo" lines. (This is
630 overdue, caused by the large number of possibilities that now exist.
631 Originally there were few.) While I was there, I also converted the
632 str(n)cmp tests so they don't re-test the leading "-" and the first
633 character, in the hope this might squeeze out yet more improvement.
635 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
636 flag allowing group syntax was set by the header_syntax check but not
637 turned off, possible causing trouble later; (2) The flag was not being
638 set at all for the header_verify test, causing "group"-style headers to
639 be rejected. I have now set it in this case, and also caused header_
640 verify to ignore an empty address taken from a group. While doing this, I
641 came across some other cases where the code for allowing group syntax
642 while scanning a header line wasn't quite right (mostly, not resetting
643 the flag correctly in the right place). These bugs could have caused
644 trouble for malformed header lines. I hope it is now all correct.
646 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
647 with the "reply" argument non-NULL. The code, however (which originally
648 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
649 but it didn't always do it. This confused somebody who was copying the
650 code for some other use. I have removed all the tests.
652 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
653 feature that was used to support insecure browsers during the U.S. crypto
654 embargo. It requires special client support, and Exim is probably the
655 only MTA that supported it -- and would never use it because real RSA is
656 always available. This code has been removed, because it had the bad
657 effect of slowing Exim down by computing (never used) parameters for the
658 RSA_EXPORT functionality.
660 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
661 authenticator to fail if there's a tab character in the incoming data
662 (there should never be unless someone is messing about, as it's supposed
663 to be base64-encoded). Also added, on Timo's advice, the "secured" option
664 if the connection is using TLS or if the remote IP is the same as the
665 local IP, and the "valid-client-cert option" if a client certificate has
668 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
669 authenticators. This can be used for authorization after authentication
670 succeeds. (In the case of plaintext, it servers for both authentication
673 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
674 if any retry times were supplied.
676 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
677 connection, where there is no client IP address to check. In this
678 situation, the verify now always succeeds.
680 PH/25 Applied John Jetmore's -Mset patch.
682 PH/26 Added -bem to be like -Mset, but loading a message from a file.
684 PH/27 In a string expansion for a processed (not raw) header when multiple
685 headers of the same name were present, leading whitespace was being
686 removed from all of them, but trailing whitespace was being removed only
687 from the last one. Now trailing whitespace is removed from each header
688 before concatenation. Completely empty headers in a concatenation (as
691 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
692 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
694 PH/29 [Removed. This was a change that I later backed out, and forgot to
695 correct the ChangeLog entry (that I had efficiently created) before
696 committing the later change.]
698 PH/30 Exim was sometimes attempting to deliver messages that had suffered
699 address errors (4xx response to RCPT) over the same connection as other
700 messages routed to the same hosts. Such deliveries are always "forced",
701 so retry times are not inspected. This resulted in far too many retries
702 for the affected addresses. The effect occurred only when there were more
703 hosts than the hosts_max_try setting in the smtp transport when it had
704 the 4xx errors. Those hosts that it had tried were not added to the list
705 of hosts for which the message was waiting, so if all were tried, there
706 was no problem. Two fixes have been applied:
708 (i) If there are any address or message errors in an SMTP delivery, none
709 of the hosts (tried or untried) are now added to the list of hosts
710 for which the message is waiting, so the message should not be a
711 candidate for sending over the same connection that was used for a
712 successful delivery of some other message. This seems entirely
713 reasonable: after all the message is NOT "waiting for some host".
714 This is so "obvious" that I'm not sure why it wasn't done
715 previously. Hope I haven't missed anything, but it can't do any
716 harm, as the worst effect is to miss an optimization.
718 (ii) If, despite (i), such a delivery is accidentally attempted, the
719 routing retry time is respected, so at least it doesn't keep
720 hammering the server.
722 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
723 in ${readsocket because some servers need this prod.
725 PH/32 Added some extra debug output when updating a wait-xxx database.
727 PH/33 The hint "could be header name not terminated by colon", which has been
728 given for certain expansion errors for a long time, was not being given
729 for the ${if def:h_colon_omitted{... case.
731 PH/34 The spec says: "With one important exception, whenever a domain list is
732 being scanned, $domain contains the subject domain." There was at least
733 one case where this was not true.
735 PH/35 The error "getsockname() failed: connection reset by peer" was being
736 written to the panic log as well as the main log, but it isn't really
737 panic-worthy as it just means the connection died rather early on. I have
738 removed the panic log writing for the ECONNRESET error when getsockname()
741 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
742 runs only) independently of the message's sender address. This meant
743 that, if the 4xx error was in fact related to the sender, a different
744 message to the same recipient with a different sender could confuse
745 things. In particualar, this can happen when sending to a greylisting
746 server, but other circumstances could also provoke similar problems.
747 I have changed the default so that the retry time for these errors is now
748 based a combination of the sender and recipient addresses. This change
749 can be overridden by setting address_retry_include_sender=false in the
752 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
753 remote server are returned as part of bounce messages. This was not
754 happening for LMTP over a pipe (the lmtp transport), but now it is the
755 same for both kinds of LMTP.
757 PH/38 Despite being documented as not happening, Exim was rewriting addresses
758 in header lines that were in fact CNAMEs. This is no longer the case.
760 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
761 and queue runs started by the daemon processed all messages. This has
762 been fixed so that -R and -S can now usefully be given with -q<time>.
764 PH/40 Import PCRE release 6.7 (fixes some bugs).
766 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
768 PH/42 Give an error if -q is specified more than once.
770 PH/43 Renamed the variables $interface_address and $interface_port as
771 $received_ip_address and $received_port, to make it clear that these
772 values apply to message reception, and not to the outgoing interface when
773 a message is delivered. (The old names remain recognized, of course.)
775 PH/44 There was no timeout on the connect() call when using a Unix domain
776 socket in the ${readsocket expansion. There now is.
778 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
779 be meaningful with "accept".
781 SC/01 Eximstats V1.43
782 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
784 SC/02 Eximstats V1.44
785 Use a glob alias rather than an array ref in the generated
786 parser. This improves both readability and performance.
788 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
789 Collect SpamAssassin and rejection statistics.
790 Don't display local sender or destination tables unless
791 there is data to show.
792 Added average volumes into the top table text output.
794 SC/04 Eximstats V1.46
795 Collect data on the number of addresses (recipients)
796 as well as the number of messages.
798 SC/05 Eximstats V1.47
799 Added 'Message too big' to the list of mail rejection
800 reasons (thanks to Marco Gaiarin).
802 SC/06 Eximstats V1.48
803 Mainlog lines which have GMT offsets and are too short to
804 have a flag are now skipped.
806 SC/07 Eximstats V1.49 (Alain Williams)
807 Added the -emptyok flag.
809 SC/08 Eximstats V1.50
810 Fixes for obtaining the IP address from reject messages.
812 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
813 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
814 whitesspace changes from 4.64-PH/27
816 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
819 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
820 are found, allow negative numbers in numeric criteria)
822 JJ/06 exipick.20061117.2, added new $message_body_missing variable
824 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
825 to match changes made in 4.64-PH/43
827 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
829 PH/47 Put in an explicit test for a DNS lookup of an address record where the
830 "domain" is actually an IP address, and force a failure. This locks out
831 those revolvers/nameservers that support "A-for-A" lookups, in
832 contravention of the specifications.
834 PH/48 When a host name was looked up from an IP address, and the subsequent
835 forward lookup of the name timed out, the host name was left in
836 $sender_host_name, contrary to the specification.
838 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
839 restricted to single-key lookups, Exim was not diagnosing an error if
840 * or *@ was used with a query-style lookup.
842 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
844 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
845 long ago, but noone interested enough thought of it. Let's just say that
846 the "1.1" means that there are some new functions that weren't there at
847 some point in the past.
849 PH/51 Error processing for expansion failure of helo_data from an smtp
850 transport during callout processing was broken.
852 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
853 tested/used via the -bh/-bhc/-bs options.
855 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
856 bug, fixed in subsequent PCRE releases).
858 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
859 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
861 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
867 SC/01 Use a glob alias rather than an array ref in eximstats generated
868 parser. This improves both readability and performance.
870 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
871 Don't display local sender or destination tables in eximstats unless
872 there is data to show.
873 Added average volumes into the eximstats top table text output.
875 SC/03 Collect data on the number of addresses (recipients) as well
876 as the number of messages in eximstats.
878 TF/01 Correct an error in the documentation for the redirect router. Exim
879 does (usually) call initgroups() when daemonizing.
881 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
882 with consistent privilege compared to when running as a daemon.
884 TF/03 Note in the spec that $authenticated_id is not set for local
885 submissions from trusted users.
887 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
888 Thanks to Dean Brooks <dean@iglou.com> for the patch.
890 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
891 by adding some example configuration directives to the default
892 configuration file. A little bit of work is required to uncomment the
893 directives and define how usernames and passwords are checked, but
894 there is now a framework to start from.
896 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
897 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
898 without this. I don't know how relevant this is to other LDAP libraries.
900 PH/02 Add the verb name to the "unknown ACL verb" error.
902 PH/03 Magnus Holmgren's patch for filter_prepend_home.
904 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
906 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
907 directory not expanded when it should be if an expanded home directory
908 was set for the address (which is overridden by the transport).
910 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
913 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
914 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
915 because it is too late at that time, and has no effect.
917 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
918 security issue with \' (bugzilla #107). I could not use the
919 PQescapeStringConn() function, because it needs a PGconn value as one of
922 PH/08 When testing addresses using -bt, indicate those final addresses that
923 are duplicates that would not cause an additional delivery. At least one
924 person was confused, thinking that -bt output corresponded to deliveries.
925 (Suppressing duplicates isn't a good idea as you lose the information
926 about possibly different redirections that led to the duplicates.)
928 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
929 systems where poll() doesn't work, in particular OS X.
931 PH/10 Added more information to debugging output for retry time not reached.
933 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
934 operations in malware.c.
936 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
939 PH/13 If write_rejectlog was set false when logging was sent to syslog with
940 syslog_duplication set false, log lines that would normally be written
941 both the the main log and to the reject log were not written to syslog at
944 PH/14 In the default configuration, change the use of "message" in ACL warn
945 statements to "add_header".
947 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
948 not followed by a command (e.g. "seen endif").
950 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
951 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
954 PH/17 Added extra conditions to the default value of delay_warning_condition
958 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
959 { match{$h_precedence:}{(?i)bulk|list|junk} } \
960 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
963 The Auto-Submitted: and various List- headers are standardised, whereas I
964 don't think Precedence: ever was.
966 PH/18 Refactored debugging code in route_finduser() to show more information,
967 in particular, the error code if getpwnam() issues one.
969 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
970 This is apparently needed in addition to the PH/07 change above to avoid
971 any possible encoding problems.
973 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
974 but not after initializing Perl.
976 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
977 output them only if debugging. By default they are written stderr,
978 apparently, which is not desirable.
980 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
983 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
986 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
988 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
989 authenticated or an ident call has been made. Suppress the default
990 values for $authenticated_id and $authenticated_sender (but permit -oMai
991 and -oMas) when testing with -bh.
993 PH/24 Re-jigged the order of the tests in the default configuration so that the
994 tests for valid domains and recipients precede the DNS black list and CSA
995 tests, on the grounds that those ones are more expensive.
997 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
998 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
999 This bug exists in every version of Exim that I still have, right back to
1002 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1003 However, an attempt to turn on case-sensitivity in a regex key by
1004 including (?-i) didn't work because the subject string was already
1005 lowercased, and the effects were non-intuitive. It turns out that a
1006 one-line patch can be used to allow (?-i) to work as expected.
1012 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1013 other effects) broke the use of negated acl sub-conditions.
1015 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1018 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1019 "Deny" causes Exim to reject the incoming connection with a 554 error.
1020 Unfortunately, if there is a major crisis, such as a disk failure,
1021 tcp-wrappers gives "deny", whereas what one would like would be some
1022 kind of temporary error. A kludge has been added to help with this.
1023 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1024 554 error is used if errno is still zero or contains ENOENT (which occurs
1025 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1028 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1030 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1031 errors. Otherwise a message that provokes a temporary error (when other
1032 messages do not) can cause a whole host to time out.
1034 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1035 addresses were routed directly to files or pipes from a redirect router.
1036 File deliveries just didn't batch; pipe deliveries might have suffered
1039 PH/06 A failure to get a lock for a hints database would erroneously always say
1040 "Failed to get write lock", even when it was really a read lock.
1042 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1043 of 0600. This has been changed to use the value of the lockfile_mode
1044 option (which defaults to 0600).
1046 PH/08 Applied small patch from the Sieve maintainer.
1048 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1049 folder from quota calculations, a direct delivery into this folder messed
1050 up the contents of the maildirsize file. This was because the regex was
1051 used only to exclude .Trash (or whatever) when the size of the mailbox
1052 was calculated. There was no check that a delivery was happening into an
1053 excluded directory. This bug has been fixed by ignoring all quota
1054 processing for deliveries into excluded directories.
1056 PH/10 Added the maildirfolder_create_regex option to appendfile.
1062 PH/01 The code for finding all the local interface addresses on a FreeBSD
1063 system running IPv6 was broken. This may well have applied to all BSD
1064 systems, as well as to others that have similar system calls. The broken
1065 code found IPv4 interfaces correctly, but gave incorrect values for the
1066 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1067 that it would not match correctly against @[] and not recognize the IPv6
1070 PH/02 The ipliteral router was not recognizing addresses of the form user@
1071 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1073 PH/03 Added disable_ipv6.
1075 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1076 decoded form, because it is most often used to construct To: headers
1077 lines in autoreplies, and the decoded form may well be syntactically
1078 invalid. However, $reply_address has leading white space removed, and all
1079 newlines turned into spaces so that the autoreply transport does not
1082 PH/05 If group was specified without a user on a router, and no group or user
1083 was specified on a transport, the group from the router was ignored.
1085 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1086 for visible compile-time settings that can be used to change these
1087 numbers, for those that want even more. Backwards compatibility with old
1088 spool files has been maintained. However, going back to a previous Exim
1089 release will lost any variables that are in spool files.
1091 PH/07 Two small changes when running in the test harness: increase delay when
1092 passing a TCP/IP connection to a new process, in case the original
1093 process has to generate a bounce, and remove special handling of
1094 127.0.0.2 (sic), which is no longer necessary.
1096 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1097 be the same on different OS.
1099 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1102 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1103 whether --show-vars was specified or not
1105 JJ/02 exipick: Added support for new ACL variable spool format introduced
1108 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1109 syntactically invalid From: or Reply-to: line, and a filter used this to
1110 generate an autoreply, and therefore failed to obtain an address for the
1111 autoreply, Exim could try to deliver to a non-existent relative file
1112 name, causing unrelated and misleading errors. What now happens is that
1113 it logs this as a hard delivery error, but does not attempt to create a
1116 PH/11 The exinext utility has a -C option for testing purposes, but although
1117 the given file was scanned by exinext itself; it wasn't being passed on
1118 when Exim was called.
1120 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1121 an end-of-file indication when reading a command response.
1123 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1124 compiled. In many other places in Exim, IPv6 addresses are always
1125 recognized, so I have changed this. It also means that IPv4 domain
1126 literals of the form [IPV4:n.n.n.n] are now always recognized.
1128 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1129 used if the router is not running as root, for example, when verifying at
1130 ACL time, or when using -bh. The debugging output from this situation was
1131 non-existent - all you got was a failure to exec. I have made two
1134 (a) Failures to set uid/gid, the current directory, or a process leader
1135 in a subprocess such as that created by queryprogram now generate
1136 suitable debugging ouput when -d is set.
1138 (b) The queryprogram router detects when it is not running as root,
1139 outputs suitable debugging information if -d is set, and then runs
1140 the subprocess without attempting to change uid/gid.
1142 PH/15 Minor change to Makefile for building test_host (undocumented testing
1145 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1146 additional section of a DNS packet that returns MX or SRV records.
1147 Instead, it always explicitly searches for A/AAAA records. This avoids
1148 major problems that occur when a DNS server includes only records of one
1149 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1150 fixed another bug: if SRV records were looked up and the corresponding
1151 address records were *not* found in the additional section, the port
1152 values from the SRV records were lost.
1154 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1155 using the correct key (the original address) when searching the retry
1156 rules in order to find which one to use for generating the retry hint.
1158 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1159 adding the default one. Similarly, if it contains a Reply-To: header, the
1160 errors_reply_to option, if set, is not used.
1162 PH/19 When calculating a retry time, Exim used to measure the "time since
1163 failure" by looking at the "first failed" field in the retry record. Now
1164 it does not use this if it is later than than the arrival time of the
1165 message. Instead it uses the arrival time. This makes for better
1166 behaviour in cases where some deliveries succeed, thus re-setting the
1167 "first failed" field. An example is a quota failure for a huge message
1168 when small messages continue to be delivered. Without this change, the
1169 "time since failure" will always be short, possible causing more frequent
1170 delivery attempts for the huge message than are intended.
1171 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1173 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1174 $1, $2, $3) because the numerical variables can be reset during some
1175 expansion items (e.g. "match"), thereby losing the authentication data.
1177 PH/21 Make -bV show the size of off_t variables so that the test suite can
1178 decide whether to run tests for quotas > 2G.
1180 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1181 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1182 filecount value is greater than 2G or if a quota value is greater than 2G
1183 on a system where the size of off_t is not greater than 4, a panic error
1186 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1187 never match. The debug and -bh output now contains an explicit error
1188 message indicating a malformed IPv4 address or mask.
1190 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1191 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1192 PH/23 above applies.
1194 PH/25 Do not write to syslog when running in the test harness. The only
1195 occasion when this arises is a failure to open the main or panic logs
1196 (for which there is an explicit test).
1198 PH/26 Added the /no_tell option to "control=freeze".
1200 PH/27 If a host name lookup failed very early in a connection, for example, if
1201 the IP address matched host_lookup and the reverse lookup yielded a name
1202 that did not have a forward lookup, an error message of the form "no IP
1203 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1204 could be logged. Now it outputs the IP address instead of "NULL".
1206 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1207 allows the sender and the authenticated sender to be set when
1208 submitting a message from within Exim. Since child_open_exim() is
1209 documented for local_scan(), the new function should be too.
1211 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1212 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1213 results in an empty string is now treated as unset.
1215 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1217 PH/31 Added sender_verify_fail logging option.
1219 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1220 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1221 tidied the source and removed it altogether.
1223 PH/33 When a queue run was abandoned because the load average was too high, a
1224 log line was always written; now it is written only if the queue_run log
1225 selector is set. In addition, the log line for abandonment now contains
1226 information about the queue run such as the pid. This is always present
1227 in "start" and "stop" lines but was omitted from the "abandon" line.
1229 PH/34 Omit spaces between a header name and the colon in the error message that
1230 is given when verify = headers_syntax fails (if there are lots of them,
1231 the message gets confusing).
1233 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1234 names, as there are now some PTR records that contain slashes. This check
1235 is only to protect against broken name servers that fall over on strange
1236 characters, so the fact that it applies to all lookups doesn't matter.
1238 PH/36 Now that the new test suite is complete, we can remove some of the
1239 special code in Exim that was needed for the old test suite. For example,
1240 sorting DNS records because real resolvers return them in an arbitrary
1241 order. The new test suite's fake resolver always returns records in the
1244 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1245 bounces) except when queue_only is set, to avoid logging races between
1246 the different processes.
1248 PH/38 Panic-die if .include specifies a non-absolute path.
1250 PH/39 A tweak to the "H" retry rule from its user.
1252 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1253 a label. They prevented compilation on older perls.
1255 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1256 a warning to be raised on newish perls.
1258 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1259 on queue. Changes to match documented behaviour of showing count of
1260 messages matching specified criteria.
1262 PH/40 Changed the default ident timeout from 30s to 5s.
1264 PH/41 Added support for the use of login_cap features, on those BSD systems
1265 that have them, for controlling the resources used by pipe deliveries.
1267 PH/42 The content-scanning code uses fopen() to create files in which to put
1268 message data. Previously it was not paying any attention to the mode of
1269 the files. Exim runs with umask(0) because the rest of the code creates
1270 files with open(), and sets the required mode explicitly. Thus, these
1271 files were ending up world-writeable. This was not a big issue, because,
1272 being within the spool directory, they were not world-accessible. I have
1273 created a function called modefopen, which takes an additional mode
1274 argument. It sets umask(777), creates the file, chmods it to the required
1275 mode, then resets the umask. All the relevant calls to fopen() in the
1276 content scanning code have been changed to use this function.
1278 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1279 to 24 hours. This avoids potential overflow problems when processing G
1280 and H retry rules. I suspect nobody ever tinkers with this value.
1282 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1284 PH/45 When the plaintext authenticator is running as a client, the server's
1285 challenges are checked to ensure they are valid base64 strings. By
1286 default, the authentication attempt is cancelled if an invalid string is
1287 received. Setting client_ignore_invalid_base64 true ignores these errors.
1288 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1289 they are received. Thus, the responses can be made to depend on the
1290 challenges. If an invalid string is ignored, an empty string is placed in
1293 PH/46 Messages that are created by the autoreply transport now contains a
1294 References: header, in accordance with RFCs 2822 and 3834.
1296 PH/47 Added authenticated_sender_force to the smtp transport.
1298 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1300 PH/49 Installed latest patch from the Sieve maintainer.
1302 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1303 was also set, the appendfile transport was unnecessarily scanning a
1304 directory of message files (e.g. for maildir delivery) to find the count
1305 of files (along with the size), even though it did not need this
1306 information. It now does the scan only if it needs to find either the
1307 size of the count of files.
1309 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1311 PH/52 Two bugs concerned with error handling when the smtp transport is
1314 (i) Exim was not creating retry information for temporary errors given
1315 for individual recipients after the DATA command when the smtp transport
1316 was used in LMTP mode. This meant that they could be retried too
1317 frequently, and not timed out correctly.
1319 (ii) Exim was setting the flag that allows error details to be returned
1320 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1321 recipients that were returned after the DATA command.
1323 PH/53 This is related to PH/52, but is more general: for any failing address,
1324 when detailed error information was permitted to be returned to the
1325 sender, but the error was temporary, then after the final timeout, only
1326 "retry timeout exceeded" was returned. Now it returns the full error as
1327 well as "retry timeout exceeded".
1329 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1330 do this, and (what is worse) MTAs that accept it.
1332 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1333 will now be deprecated.
1335 PH/56 New os.c-cygwin from the Cygwin maintainer.
1337 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1338 formats (previously only available in exim formats via -bpr, -bpru,
1339 and -bpra. Now also available in native and exiqgrep formats)
1341 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1342 with very large, slow to parse queues
1344 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1346 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1348 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1349 responses to authentication challenges, though it was showing the
1350 challenges; (ii) I've removed the CR characters from the debug output for
1353 PH/58 Allow for the insertion of a newline as well as a space when a string
1354 is turned into more than one encoded-word during RFC 2047 encoding. The
1355 Sieve code now uses this.
1357 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1358 data_4xx, lost_connection, tls_required.
1360 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1361 message was being sent as an SMTP response.
1363 PH/61 Add -l and -k options to exicyclog.
1365 PH/62 When verifying, if an address was redirected to one new address, so that
1366 verification continued, and the new address failed or deferred after
1367 having set something in $address_data, the value of $address_data was not
1368 passed back to the ACL. This was different to the case when no
1369 redirection occurred. The value is now passed back in both cases.
1371 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1372 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1373 use login_cap.h, so on its own it isn't the distinguishing feature. The
1374 new name refers directly to the setclassresources() function.
1376 PH/65 Added configuration files for NetBSD3.
1378 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1380 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1381 is preferred over IPv4.
1383 PH/68 The bounce_return_message and bounce_return_body options were not being
1384 honoured for bounces generated during the reception of non-SMTP messages.
1385 In particular, this applied to messages rejected by the ACL. This bug has
1386 been fixed. However, if bounce_return_message is true and bounce_return_
1387 body is false, the headers that are returned for a non-SMTP message
1388 include only those that have been read before the error was detected.
1389 (In the case of an ACL rejection, they have all been read.)
1391 PH/69 The HTML version of the specification is now built in a directory called
1392 spec_html instead of spec.html, because the latter looks like a path with
1393 a MIME-type, and this confuses some software.
1395 PH/70 Catch two compiler warnings in sieve.c.
1397 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1398 function verify_get_ident() calls ip_connect() to connect a socket, but
1399 if the "connect()" function timed out, ip_connect() used to close the
1400 socket. However, verify_get_ident() also closes the socket later, and in
1401 between Exim writes to the log, which may get opened at this point. When
1402 the socket was closed in ip_connect(), the log could get the same file
1403 descriptor number as the socket. This naturally causes chaos. The fix is
1404 not to close the socket in ip_connect(); the socket should be closed by
1405 the function that creates it. There was only one place in the code where
1406 this was missing, in the iplookup router, which I don't think anybody now
1407 uses, but I've fixed it anyway.
1409 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1410 well as to direct DNS lookups. Otherwise the handling of names in host
1411 lists is inconsistent and therefore confusing.
1417 PH/01 Two changes to the default runtime configuration:
1419 (1) Move the checks for relay_from_hosts and authenticated clients from
1420 after to before the (commented out) DNS black list checks.
1422 (2) Add control=submission to the relay_from_hosts and authenticated
1423 clients checks, on the grounds that messages accepted by these
1424 statements are most likely to be submissions.
1426 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1428 (1) Generate an error if the third argument for the ${prvs expansion is
1431 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1434 (3) Reset the variables that are obtained from the first argument of
1435 ${prvscheck and used in the second argument before leaving the code,
1436 because their memory is reclaimed, so using them afterwards may do
1439 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1440 one (it's much easier than Tom thought :-).
1442 (5) Because of (4), we can now allow for the use of $prvscheck_result
1443 inside the third argument.
1445 PH/03 For some reason, the default setting of PATH when running a command from
1446 a pipe transport was just "/usr/bin". I have changed it to
1449 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1450 anything to be listed in the output from -bV.
1452 PH/05 When a filter generated an autoreply, the entire To: header line was
1453 quoted in the delivery log line, like this:
1455 => >A.N.Other <ano@some.domain> <original@ddress> ...
1457 This has been changed so that it extracts the operative address. There
1458 may be more than one such address. If so, they are comma-separated, like
1461 => >ano@some.domain,ona@other.domain <original@ddress> ...
1463 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1464 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1465 not being looked up in the rDNS to get a host name, Exim was showing the
1466 IP address twice in Received: lines, even though the IP addresses were
1467 identical. For example:
1469 Received: from [1.2.3.4] (helo=[1.2.3.4])
1471 However, if the real host name was known, it was omitting the HELO data
1472 if it matched the actual IP address. This has been tidied up so that it
1473 doesn't show the same IP address twice.
1475 PH/07 When both +timestamp and +memory debugging was on, the value given by
1476 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1477 called by the debug printing, thereby overwriting the timestamp buffer.
1478 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1480 PH/08 When the original message was included in an autoreply transport, it
1481 always said "this is a copy of the message, including all the headers",
1482 even if body_only or headers_only was set. It now gives an appropriate
1485 PH/09 Applied a patch from the Sieve maintainer which:
1487 o fixes some comments
1488 o adds the (disabled) notify extension core
1489 o adds some debug output for the result of if/elsif tests
1490 o points to the current vacation draft in the documentation
1491 and documents the missing references header update
1495 o fixes a bug in processing the envelope test (when testing
1496 multiple envelope elements, the last element determinted the
1499 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1500 Electronic Mail") by including:
1502 Auto-submitted: auto-generated
1504 in the messages that it generates (bounce messages and others, such as
1505 warnings). In the case of bounce messages for non-SMTP mesages, there was
1506 also a typo: it was using "Auto_submitted" (underscore instead of
1507 hyphen). Since every message generated by Exim is necessarily in response
1508 to another message, thes have all been changed to:
1510 Auto-Submitted: auto-replied
1512 in accordance with these statements in the RFC:
1514 The auto-replied keyword:
1516 - SHOULD be used on messages sent in direct response to another
1517 message by an automatic process,
1519 - MUST NOT be used on manually-generated messages,
1521 - MAY be used on Delivery Status Notifications (DSNs) and Message
1522 Disposition Notifications (MDNs),
1524 - MUST NOT be used on messages generated by automatic or periodic
1525 processes, except for messages which are automatic responses to
1528 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1529 to the default Received: header definition.
1531 PH/12 Added log selector acl_warn_skipped (default on).
1533 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1534 variables because (a) they are in the wrong storage pool and (b) even if
1535 they were copied, it wouldn't work properly because of the caching.
1537 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1538 checking when decoding. Apparently there are clients that generate
1539 overlong encoded strings. Why am I not surprised?
1541 PH/15 If the first argument of "${if match_address" was not empty, but did not
1542 contain an "@" character, Exim crashed. Now it writes a panic log message
1543 and treats the condition as false.
1545 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1547 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1548 extension "envelope-auth". The code is finished and in agreement with
1549 other implementations, but there is no documentation so far and in fact,
1550 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1551 not changing the active code.
1553 Print executed "if" and "elsif" statements when debugging is used. This
1554 helps a great deal to understand what a filter does.
1556 Document more things not specified clearly in RFC3028. I had all this
1557 sorted out, when out of a sudden new issues came to my mind. Oops."
1559 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1562 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1563 canonical form (as documented). However, after a host name lookup from
1564 the IP address, check_host() was doing a simple string comparison with
1565 addresses acquired from the DNS when checking that the found name did
1566 have the original IP as one of its addresses. Since any found IPv6
1567 addresses are likely to be in abbreviated form, the comparison could
1568 fail. Luckily, there already exists a function for doing the comparison
1569 by converting both addresses to binary, so now that is used instead of
1570 the text comparison.
1572 PH/20 There was another similar case to PH/19, when a complete host name was
1573 given in a host list; looking up its IP address could give an abbreviated
1574 form, whereas the current host's name might or might not be abbreviated.
1575 The same fix has been applied.
1581 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1582 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1585 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1586 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1588 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1590 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1591 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1592 Also, for "Auto-Submitted", treat anything other than "no" as
1593 non-personal, in accordance with RFC 3834. (Previously it treated
1594 anything starting "auto-" as non-personal.)
1596 TF/01 The control=submission/name=... option had a problem with syntax
1597 errors if the name included a slash character. The /name= option
1598 now slurps the rest of the string, so it can include any characters
1599 but it must come last in the list of options (after /sender_retain
1602 PH/05 Some modifications to the interface to the fake nameserver for the new
1610 TK/01 Added the "success_on_redirect" address verification option. See
1611 NewStuff for rationale and an example.
1613 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1615 PH/02 Patch to exigrep to allow it to work on syslog lines.
1617 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1618 fread() to skip over the body file's header line, because in Cygwin the
1619 header line is locked and is inaccessible.
1621 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1622 co-exist for some time) to make it clear that it is the Exim ID that is
1623 referenced, not the Message-ID: header line.
1625 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1626 string_format() function, because snprintf() does not exist on all
1629 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1630 ${stat: expansion item.
1632 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1633 protocol synchronization error", to keep the pedants happy.
1635 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1636 well as for IRIX systems, when gcc is being used. See the host.c source
1639 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1641 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1644 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1645 transport and to the smtp transport in LMTP mode.
1647 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1649 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1650 run a filter in a subprocess. This could lead to confusion in subsequent
1651 lookups in the parent process. There should also be a search_tidyup() at
1652 the end of the subprocess.
1654 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1655 only if the host matched helo_try_verify_hosts, which caused the
1656 verification to occur when the EHLO/HELO command was issued. The ACL just
1657 tested the remembered result. Now, if a previous verification attempt has
1658 not happened, "verify = helo" does it there and then.
1660 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1662 TK/03 Fix log output including CR from clamd.
1664 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1665 exist provoked a memory error which could cause a segfault.
1667 PH/15 Installed PCRE 6.2
1669 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1671 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1672 of the problem. Specifically, suggested +O2 rather than +O1 for the
1675 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1677 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1678 record, the port was not correctly logged when the outgoing_port log
1679 selector was set (it logged the transort's default port).
1681 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1682 fallback_hosts, and "hosts" in the smtp transport.
1684 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1685 host errors such as "Connection refused".
1687 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1688 authentication with radiusclient 0.4.9:
1690 - Error returned from rc_read_config was caught wrongly
1691 - Username/password not passed on to radius server due to wrong length.
1693 The presumption is that some radiusclient API changes for 4.51/PH/17
1694 were not taken care of correctly. The code is still untested by me (my
1695 Linux distribution still has 0.3.2 of radiusclient), but it was
1696 contributed by a Radius user.
1698 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1699 expanding the "port" option of the smtp transport.
1701 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1702 while reading a MIME header. Thanks to Tom Hughes for a patch.
1704 PH/24 Include config.h inside local_scan.h so that configuration settings are
1707 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1708 that in an ACL for RCPT (for example), you can examine exactly what was
1711 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1712 commands, but it was not correctly comparing the address with the actual
1713 client host address. Thus, it would show the EHLO address in Received:
1714 header lines when this was not necessary.
1716 PH/27 Added the % operator to ${eval:}.
1718 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1719 it should be ignoring failures (because with -C, for example, it has lost
1720 privilege). It wasn't ignoring creation failures other than "already
1723 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1726 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1727 because an input line was too long, either on its own, or by virtue of
1728 too many continuations, the temporary file was not being removed, and the
1729 return code was incorrect.
1731 PH/31 Missing "BOOL" in function definition in filtertest.c.
1733 PH/32 Applied Sieve patches from the maintainer.
1735 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1737 PH/33 Added "verify = not_blind".
1739 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1740 Local/Makefile (with some defaults set). These are used in built scripts
1741 such as exicyclog, but they have never been used in the exim_install
1742 script (though there are many overriding facilities there). I have
1743 arranged that the exim_install script now takes note of these two
1746 PH/35 Installed configuration files for Dragonfly.
1748 PH/36 When a locally submitted message by a trusted user did not contain a
1749 From: header, and the sender address was obtained from -f or from an SMTP
1750 MAIL command, and the trusted user did not use -F to supply a sender
1751 name, $originator_name was incorrectly used when constructing a From:
1752 header. Furthermore, $originator_name was used for submission mode
1753 messages from external hosts without From: headers in a similar way,
1754 which is clearly wrong.
1756 PH/37 Added control=suppress_local_fixups.
1758 PH/38 When log_selector = +received_sender was set, and the addition of the
1759 sender made the log line's construction buffer exactly full, or one byte
1760 less than full, an overflow happened when the terminating "\n" was
1763 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1764 when the result of a list match is failure because a DNS lookup failed.
1766 PH/40 RM_COMMAND is now used in the building process.
1768 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1769 the "build-* directories that it finds.
1771 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1772 address in a domain literal was a prefix of an interface address.
1774 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1775 when verifying a sender address, unless rewrite_headers is false.
1777 PH/44 Wrote a long comment about why errors_to addresses are verified as
1778 recipients, not senders.
1780 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1781 the ratelimit ACL was added.
1783 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1785 PH/46 Added extra information about PostgreSQL errors to the error string.
1787 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1788 avoiding the need to install special zones in a real server. This is
1789 backwards compatible; if it can't find the fake resolver, it drops back.
1790 Thus, both old and new test suites can be run.
1792 TF/02 Added util/ratelimit.pl
1794 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1795 clock is set back in time.
1797 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1798 Candler <B.Candler@pobox.com>.
1800 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1801 OK for addresses that are the result of redirections.
1803 PH/48 A number of further additions for the benefit of the new test suite,
1804 including a fake gethostbyname() that interfaces to the fake DNS resolver
1807 TF/06 The fix for widen_domains has also been applied to qualify_single and
1808 search_parents which are the other dnslookup options that can cause
1811 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1814 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1816 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1817 service" if it encountered an MX record with an empty target hostname.
1818 The message is now "an MX or SRV record indicated no SMTP service".
1820 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1821 if the DNS of the sending site is misconfigured. This is quite a
1822 common situation. This change restores the behaviour of treating a
1823 helo verification defer as a failure.
1825 PH/51 If self=fail was set on a router, the bounce message did not include the
1826 actual error message.
1832 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1834 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1835 command itself ended in error, the underlying message about the transport
1836 filter timeout was being overwritten with the pipe command error. Now the
1837 underlying error message should be appended to the second error message.
1839 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1841 PH/02 Reduce the amount of output that "make" produces by default. Full output
1842 can still be requested.
1844 PH/03 The warning log line about a condition test deferring for a "warn" verb
1845 was being output only once per connection, rather than after each
1846 occurrence (because it was using the same function as for successful
1847 "warn" verbs). This seems wrong, so I have changed it.
1849 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1850 it should not have, which might have caused a crash in the right
1851 circumstances, but probably never did.
1853 PH/04 Installed a modified version of Tony Finch's patch to make submission
1854 mode fix the return path as well as the Sender: header line, and to
1855 add a /name= option so that you can make the user's friendly name appear
1858 TF/03 Added the control = fakedefer ACL modifier.
1860 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1861 Mark Lowes for thorough testing.
1863 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1865 TK/03 Merged latest SRS patch from Miles Wilton.
1867 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1868 with the definition in sysexits.h (which is #included earlier).
1869 Fortunately, Exim does not actually use EX_OK. The code used to try to
1870 preserve the sysexits.h value, by assumimg that macro definitions were
1871 scanned for macro replacements. I have been disabused of this notion,
1872 so now the code just undefines EX_OK before #including unistd.h.
1874 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1875 in the smtp transport. When a block could not be written in a single
1876 write() function, the timeout was being re-applied to each part-write.
1877 This seems wrong - if the receiver was accepting one byte at a time it
1878 would take for ever. The timeout is now adjusted when this happens. It
1879 doesn't have to be particularly precise.
1881 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1882 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1884 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1885 without a domain if the check to <postmaster@domain> fails.
1887 SC/01 Eximstats: added -xls and the ability to specify output files
1888 (patch written by Frank Heydlauf).
1890 SC/02 Eximstats: use FileHandles for outputing results.
1892 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1894 SC/04 Eximstats: fixed display of large numbers with -nvr option
1896 SC/05 Eximstats: fixed merging of reports with empty tables.
1898 SC/06 Eximstats: added the -include_original_destination flag
1900 SC/07 Eximstats: removed tabs and trailing whitespace.
1902 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1904 TK/06 MBOX spool code: Add real "From " MBOX separator line
1905 so the .eml file is really in mbox format (even though
1906 most programs do not really care). Patch from Alex Miller.
1908 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1909 The latter is generated from $received_to and is only set if the
1910 message has one envelope recipient. SA can use these headers,
1911 obviously out-of-the-box. Patch from Alex Miller.
1913 PH/08 The ${def test on a variable was returning false if the variable's
1914 value was "0", contrary to what the specification has always said!
1915 The result should be true unless the variable is empty.
1917 PH/09 The syntax error of a character other than { following "${if
1918 def:variable_name" (after optional whitespace) was not being diagnosed.
1919 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1920 accidental colon was present, for example, could give incorrect results.
1922 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1923 result is used (not including appendfile, where other changes are about
1926 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1927 This involved changing a lot of size variables from int to off_t. It
1928 should work with maildirs and everything.
1930 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1931 spamd dying while we are connected to it.
1933 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1936 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1937 function for BDB 4.3.
1939 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1941 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1942 the sources for PCRE have been re-arranged and more files are now
1945 PH/15 The code I had for printing potentially long long variables in PH/11
1946 above was not the best (it lost precision). The length of off_t variables
1947 is now inspected at build time, and an appropriate printing format (%ld
1948 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1949 to be "long long int" or "long int". This is needed for the internal
1950 formatting function string_vformat().
1952 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1953 the configuration file to be ":syslog", then the script "guesses" where
1954 the logs files are, rather than using the compiled in default. In our
1955 case the guess is not the same as the compiled default, so the script
1956 suddenly stopped working when I started to use syslog. The patch checks
1957 to see if log_file_path is "". If so, it attempts to read it from exim
1958 with no configuration file to get the compiled in version, before it
1959 falls back to the previous guessing code."
1961 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1962 implementing BATV in an Exim configuration. See NewStuff for the gory
1965 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1966 Makefile that are specific to HP-UX.
1968 PH/18 If the "use_postmaster" option was set for a recipient callout together
1969 with the "random" option, the postmaster address was used as the MAIL
1970 FROM address for the random test, but not for the subsequent recipient
1971 test. It is now used for both.
1973 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1974 patch removes a few documentation additions to RFC 3028, because the
1975 latest draft now contains them. It adds the new en;ascii-case comparator
1976 and a new error check for 8bit text in MIME parts. Comparator and
1977 require names are now matched exactly. I enabled the subaddress
1978 extension, but it is not well tested yet (read: it works for me)."
1980 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1981 rework some of the code of TK/09 above to avoid the hardwired use of
1982 "%lld" and "long long". Replaced the call to snprintf() with a call to
1985 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1986 records point to non-existent hosts", "retry timeout exceeded", and
1987 "retry time not reached for any host after a long failure period".
1989 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1990 experimental DomainKeys support:
1992 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1993 (2) On an error such as an illegally used "control", the wrong name for
1994 the control was given.
1996 These problems did NOT occur unless DomainKeys support was compiled.
1998 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2000 PH/24 Added ${if match_ip condition.
2002 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2003 databases so that it will be absolutely obvious if a crash occurs in the
2004 DB library. This is a regular occurrence (often caused by mis-matched
2007 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2008 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2009 picked up on a user's system that detects such things. There doesn't seem
2010 to be a gcc warning option for this - only an attribute that has to be
2011 put on the function's prototype. It seems that in Fedora Core 4 they have
2012 set this on a number of new functions. No doubt there will be more in due
2015 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2016 specify a transport. However, if an address that was verified by such a
2017 router was the subject of a callout, Exim crashed because it tried to
2018 read the rcpt_include_affixes from the non-existent transport. Now it
2019 just assumes that the setting of that option is false. This bug was
2020 introduced by 4.51/PH/31.
2022 PH/28 Changed -d+all to exclude +memory, because that information is very
2023 rarely of interest, but it makes the output a lot bigger. People tend to
2024 do -d+all out of habit.
2026 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2027 code in os-type was giving problems when libc.so lives in lib64, like on
2030 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2031 aren't the modern standard, and it seems that some systems' include files
2032 don't always have them. Exim was already checking for some of the newer
2033 ones like T_AAAA, and defining it itself. I've added checks for all the
2034 record types that Exim uses.
2036 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2037 not automatically generating a new one, as it is supposed to. This
2038 prevented TLS from working. If the file did exist, but contained invalid
2039 data, a new version was generated, as expected. It was only the case of a
2040 non-existent file that was broken.
2042 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2043 with a change in libdomainkeys > 0.64.
2045 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2046 from DNS. If the selector record carries the flag, it now has
2047 precedence over the domain-wide flag.
2049 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2051 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2052 the use of a port name that isn't defined in /etc/services) occurred, the
2053 message was deferred as in a normal delivery, and thus remained on the
2054 spool, instead of being failed because of the mua_wrapper setting. This
2055 is now fixed, and I tidied up some of the mua_wrapper messages at the
2058 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2059 the messages in a hash of arrays rather than using individual hashes.
2060 This is a bit cleaner and results in dramatic memory savings, albeit
2061 at a slight CPU cost.
2063 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2064 as requested by Marc Sherman.
2066 SC/10 Eximstats: added histograms for user specified patterns as requested
2069 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2071 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2072 fopen() in the content-scanning modules that did not already have it.
2078 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2079 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2081 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2083 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2085 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2086 to test Sieve filters that use "vacation".
2088 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2089 that changes the way the GnuTLS parameters are stored in the cache file.
2090 The new format can be generated externally. For backward compatibility,
2091 if the data in the cache doesn't make sense, Exim assumes it has read an
2092 old-format file, and it generates new data and writes a new file. This
2093 means that you can't go back to an older release without removing the
2096 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2097 work if there are any delivery delays because "one_time" forces the
2098 parent to be marked "delivered", so its unseen clone is never tried
2099 again. For this reason, Exim now forbids the simultaneous setting of
2102 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2103 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2104 entry if you want to know the details. The fix, however, neglected to
2105 consider the case when local delivery batching is involved. The test for
2106 "previously delivered" was not happening when checking to see if an
2107 address could be batched with a previous (undelivered) one; under
2108 certain circumstances this could lead to multiple deliveries to the same
2111 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2112 in its include files, and this causes problems building Exim.
2114 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2115 header_syntax) but Exim was just ignoring anything given after a slash.
2116 In particular, this caused confusion with an attempt to use "verify =
2117 reverse_host_lookup/defer_ok". An error is now given when options are
2118 supplied for verify items that do not have them. (Maybe reverse_host_
2119 lookup should have a defer_ok option, but that's a different point.)
2121 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2122 defined by RFC 821) to 2048, because there were problems with some AUTH
2123 commands, and RFC 1869 says the size should be increased for extended
2124 SMTP commands that take arguments.
2126 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2129 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2130 "unknown" error; now it says that the functionality isn't in the binary.
2132 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2133 an address' error message when a string expansion fails (syntax or
2134 whatever). Otherwise the password may appear in the log. Following change
2135 PH/42 below, there is no longer a chance of it appearing in a bounce
2138 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2140 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2141 of its addresses was ever tried. (Bugzilla bug #2.)
2143 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2144 the result incorrectly in the debug output. (It correctly added a newline
2145 to what was transported.)
2147 TF/01 Added $received_time.
2149 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2150 commented out examples of how to interface to a virus scanner and to
2151 SpamAssassin. Also added commented examples of av_scanner and
2152 spamd_address settings.
2154 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2155 and controls are allowed in which ACLs. There were a couple of minor
2156 errors. Some of the entries in the conditions table (which is a table of
2157 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2158 negation of where the condition IS allowed.
2160 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2162 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2163 header file does not have a version number, so I've had to invent a new
2164 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2165 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2166 radiusclient), but it was contributed by a Radius user.
2168 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2169 files or directories, for OpenSSL.
2171 PH/19 When an Exim process that is running as root has to create an Exim log
2172 file, it does so in a subprocess that runs as exim:exim so as to get the
2173 ownership right at creation (otherwise, other Exim processes might see
2174 the file with the wrong ownership). There was no test for failure of this
2175 fork() call, which would lead to the process getting stuck as it waited
2176 for a non-existent subprocess. Forks do occasionally fail when resources
2177 run out. I reviewed all the other calls to fork(); they all seem to check
2180 PH/20 When checking for unexpected SMTP input at connect time (before writing
2181 the banner), Exim was not dealing correctly with a non-positive return
2182 from the read() function. If the client had disconnected by this time,
2183 the result was a log entry for a synchronization error with an empty
2184 string after "input=" when read() returned zero. If read() returned -1
2185 (an event I could not check), uninitialized data bytes were printed.
2186 There were reports of junk text (parts of files, etc) appearing after
2189 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2191 PH/22 Added support for macro redefinition, and (re)definition in between
2192 driver and ACL definitions.
2194 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2195 forgetting to use the resulting value; it was using the unexpanded value.
2197 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2198 hadn't been configured. The fix is from Juergen Kreileder, who
2199 understands it better than I do:
2201 "Here's what I see happening with three configured cyrus_sasl
2202 authenticators configured (plain, login, cram-md5):
2204 On startup auth_cyrus_sasl_init() gets called for each of these.
2205 This means three calls to sasl_listmech() without a specified mech_list.
2206 => SASL tests which mechs of all available mechs actually work
2207 => three warnings about OTP not working
2208 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2210 With the patch, sasl_listmech() also gets called three times. But now
2211 SASL's mech_list option is set to the server_mech specified in the the
2212 authenticator. Or in other words, the answer from sasl_listmech()
2213 gets limited to just the mech you're testing for (which is different
2215 => the return list contains just 'plain' or 'login', 'cram-md5' or
2216 nothing depending on the value of ob->server_mech.
2218 I've just tested the patch: Authentication still works fine,
2219 unavailable mechs specified in the exim configuration are still
2220 caught, and the auth.log warnings about OTP are gone."
2222 PH/25 When debugging is enabled, the contents of the command line are added
2223 to the debugging output, even when log_selector=+arguments is not
2226 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2227 answer is "GNU", and only if the return is "GNU/something" is the answer
2230 PH/27 $acl_verify_message is now set immediately after the failure of a
2231 verification in an ACL, and so is available in subsequent modifiers. In
2232 particular, the message can be preserved by coding like this:
2234 warn !verify = sender
2235 set acl_m0 = $acl_verify_message
2237 Previously, $acl_verify_message was set only while expanding "message"
2238 and "log_message" when a very denied access.
2240 PH/28 Modified OS/os.c-Linux with
2242 -#ifndef OS_LOAD_AVERAGE
2243 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2245 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2246 nomenclature these days.)
2248 PH/29 Installed patch from the Sieve maintainer that adds the options
2249 sieve_useraddress and sieve_subaddress to the redirect router.
2251 PH/30 In these circumstances:
2252 . Two addresses routed to the same list of hosts;
2253 . First host does not offer TLS;
2254 . First host accepts first address;
2255 . First host gives temporary error to second address;
2256 . Second host offers TLS and a TLS session is established;
2257 . Second host accepts second address.
2258 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2259 and peerdn, if requested) that were in fact used only for the second
2262 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2263 attention to any local part prefix or suffix that was matched by the
2264 router that accepted the address. It now behaves in the same way as it
2265 does for delivery: the affixes are removed from the local part unless
2266 rcpt_include_affixes is set on the transport.
2268 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2269 timeout during the DATA phase of an incoming message.
2271 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2272 applied a patch sanctioned by the Sieve maintainer.
2274 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2275 the uid or gid is negative. A case of a negative gid caused this to be
2276 noticed. The fix allows for either to be negative.
2278 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2279 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2282 PH/36 The change PH/12 above was broken. Fixed it.
2284 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2285 the grounds that routing the same address twice would always produce the
2286 same answer. This might have been true once, but it is certainly no
2287 longer true now. Routing a child address may depend on the previous
2288 routing that produced that child. Some complicated redirection strategies
2289 went wrong when messages had multiple recipients, and made Exim's
2290 behaviour dependent on the order in which the addresses were given.
2292 I have moved the duplicate checking until after the routing is complete.
2293 Exim scans the addresses that are assigned to local and remote
2294 transports, and removes any duplicates. This means that more work will be
2295 done, as duplicates will always all be routed, but duplicates are
2296 presumably rare, so I don't expect this is of any significance.
2298 For deliveries to pipes, files, and autoreplies, the duplicate checking
2299 still happens during the routing process, since they are not going to be
2302 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2303 It corrects a timeout issue with spamd. This is Ian's comment: "The
2304 background is that sometimes spamd either never reads data from a
2305 connection it has accepted, or it never writes response data. The exiscan
2306 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2307 blindly assumes that writes won't block so it may never time out."
2309 PH/39 Allow G after quota size as well as K and M.
2311 PH/40 The value set for $authenticated_id in an authenticator may not contain
2312 binary zeroes or newlines because the value is written to log lines and
2313 to spool files. There was no check on this. Now the value is run through
2314 the string_printing() function so that such characters are converted to
2315 printable escape sequences.
2317 PH/41 $message_linecount is a new variable that contains the total number of
2318 lines in the message. Compare $body_linecount, which is the count for the
2321 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2322 bounce and delay warning messages, except in certain special cases, which
2325 (a) An SMTP error message from a remote host;
2326 (b) A message specified in a :fail: redirection;
2327 (c) A message specified in a "fail" command in a system filter;
2328 (d) A message specified in a FAIL return from the queryprogram router;
2329 (e) A message specified by the cannot_route_message router option.
2331 In these cases only, Exim does include the error details in bounce and
2332 warning messages. There are also a few cases where bland messages such
2333 as "unrouteable address" or "local delivery error" are given.
2335 PH/43 $value is now also set for the "else" part of a ${run expansion.
2337 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2338 being worked on, but at least Exim now implements the latest version to
2341 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2342 process to complete was treated as a delivery failure, a timeout while
2343 writing the message to the pipe was logged, but erroneously treated as a
2344 successful delivery. Such timeouts include transport filter timeouts. For
2345 consistency with the overall process timeout, these timeouts are now
2346 treated as errors, giving rise to delivery failures by default. However,
2347 there is now a new Boolean option for the pipe transport called
2348 timeout_defer, which, if set TRUE, converts the failures into defers for
2349 both kinds of timeout. A transport filter timeout is now identified in
2352 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2353 systems where "make" and "gmake" are different, calling "gmake" at top
2354 level broke things. I've arranged for the value of $(MAKE) to be passed
2355 from the Makefile to this script so that it can call the same version of
2359 A note about Exim versions 4.44 and 4.50
2360 ----------------------------------------
2362 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2363 changes of various kinds. As a consequence, a big documentation update was
2364 needed. This delayed the release for rather longer than seemed good, especially
2365 in the light of a couple of (minor) security issues. Therefore, the changes
2366 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2367 release. So 4.44 and 4.50 are in effect two different branches that both start
2370 I have left the 4.50 change log unchanged; it contains all the changes since
2371 4.43. The change log for 4.44 is below; many of its items are identical to
2372 those for 4.50. This seems to be the most sensible way to preserve the
2373 historical information.
2379 1. Minor wording change to the doc/README.SIEVE file.
2381 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2382 computation of the current number of files was incorrect.
2384 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2385 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2388 4. Give more explanation in the error message when the command for a transport
2389 filter fails to execute.
2391 5. There are several places where Exim runs a non-Exim command in a
2392 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2393 was being done only for the command run by the queryprogram router. It is
2394 now done for all such subprocesses. The other cases are: ${run, transport
2395 filters, and the commands run by the lmtp and pipe transports.
2397 6. Added CONFIGURE_GROUP build-time option.
2399 7. Some older OS have a limit of 256 on the maximum number of file
2400 descriptors. Exim was using setrlimit() to set 1000 as a large value
2401 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2402 systems. I've change it so that if it can't get 1000, it tries for 256.
2404 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2405 was an oversight, and furthermore, ever since the addition of extra
2406 controls (e.g. 4.43/32), the checks on when to allow different forms of
2407 "control" were broken. There should now be diagnostics for all cases when a
2408 control that does not make sense is encountered.
2410 9. Added the /retain_sender option to "control=submission".
2412 10. $recipients is now available in the predata ACL (oversight).
2414 11. Tidy the search cache before the fork to do a delivery from a message
2415 received from the command line. Otherwise the child will trigger a lookup
2416 failure and thereby defer the delivery if it tries to use (for example) a
2417 cached ldap connection that the parent has called unbind on.
2419 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2420 of $address_data from the recipient verification was clobbered by the
2421 sender verification.
2423 13. The value of address_data from a sender verification is now available in
2424 $sender_address_data in subsequent conditions in the ACL statement.
2426 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2428 15. Added a new option "connect=<time>" to callout options, to set a different
2431 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2432 was its contents. (It was OK if the option was not defined at all.)
2434 17. A "Completed" log line is now written for messages that are removed from
2435 the spool by the -Mrm option.
2437 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2438 information about exactly what failed.
2440 19. Added -dd to debug only the daemon process.
2442 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2443 handles timeouts, both on the server side and network timeouts. Renamed the
2444 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2446 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2447 It is now set to "smtps".
2449 22. $host_address is now set to the target address during the checking of
2450 ignore_target_hosts.
2452 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2453 being passed; this would have caused $sender_host_name to have been used if
2454 matching the list had actually called for a host name (not very likely,
2455 since this list is usually IP addresses). A host name is now passed as
2458 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2459 code that specifies a non-restarting handler (typically sigaction() in
2460 modern systems) in an attempt to fix a rare and obscure crash bug.
2462 25. Narrowed the window for a race in the daemon that could cause it to ignore
2463 SIGCHLD signals. This is not a major problem, because they are used only to
2464 wake it up if nothing else does.
2466 26. A malformed maildirsize file could cause Exim to calculate negative values
2467 for the mailbox size or file count. Odd effects could occur as a result.
2468 The maildirsize information is now recalculated if the size or filecount
2471 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2472 support for a long time. Removed HAVE_SYS_VFS_H.
2474 28. Installed the latest version of exipick from John Jetmore.
2476 29. In an address list, if the pattern was not a regular expression, an empty
2477 subject address (from a bounce message) matched only if the pattern was an
2478 empty string. Non-empty patterns were not even tested. This was the wrong
2479 because it is perfectly reasonable to use an empty address as part of a
2480 database query. An empty address is now tested by patterns that are
2481 lookups. However, all the other forms of pattern expect the subject to
2482 contain a local part and a domain, and therefore, for them, an empty
2483 address still always fails if the pattern is not itself empty.
2485 30. Exim went into a mad DNS loop when attempting to do a callout where the
2486 host was specified on an smtp transport, and looking it up yielded more
2487 than one IP address.
2489 31. Re-factored the code for checking spool and log partition space into a
2490 function that finds that data and another that does the check. The former
2491 is then used to implement four new variables: $spool_space, $log_space,
2492 $spool_inodes, and $log_inodes.
2494 32. The RFC2047 encoding function was originally intended for short strings
2495 such as real names; it was not keeping to the 75-character limit for
2496 encoded words that the RFC imposes. It now respects the limit, and
2497 generates multiple encoded words if necessary. To be on the safe side, I
2498 have increased the buffer size for the ${rfc2047: expansion operator from
2501 33. It is now permitted to omit both strings after an "if" condition; if the
2502 condition is true, the result is "true". As before, when the second string
2503 is omitted, a false condition yields an empty string. This makes it less
2504 cumbersome to write custom ACL and router conditions.
2506 34. Failure to deliver a bounce message always caused it to be frozen, even if
2507 there was an errors_to setting on the router. The errors_to setting is now
2510 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2511 canonical form (fully expanded) before being placed in
2512 $sender_host_address.
2514 36. The table in the code that translates DNS record types into text (T_A to
2515 "A" for instance) was missing entries for NS and CNAME. It is just possible
2516 that this could have caused confusion if both these types were looked up
2517 for the same domain, because the text type is used as part of Exim's
2518 per-process caching. But the chance of anyone hitting this buglet seems
2521 37. The dnsdb lookup has been extended in a number of ways.
2523 (1) There is a new type, "zns", which walks up the domain tree until it
2524 finds some nameserver records. It should be used with care.
2526 (2) There is a new type, "mxh", which is like "mx" except that it returns
2527 just the host names, not the priorities.
2529 (3) It is now possible to give a list of domains (or IP addresses) to be
2530 looked up. The behaviour when one of the lookups defers can be
2531 controlled by a keyword.
2533 (4) It is now possible to specify the separator character for use when
2534 multiple records are returned.
2536 38. The dnslists ACL condition has been extended: it is now possible to supply
2537 a list of IP addresses and/or domains to be looked up in a particular DNS
2540 39. Added log_selector=+queue_time_overall.
2542 40. When running the queue in the test harness, wait just a tad after forking a
2543 delivery process, to get repeatability of debugging output.
2545 41. Include certificate and key file names in error message when GnuTLS fails
2546 to set them up, because the GnuTLS error message doesn't include the name
2547 of the failing file when there is a problem reading it.
2549 42. Allow both -bf and -bF in the same test run.
2551 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2553 44. The "Exiscan patch" is now merged into the mainline Exim source.
2555 45. Sometimes the final signoff response after QUIT could fail to get
2556 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2557 < 0 before doing a fflush(). This bug looks as though it goes back to the
2558 introduction of TLS in release 3.20, but "sometimes" must have been rare
2559 because the tests only now provoked it.
2561 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2562 (this can affect the format of dates).
2564 47. exim_tidydb, when checking for the continued existence of a message for
2565 which it has found a message-specific retry record, was not finding
2566 messages that were in split spool directories. Consequently, it was
2567 deleting retry records that should have stayed in existence.
2569 48. Steve fixed some bugs in eximstats.
2571 49. The SPA authentication driver was not abandoning authentication and moving
2572 on to the next authenticator when an expansion was forced to fail,
2573 contradicting the general specification for all authenticators. Instead it
2574 was generating a temporary error. It now behaves as specified.
2576 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2577 (the order specifies the preference for clients). The order is now AES256,
2578 AES128, 3DES, ARCFOUR128.
2580 51. Small patch to Sieve code - explicitly set From: when generating an
2583 52. Exim crashed if a remote delivery caused a very long error message to be
2584 recorded - for instance if somebody sent an entire SpamAssassin report back
2585 as a large number of 550 error lines. This bug was coincidentally fixed by
2586 increasing the size of one of Exim's internal buffers (big_buffer) that
2587 happened as part of the Exiscan merge. However, to be on the safe side, I
2588 have made the code more robust (and fixed the comments that describe what
2591 53. Now that there can be additional text after "Completed" in log lines (if
2592 the queue_time_overall log selector is set), a one-byte patch to exigrep
2593 was needed to allow it to recognize "Completed" as not the last thing in
2596 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2597 patch that reportedly fixes this has been added. I am not expert enough to
2598 create a test for it. This is what the patch creator wrote:
2600 "I found a little strange behaviour of ldap code when working with
2601 Windows 2003 AD Domain, where users was placed in more than one
2602 Organization Units. When I tried to give exim partial DN, the exit code
2603 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2604 But simultaneously result of request was absolutely normal ldap result,
2605 so I produce this patch..."
2607 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2608 REFERENCE, so I have modified the code to exclude the patch when that macro
2611 55. Some experimental protocols are using DNS PTR records for new purposes. The
2612 keys for these records are domain names, not reversed IP addresses. The
2613 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2614 leaves it alone. Component reversal etc. now happens only for IP addresses.
2617 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2619 57. Double the size of the debug message buffer (to 2048) so that more of very
2620 long debug lines gets shown.
2622 58. The exicyclog utility now does better if the number of log files to keep
2623 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2625 59. Two changes related to the smtp_active_hostname option:
2627 (1) $smtp_active_hostname is now available as a variable.
2628 (2) The default for smtp_banner uses $smtp_active_hostname instead
2629 of $primary_hostname.
2631 60. The host_aton() function is supposed to be passed a string that is known
2632 to be a valid IP address. However, in the case of IPv6 addresses, it was
2633 not checking this. This is a hostage to fortune. Exim now panics and dies
2634 if the condition is not met. A case was found where this could be provoked
2635 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2636 components; fortuitously, this particular loophole had already been fixed
2637 by change 4.50/55 above.
2639 If there are any other similar loopholes, the new check in host_aton()
2640 itself should stop them being exploited. The report I received stated that
2641 data on the command line could provoke the exploit when Exim was running as
2642 exim, but did not say which command line option was involved. All I could
2643 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2644 running as the user.
2647 61. There was a buffer overflow vulnerability in the SPA authentication code
2648 (which came originally from the Samba project). I have added a test to the
2649 spa_base64_to_bits() function which I hope fixes it.
2652 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2653 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2655 63. The daemon start-up calls getloadavg() while still root for those OS that
2656 need the first call to be done as root, but it missed one case: when
2657 deliver_queue_load_max is set with deliver_drop_privilege. This is
2658 necessary for the benefit of the queue runner, because there is no re-exec
2659 when deliver_drop_privilege is set.
2661 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2662 This has been fixed.
2664 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2665 was in use, was not putting the data itself into the right store pool;
2666 consequently, it could be overwritten for a subsequent message in the same
2667 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2670 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2672 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2673 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2674 the function were treating the return as a boolean value, which happened to
2675 work because 0=false and not-0=true, but is not correct code.
2677 68. The host_aton() function was not handling scoped IPv6 addresses (those
2678 with, for example, "%eth0" on the end) correctly.
2680 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2681 negated items (that is, ~something) in unsigned ints. Some compilers
2682 apparently mutter when there is no cast.
2684 70. If an address verification called from an ACL failed, and did not produce a
2685 user-specific message (i.e. there was only a "system" message), nothing was
2686 put in $acl_verify_message. In this situation, it now puts the system
2689 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2690 session; change 4.31/43 added the unwanted input to the log line - except
2691 that it did not do this in the start of session case. It now does.
2693 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2694 This is wrong and can cause the other end to generate a synchronization
2695 error if it is another Exim or anything else that does the synchronization
2696 check. A QUIT command is no longer sent after a timeout.
2698 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2699 during host lookups.
2701 74. The defer_ok option of callout verification was not working if it was used
2702 when verifying addresses in header lines, that is, for this case:
2704 verify = header_sender/callout=defer_ok
2706 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2707 those file descriptors could be used for SMTP connections. If anything
2708 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2709 could be sent to the SMTP client, causing chaos. The daemon now opens
2710 stdin, stdout, and stderr to /dev/null when it puts itself into the
2713 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2714 log by default. The user can override this with suitable Perl magic.
2716 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2717 the log message when discard triggers, was not working for the DATA ACL or
2718 for the non-SMTP ACL.
2720 78. Error message wording change in sieve.c.
2722 79. If smtp_accept_max_per_host was set, the number of connections could be
2723 restricted to fewer than expected, because the daemon was trying to set up
2724 a new connection before checking whether the processes handling previous
2725 connections had finished. The check for completed processes is now done
2726 earlier. On busy systems, this bug wouldn't be noticed because something
2727 else would have woken the daemon, and it would have reaped the completed
2730 80. If a message was submitted locally by a user whose login name contained one
2731 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2732 It caused a spool format error. I have fixed the spool reading code. A
2733 related problem was that the "from" clause in the Received: line became
2734 illegal because of the space(s). It is now covered by ${quote_local_part.
2736 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2739 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2741 83. Patch from Sieve maintainer to fix unterminated string problem in
2742 "vacation" handling.
2744 84. Some minor changes to the Linux configuration files to help with other
2745 OS variants using glibc.
2747 85. One more patch for Sieve to update vacation handling to latest spec.
2750 ----------------------------------------------------
2751 See the note above about the 4.44 and 4.50 releases.
2752 ----------------------------------------------------
2758 1. Change 4.43/35 introduced a bug that caused file counts to be
2759 incorrectly computed when quota_filecount was set in an appendfile
2762 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2763 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2766 3. Give more explanation in the error message when the command for a transport
2767 filter fails to execute.
2769 4. There are several places where Exim runs a non-Exim command in a
2770 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2771 was being done only for the command run by the queryprogram router. It is
2772 now done for all such subprocesses. The other cases are: ${run, transport
2773 filters, and the commands run by the lmtp and pipe transports.
2775 5. Some older OS have a limit of 256 on the maximum number of file
2776 descriptors. Exim was using setrlimit() to set 1000 as a large value
2777 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2778 systems. I've change it so that if it can't get 1000, it tries for 256.
2780 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2781 was an oversight, and furthermore, ever since the addition of extra
2782 controls (e.g. 4.43/32), the checks on when to allow different forms of
2783 "control" were broken. There should now be diagnostics for all cases when a
2784 control that does not make sense is encountered.
2786 7. $recipients is now available in the predata ACL (oversight).
2788 8. Tidy the search cache before the fork to do a delivery from a message
2789 received from the command line. Otherwise the child will trigger a lookup
2790 failure and thereby defer the delivery if it tries to use (for example) a
2791 cached ldap connection that the parent has called unbind on.
2793 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2794 of $address_data from the recipient verification was clobbered by the
2795 sender verification.
2797 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2798 was its contents. (It was OK if the option was not defined at all.)
2800 11. A "Completed" log line is now written for messages that are removed from
2801 the spool by the -Mrm option.
2803 12. $host_address is now set to the target address during the checking of
2804 ignore_target_hosts.
2806 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2807 being passed; this would have caused $sender_host_name to have been used if
2808 matching the list had actually called for a host name (not very likely,
2809 since this list is usually IP addresses). A host name is now passed as
2812 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2813 code that specifies a non-restarting handler (typically sigaction() in
2814 modern systems) in an attempt to fix a rare and obscure crash bug.
2816 15. Narrowed the window for a race in the daemon that could cause it to ignore
2817 SIGCHLD signals. This is not a major problem, because they are used only to
2818 wake it up if nothing else does.
2820 16. A malformed maildirsize file could cause Exim to calculate negative values
2821 for the mailbox size or file count. Odd effects could occur as a result.
2822 The maildirsize information is now recalculated if the size or filecount
2825 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2826 support for a long time. Removed HAVE_SYS_VFS_H.
2828 18. Updated exipick to current release from John Jetmore.
2830 19. Allow an empty sender to be matched against a lookup in an address list.
2831 Previously the only cases considered were a regular expression, or an
2834 20. Exim went into a mad DNS lookup loop when doing a callout where the
2835 host was specified on the transport, if the DNS lookup yielded more than
2838 21. The RFC2047 encoding function was originally intended for short strings
2839 such as real names; it was not keeping to the 75-character limit for
2840 encoded words that the RFC imposes. It now respects the limit, and
2841 generates multiple encoded words if necessary. To be on the safe side, I
2842 have increased the buffer size for the ${rfc2047: expansion operator from
2845 22. Failure to deliver a bounce message always caused it to be frozen, even if
2846 there was an errors_to setting on the router. The errors_to setting is now
2849 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2850 canonical form (fully expanded) before being placed in
2851 $sender_host_address.
2853 24. Updated eximstats to version 1.33
2855 25. Include certificate and key file names in error message when GnuTLS fails
2856 to set them up, because the GnuTLS error message doesn't include the name
2857 of the failing file when there is a problem reading it.
2859 26. Expand error message when OpenSSL has problems setting up cert/key files.
2862 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2863 (this can affect the format of dates).
2865 28. exim_tidydb, when checking for the continued existence of a message for
2866 which it has found a message-specific retry record, was not finding
2867 messages that were in split spool directories. Consequently, it was
2868 deleting retry records that should have stayed in existence.
2870 29. eximstats updated to version 1.35
2871 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2872 1.35 - bugfix such that pie charts by volume are generated correctly
2874 30. The SPA authentication driver was not abandoning authentication and moving
2875 on to the next authenticator when an expansion was forced to fail,
2876 contradicting the general specification for all authenticators. Instead it
2877 was generating a temporary error. It now behaves as specified.
2879 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2880 (the order specifies the preference for clients). The order is now AES256,
2881 AES128, 3DES, ARCFOUR128.
2883 31. Small patch to Sieve code - explicitly set From: when generating an
2886 32. Exim crashed if a remote delivery caused a very long error message to be
2887 recorded - for instance if somebody sent an entire SpamAssassin report back
2888 as a large number of 550 error lines. This bug was coincidentally fixed by
2889 increasing the size of one of Exim's internal buffers (big_buffer) that
2890 happened as part of the Exiscan merge. However, to be on the safe side, I
2891 have made the code more robust (and fixed the comments that describe what
2894 33. Some experimental protocols are using DNS PTR records for new purposes. The
2895 keys for these records are domain names, not reversed IP addresses. The
2896 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2897 leaves it alone. Component reversal etc. now happens only for IP addresses.
2900 34. The host_aton() function is supposed to be passed a string that is known
2901 to be a valid IP address. However, in the case of IPv6 addresses, it was
2902 not checking this. This is a hostage to fortune. Exim now panics and dies
2903 if the condition is not met. A case was found where this could be provoked
2904 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2905 components; fortuitously, this particular loophole had already been fixed
2906 by change 4.50/55 or 4.44/33 above.
2908 If there are any other similar loopholes, the new check in host_aton()
2909 itself should stop them being exploited. The report I received stated that
2910 data on the command line could provoke the exploit when Exim was running as
2911 exim, but did not say which command line option was involved. All I could
2912 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2913 running as the user.
2916 35. There was a buffer overflow vulnerability in the SPA authentication code
2917 (which came originally from the Samba project). I have added a test to the
2918 spa_base64_to_bits() function which I hope fixes it.
2921 36. The daemon start-up calls getloadavg() while still root for those OS that
2922 need the first call to be done as root, but it missed one case: when
2923 deliver_queue_load_max is set with deliver_drop_privilege. This is
2924 necessary for the benefit of the queue runner, because there is no re-exec
2925 when deliver_drop_privilege is set.
2927 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2928 was in use, was not putting the data itself into the right store pool;
2929 consequently, it could be overwritten for a subsequent message in the same
2930 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2933 38. Sometimes the final signoff response after QUIT could fail to get
2934 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2935 < 0 before doing a fflush(). This bug looks as though it goes back to the
2936 introduction of TLS in release 3.20, but "sometimes" must have been rare
2937 because the tests only now provoked it.
2943 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2944 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2945 Now it returns an integer. A number of calls were still expecting a T/F
2946 return. Fortuitously, in all cases, the tests worked in OK situations,
2947 which is the norm. However, things would have gone wrong on any write
2948 failures on the smtp file descriptor. This function is used when sending
2949 messages over SMTP and also when doing verify callouts.
2951 2. When Exim is called to do synchronous delivery of a locally submitted
2952 message (the -odf or -odi options), it no longer closes stderr before doing
2955 3. Implemented the mua_wrapper option.
2957 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2959 5. Implemented the functions header_remove(), header_testname(),
2960 header_add_at_position(), and receive_remove_recipient(), and exported them
2963 6. If an ACL "warn" statement specified the addition of headers, Exim already
2964 inserted X-ACL-Warn: at the start if there was no header name. However, it
2965 was not making this test for the second and subsequent header lines if
2966 there were newlines in the string. This meant that an invalid header could
2967 be inserted if Exim was badly configured.
2969 7. Allow an ACL "warn" statement to add header lines at the start or after all
2970 the Received: headers, as well as at the end.
2972 8. Added the rcpt_4xx retry error code.
2974 9. Added postmaster_mailfrom=xxx to callout verification option.
2976 10. Added mailfrom=xxxx to the callout verification option, for verify=
2979 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2980 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2982 12. Inserted some casts to stop certain compilers warning when using pointer
2983 differences as field lengths or precisions in printf-type calls (mostly
2984 affecting debugging statements).
2986 13. Added optional readline() support for -be (dynamically loaded).
2988 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2989 same clock tick as a message's arrival, so that its received time was the
2990 same as the "first fail" time on the retry record, and that message
2991 remained on the queue past the ultimate address timeout, every queue runner
2992 would try a delivery (because it was past the ultimate address timeout) but
2993 after another failure, the ultimate address timeout, which should have then
2994 bounced the address, did not kick in. This was a "< instead of <=" error;
2995 in most cases the first failure would have been in the next clock tick
2996 after the received time, and all would be well.
2998 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2999 being recognized when the domain list was tested by the match_domain
3000 condition in an expansion string.
3002 16. Added the ${str2b64: operator.
3004 17. Exim was always calling setrlimit() to set a large limit for the number of
3005 processes, without checking whether the existing limit was already
3006 adequate. (It did check for the limit on file descriptors.) Furthermore,
3007 errors from getrlimit() and setrlimit() were being ignored. Now they are
3008 logged to the main and panic logs, but Exim does carry on, to try to do its
3009 job under whatever limits there are.
3011 18. Imported PCRE 5.0.
3013 19. Trivial typo in log message " temporarily refused connection" (the leading
3016 20. If the log selector return_path_on_delivery was set and an address was
3017 redirected to /dev/null, the delivery process crashed because it assumed
3018 that a return path would always be set for a "successful" delivery. In this
3019 case, the whole delivery is bypassed as an optimization, and therefore no
3022 21. Internal re-arrangement: the function for sending a challenge and reading
3023 a response while authentication was assuming a zero-terminated challenge
3024 string. It's now changed to take a pointer and a length, to allow for
3025 binary data in such strings.
3027 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3029 23. Exim was not respecting finduser_retries when seeking the login of the
3030 uid under which it was called; it was always trying 10 times. (The default
3031 setting of finduser_retries is zero.) Also, it was sleeping after the final
3032 failure, which is pointless.
3034 24. Implemented tls_on_connect_ports.
3036 25. Implemented acl_smtp_predata.
3038 26. If the domain in control=submission is set empty, Exim assumes that the
3039 authenticated id is a complete email address when it generates From: or
3040 Sender: header lines.
3042 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3043 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3044 chown and chgrp in /bin and hostname in /usr/bin.
3046 28. Exim was keeping the "process log" file open after each use, just as it
3047 does for the main log. This opens the possibility of it remaining open for
3048 long periods when the USR1 signal hits a daemon. Occasional processlog
3049 errors were reported, that could have been caused by this. Anyway, it seems
3050 much more sensible not to leave this file open at all, so that is what now
3053 29. The long-running daemon process does not normally write to the log once it
3054 has entered its main loop, and it closes the log before doing so. This is
3055 so that log files can straightforwardly be renamed and moved. However,
3056 there are a couple of unusual error situations where the daemon does write
3057 log entries, and I had neglected to close the log afterwards.
3059 30. The text of an SMTP error response that was received during a remote
3060 delivery was being truncated at 512 bytes. This is too short for some of
3061 the long messages that one sometimes sees. I've increased the limit to
3064 31. It is now possible to make retry rules that apply only when a message has a
3065 specific sender, in particular, an empty sender.
3067 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3068 it possible to be selective about when SMTP synchronization is enforced.
3070 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3072 32. Implemented hosts_connection_nolog.
3074 33. Added an ACL for QUIT.
3076 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3079 35. Added mailbox_size and mailbox_filecount to appendfile.
3081 36. Added control = no_multiline_responses to ACLs.
3083 37. There was a bug in the logic of the code that waits for the clock to tick
3084 in the case where the clock went backwards by a substantial amount such
3085 that the microsecond fraction of "now" was more than the microsecond
3086 fraction of "then" (but the whole seconds number was less).
3088 38. Added support for the libradius Radius client library this is found on
3089 FreeBSD (previously only the radiusclient library was supported).
3095 1. When certain lookups returned multiple values in the form name=value, the
3096 quoting of the values was not always being done properly. Specifically:
3097 (a) If the value started with a double quote, but contained no whitespace,
3099 (b) If the value contained whitespace other than a space character (i.e.
3100 tabs or newlines or carriage returns) it was not quoted.
3101 This fix has been applied to the mysql and pgsql lookups by writing a
3102 separate quoting function and calling it from the lookup code. The fix
3103 should probably also be applied to nisplus, ibase and oracle lookups, but
3104 since I cannot test any of those, I have not disturbed their existing code.
3106 2. A hit in the callout cache for a specific address caused a log line with no
3107 reason for rejecting RCPT. Now it says "Previous (cached) callout
3108 verification failure".
3110 3. There was an off-by-one bug in the queryprogram router. An over-long
3111 return line was truncated at 256 instead of 255 characters, thereby
3112 overflowing its buffer with the terminating zero. As well as fixing this, I
3113 have increased the buffer size to 1024 (and made a note to document this).
3115 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3116 when Exim is waiting for an SMTP response from a remote server, Exim
3117 restarts its select() call on the socket, thereby resetting its timeout.
3118 This is not a problem when such interrupts are rare. Somebody set up a cron
3119 job to run exiwhat every 2 minutes, which is less than the normal select()
3120 timeout (5 or 10 minutes). This meant that the select() timeout never
3121 kicked in because it was always reset. I have fixed this by comparing the
3122 time when an interrupt arrives with the time at the start of the first call
3123 to select(). If more time than the timeout has elapsed, the interrupt is
3124 treated as a timeout.
3126 5. Some internal re-factoring in preparation for the addition of Sieve
3127 extensions (by MH). In particular, the "personal" test is moved to a
3128 separate function, and given an option for scanning Cc: and Bcc: (which is
3129 not set for Exim filters).
3131 6. When Exim created an email address using the login of the caller as the
3132 local part (e.g. when creating a From: or Sender: header line), it was not
3133 quoting the local part when it contained special characters such as @.
3135 7. Installed new OpenBSD configuration files.
3137 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3138 try to make them clearer.
3140 9. Callout options, other than the timeout value, were being ignored when
3141 verifying sender addresses in header lines. For example, when using
3143 verify = header_sender/callout=no_cache
3145 the cache was (incorrectly) being used.
3147 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3148 only the Cygwin environment.
3150 11. When return_path_on_delivery was set as a log selector, if different remote
3151 addresses in the same message used different return paths and parallel
3152 remote delivery occurred, the wrong values would sometimes be logged.
3153 (Whenever a remote delivery process finished, the return path value from
3154 the most recently started remote delivery process was logged.)
3156 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3157 lines when AUTH and/or TLS are in use. This is the "received protocol"
3158 field. Exim used to use "asmtp" for authenticated SMTP, without any
3159 indication (in the protocol name) for TLS use. Now it follows the RFC and
3160 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3161 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3162 names appear in log lines as well as in Received: header lines.
3164 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3165 extensions, and comparison tests, and to fix some bugs.
3167 14. Changes to the "personal" filter test:
3169 (1) The test was buggy in that it was just doing the equivalent of
3170 "contains" tests on header lines. For example, if a user's address was
3171 anne@some.where, the "personal" test would incorrectly be true for
3173 To: susanne@some.where
3175 This test is now done by extracting each address from the header in turn,
3176 and checking the entire address. Other tests that are part of "personal"
3177 are now done using regular expressions (for example, to check local parts
3178 of addresses in From: header lines).
3180 (2) The list of non-personal local parts in From: addresses has been
3181 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3182 taken from the Sieve specification recommendations.
3184 (3) If the message contains any header line starting with "List-" it is
3185 treated as non-personal.
3187 (4) The test for "circular" in the Subject: header line has been removed
3188 because it now seems ill-conceived.
3190 15. Minor typos in src/EDITME comments corrected.
3192 16. Installed latest exipick from John Jetmore.
3194 17. If headers_add on a router specified a text string that was too long for
3195 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3196 of string_sprintf() is now avoided.
3198 18. $message_body_size was not set (it was always zero) when running the DATA
3199 ACL and the local_scan() function.
3201 19. For the "mail" command in an Exim filter, no default was being set for
3202 the once_repeat time, causing a random time value to be used if "once" was
3203 specified. (If the value happened to be <= 0, no repeat happened.) The
3204 default is now 0s, meaning "never repeat". The "vacation" command was OK
3205 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3206 (I found it when inspecting the code).
3208 20. There is now an overall timeout for performing a callout verification. It
3209 defaults to 4 times the callout timeout, which applies to individual SMTP
3210 commands during the callout. The overall timeout applies when there is more
3211 than one host that can be tried. The timeout is checked before trying the
3212 next host. This prevents very long delays if there are a large number of
3213 hosts and all are timing out (e.g. when the network connections are timing
3214 out). The value of the overall timeout can be changed by specifying an
3215 additional sub-option for "callout", called "maxwait". For example:
3217 verify = sender/callout=5s,maxwait=20s
3219 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3220 to the end before writing, but this should make it even safer).
3222 22. Exim was forgetting that it had advertised PIPELINING for the second and
3223 subsequent messages on an SMTP connection. It was also not resetting its
3224 memory on STARTTLS and an internal HELO.
3226 23. When Exim logs an SMTP synchronization error within a session, it now
3227 records whether PIPELINING has been advertised or not.
3229 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3230 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3231 rather than long int.
3233 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3235 26. Added the never_mail option to autoreply.
3241 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3242 crash if the getsockname() call failed; this can happen if a connection is
3243 closed very soon after it is established. The problem was simply in the
3244 order in which certain operations were done, causing Exim to try to write
3245 to the SMTP stream before it had set up the file descriptor. The bug has
3246 been fixed by making things happen in the correct order.
3252 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3253 before the connection was closed, thus losing the rejection response.
3255 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3256 some early Solaris releases, but causes trouble in current releases where
3257 socklen_t is defined.
3259 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3262 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3265 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3266 that releasing the top of it at the end releases what was used for sub-
3267 expansions (unless the block got too big). However, discard this block if
3268 the first thing is a variable or header, so that we can use its block when
3269 it is dynamic (useful for very large $message_headers, for example).
3271 6. Lookups now cache *every* query, not just the most recent. A new, separate
3272 store pool is used for this. It can be recovered when all lookup caches are
3273 flushed. Lookups now release memory at the end of their result strings.
3274 This has involved some general refactoring of the lookup sources.
3276 7. Some code has been added to the store_xxx() functions to reduce the amount
3277 of flapping under certain conditions.
3279 8. log_incoming_interface used to affect only the <= reception log lines. Now
3280 it causes the local interface and port to be added to several more SMTP log
3281 lines, for example "SMTP connection from", and rejection lines.
3283 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3285 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3287 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3288 could be overwritten at the end of the current message (or the start of a
3289 new message if it was set in a HELO ACL). The value is now preserved for
3290 the duration of the SMTP connection.
3292 12. If a transport had a headers_rewrite setting, and a matching header line
3293 contained an unqualified address, that address was qualified, even if it
3294 did not match any rewriting rules. The underlying bug was that the values
3295 of the flags that permit the existence of unqualified sender and recipient
3296 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3297 non-local messages, and by -bnq for local messages) were not being
3298 preserved with the message after it was received.
3300 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3301 "next input=" as part of the text comprising the host identity instead of
3302 the correct text. The code was using the same buffer for two different
3303 strings. However, depending on which order the printing function evaluated
3304 its arguments, the bug did not always show up. Under Linux, for example, my
3305 test suite worked just fine.
3307 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3308 doesn't work with some older versions of Perl. It has been changed to "my",
3309 which in any case is probably the better facility to use.
3311 15. A really picky compiler found some instances of statements for creating
3312 error messages that either had too many or two few arguments for the format
3315 16. The size of the buffer for calls to the DNS resolver has been increased
3316 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3317 for addresses that have a lot of PTR records. This alleviates a problem; it
3318 does not fully solve it.
3320 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3321 buffer now truncates the list and logs the incident, which is the same
3322 action as happens when Exim is looking up a host name and its aliases.
3323 Previously in this situation something unpredictable would happen;
3324 sometimes it was "internal error: store_reset failed".
3326 18. If a server dropped the connection unexpectedly when an Exim client was
3327 using GnuTLS and trying to read a response, the client delivery process
3328 crashed while trying to generate an error log message.
3330 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3331 string, for example:
3333 warn message = H1: something\nH2: something
3335 the text was added as a single header line from Exim's point of view
3336 though it ended up OK in the delivered message. However, searching for the
3337 second and subsequent header lines using $h_h2: did not work. This has been
3338 fixed. Similarly, if a system filter added multiple headers in this way,
3339 the routers could not see them.
3341 20. Expanded the error message when iplsearch is called with an invalid key to
3342 suggest using net-iplsearch in a host list.
3344 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3345 ACLs are no longer actually imposed (and a message to that effect is
3348 22. If a "gecos" field in a passwd entry contained escaped characters, in
3349 particular, if it contained a \" sequence, Exim got it wrong when building
3350 a From: or a Sender: header from that name. A second bug also caused
3351 incorrect handling when an unquoted " was present following a character
3352 that needed quoting.
3354 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3355 was not being matched caselessly.
3357 24. Arranged for all hyphens in the exim.8 source to be escaped with
3360 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3361 properly. Recipient callout cache records were still being keyed to include
3362 the sender, even when use_sender was set false. This led to far more
3363 callouts that were necessary. The sender is no longer included in the key
3364 when use_sender is false.
3366 26. Added "control = submission" modifier to ACLs.
3368 27. Added the ${base62d: operator to decode base 62 numbers.
3370 28. dnsdb lookups can now access SRV records.
3372 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3373 the configuration file.
3375 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3376 (-v) mode. This makes the output for a verbose queue run more intelligible.
3378 31. Added a use_postmaster feature to recipient callouts.
3380 32. Added the $body_zerocount variable, containing the number of binary zero
3381 bytes in the message body.
3383 33. The time of last modification of the "new" subdirectory is now used as the
3384 "mailbox time last read" when there is a quota error for a maildir
3387 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3389 35. Added +ignore_unknown as a special item in host lists.
3391 36. Code for decoding IPv6 addresses in host lists is now included, even if
3392 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3393 address was recognized as an IP address, but was then not correctly decoded
3394 into binary, causing unexpected and incorrect effects when compared with
3401 1. Very minor rewording of debugging text in manualroute to say "list of
3402 hosts" instead of "hostlist".
3404 2. If verify=header_syntax was set, and a header line with an unqualified
3405 address (no domain) and a large number of spaces between the end of the
3406 name and the colon was received, the reception process suffered a buffer
3407 overflow, and (when I tested it) crashed. This was caused by some obsolete
3408 code that should have been removed. The fix is to remove it!
3410 3. When running in the test harness, delay a bit after writing a bounce
3411 message to get a bit more predictability in the log output.
3413 4. Added a call to search_tidyup() just before forking a reception process. In
3414 theory, someone could use a lookup in the expansion of smtp_accept_max_
3415 per_host which, without the tidyup, could leave open a database connection.
3417 5. Added the variables $recipient_data and $sender_data which get set from a
3418 lookup success in an ACL "recipients" or "senders" condition, or a router
3419 "senders" option, similar to $domain_data and $local_part_data.
3421 6. Moved the writing of debug_print from before to after the "senders" test
3424 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3425 problems for message scanning, either using a data ACL, or using
3426 local_scan() because the Received: header was not generated till after they
3427 were called (in order to set the time as the time of reception completion).
3428 I have revised the way this works. The header is now generated after the
3429 body is received, but before the ACL or local_scan() are called. After they
3430 are run, the timestamp in the header is updated.
3436 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3437 before starting a queue runner without re-exec. This happened only when
3438 deliver_drop_privilege was set or when the Exim user was set to root. The
3439 effect of the bug was that timeouts during subsequent deliveries caused
3440 crashes instead of being properly handled. The handler is now left at its
3441 default (and expected) setting.
3443 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3444 message, again when deliver_drop_privilege is set or Exim is run as root.
3445 The bug described in (1) was not present in this case, but the tidying up
3446 of the other signals was missing. I have made the two cases consistent.
3448 3. The ignore_target_hosts setting on a manualroute router was being ignored
3449 for hosts that were looked up using the /MX notation.
3451 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3454 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3455 operated on the sender address. After changing the $sender_address to <>
3456 for the sender address verify, Exim was re-instated it as the original
3457 (before rewriting) address, but remembering that it had rewritten it, so it
3458 wasn't rewriting it again. This bug also had the effect of breaking the
3459 sender address verification caching when the sender address was rewritten.
3461 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3462 This has been changed so that if the ip literal address matches
3463 ignore_target_hosts, the router declines.
3465 7. Added expansion conditions match_domain, match_address, and match_local_
3466 part (NOT match_host).
3468 8. The placeholder for the Received: header didn't have a length field set.
3470 9. Added code to Exim itself and to exim_lock to test for a specific race
3471 condition that could lead to file corruption when using MBX delivery. The
3472 issue is with the lockfile that is created in /tmp. If this file is removed
3473 after a process has opened it but before that process has acquired a lock,
3474 there is the potential for a second process to recreate the file and also
3475 acquire a lock. This could lead to two Exim processes writing to the file
3476 at the same time. The added code performs the same test as UW imapd; it
3477 checks after acquiring the lock that its file descriptor still refers to
3478 the same named file.
3480 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3481 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3484 11. Added the smtp_active_hostname option. If used, this will typically be made
3485 to depend on the incoming interface address. Because $interface_address is
3486 not set up until the daemon has forked a reception process, error responses
3487 that can happen earlier (such as "too many connections") no longer contain
3490 12. If an expansion in a condition on a "warn" statement fails because a lookup
3491 defers, the "warn" statement is abandoned, and the next ACL statement is
3492 processed. Previously this caused the whole ACL to be aborted.
3494 13. Added the iplsearch lookup type.
3496 14. Added ident_timeout as a log selector.
3498 15. Added tls_certificate_verified as a log selector.
3500 16. Added a global option tls_require_ciphers (compare the smtp transport
3501 option of the same name). This controls incoming TLS connections.
3503 17. I finally figured out how to make tls_require_ciphers do a similar thing
3504 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3505 before starting the TLS session.
3507 18. Tabs are now shown as \t in -bP output.
3509 19. If the log selector return_path_on_delivery was set, Exim crashed when
3510 bouncing a message because it had too many Received: header lines.
3512 20. If two routers both had headers_remove settings, and the first one included
3513 a superfluous trailing colon, the final name in the first list and the
3514 first name in the second list were incorrectly joined into one item (with a
3515 colon in the middle).
3521 1. Added -C and -D options to the exinext utility, mainly to make it easier
3522 to include in the automated testing, but these could be helpful when
3523 multiple configurations are in use.
3525 2. The exinext utility was not formatting the output nicely when there was
3526 an alternate port involved in the retry record key, nor when there was a
3527 message id as well (for retries that were specific to a specific message
3528 and a specific host). It was also confused by IPv6 addresses, because of
3529 the additional colons they contain. I have fixed the IPv4 problem, and
3530 patched it up to do a reasonable job for IPv6.
3532 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3533 delivery, the log line now contains "pipelined" if PIPELINING was used.
3535 4. An SMTP transport process used to panic and die if the bind() call to set
3536 an explicit outgoing interface failed. This has been changed; it is now
3537 treated in the same way as a connect() failure.
3539 5. A reference to $sender_host_name in the part of a conditional expansion
3540 that was being skipped was still causing a DNS lookup. This no longer
3543 6. The def: expansion condition was not recognizing references to header lines
3544 that used bh_ and bheader_.
3546 7. Added the _cache feature to named lists.
3548 8. The code for checking quota_filecount in the appendfile transport was
3549 allowing one more file than it should have been.
3551 9. For compatibility with Sendmail, the command line option
3559 and sets the incoming protocol and host name (for trusted callers). The
3560 host name and its colon can be omitted when only the protocol is to be set.
3561 Note the Exim already has two private options, -pd and -ps, that refer to
3562 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3563 "s", but I don't think that's a major issue.
3565 10. A number of refactoring changes to the code, none of which should affect
3568 (a) The number of logging options was getting close to filling up the
3569 32-bit word that was used as a bit map. I have split them into two classes:
3570 those that are passed in the argument to log_write(), and those that are
3571 only ever tested independently outside of that function. These are now in
3572 separate 32-bit words, so there is plenty of room for expansion again.
3573 There is no change in the user interface or the logging behaviour.
3575 (b) When building, for example, log lines, the code previously used a
3576 macro that called string_cat() twice, in order to add two strings. This is
3577 not really sufficiently general. Furthermore, there was one instance where
3578 it was actually wrong because one of the argument was used twice, and in
3579 one call a function was used. (As it happened, calling the function twice
3580 did not affect the overall behaviour.) The macro has been replaced by a
3581 function that can join an arbitrary number of extra strings onto a growing
3584 (c) The code for expansion conditions now uses a table and a binary chop
3585 instead of a serial search (which was left over from when there were very
3586 few conditions). Also, it now recognizes conditions like "pam" even when
3587 the relevant support is not compiled in: a suitably worded error message is
3588 given if an attempt is made to use such a condition.
3590 11. Added ${time_interval:xxxxx}.
3592 12. A bug was causing one of the ddress fields not to be passed back correctly
3593 from remote delivery subprocesses. The field in question was not being
3594 subsequently used, so this caused to problems in practice.
3596 13. Added new log selectors queue_time and deliver_time.
3598 14. Might have fixed a bug in maildirsizefile handling that threw up
3599 "unexpected character" debug warnings, and recalculated the data
3600 unnecessarily. In any case, I expanded the warning message to give more
3603 15. Added the message "Restricted characters in address" to the statements in
3604 the default ACL that block characters like @ and % in local parts.
3606 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3607 Three changes have been made:
3609 (a) There was a serious bug; a negative response to MAIL caused the whole
3610 recipient domain to be cached as invalid, thereby blocking all messages
3611 to all local parts at the same domain, from all senders. This bug has
3612 been fixed. The domain is no longer cached after a negative response to
3613 MAIL if the sender used is not empty.
3615 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3618 (c) A new callout option, "use_sender" has been added for people who want
3619 the modified behaviour.
3625 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3628 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3629 indeed breaks things for older releases.
3631 3. Added additional logging to the case where there is a problem reading data
3632 from a filter that is running in a subprocess using a pipe, in order to
3633 try to track down a specific problem.
3635 4. Testing facility fudge: when running in the test harness and attempting
3636 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3637 getting "No route to host". Convert this to a timeout.
3639 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3642 6. Some OS don't have socklen_t but use size_t instead. This affects the
3643 fifth argument of getsockopt() amongst other things. This is now
3644 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3645 can be set for individual OS. I have set it for SunOS5, OSF1, and
3646 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3647 some earlier ones do not.
3649 7. Change 4.30/15 was not doing the test caselessly.
3651 8. The standard form for an IPv6 address literal was being rejected by address
3652 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3653 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3654 this, as well as the form without the "IPv6" on the front (but only when
3655 address literals are enabled, of course).
3657 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3659 10. Exim crashed if a message with an empty sender address specified by -f
3660 encountered a router with an errors_to setting. This could be provoked only
3661 by a command such as
3665 where an empty string was supplied; "<>" did not hit this bug.
3667 11. Installed PCRE release 4.5.
3669 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3670 remained set. It is now erased.
3672 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3673 times from message ids (which are base 36 rather than the normal 62).
3675 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3676 were being counted as actual protocol errors, and logged if the log
3677 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3678 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3679 and DATA following a set of rejected RCPTs do not count as protocol errors.
3680 In other words, Exim assumes they were pipelined, though this may not
3681 actually be the case. Of course, in all cases the client gets an
3682 appropriate error code.
3684 15. If a lookup fails in an ACL condition, a message about the failure may
3685 be available; it is used if testing the ACL cannot continue, because most
3686 such messages specify what the cause of the deferral is. However, some
3687 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3688 that caused an old message to be retained and used if a later statement
3689 caused a defer, replacing the real cause of the deferral.
3691 16. If an IP address had so many PTR records that the DNS lookup buffer
3692 was not large enough to hold them, Exim could crash while trying to process
3693 the truncated data. It now detects and logs this case.
3695 17. Further to 4.21/58, another change has been made: if (and only if) the
3696 first line of a message (the first header line) ends with CRLF, a bare LF
3697 in a subsequent header line has a space inserted after it, so as not to
3698 terminate the header.
3700 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3701 unnecessarily, used atoi() instead of strtol(), and didn't check the
3702 termination when getting file sizes from file names by regex.
3704 19. Completely re-implemented the support for maildirsize files, in the light
3705 of a number of problems with the previous contributed implementation
3706 (4.30/29). In particular:
3708 . If the quota is zero, the maildirsize file is maintained, but no quota is
3711 . If the maildir directory does not exist, it is created before any attempt
3712 to write a maildirsize file.
3714 . The quota value in the file is just a cache; if the quota is changed in
3715 the transport, the new value overrides.
3717 . A regular expression is available for excluding directories from the
3720 20. The autoreply transport checks the characters in options that define the
3721 message's headers; it allows continued headers, but it was checking with
3722 isspace() after an embedded newline instead of explicitly looking for a
3725 21. If all the "regular" hosts to which an address was routed had passed their
3726 expiry times, and had not reached their retry times, the address was
3727 bounced, even if fallback hosts were defined. Now Exim should go on to try
3730 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3731 equivalent code in the SMTP transport. Some hosts send humungous responses
3732 to HELO/EHLO, more than 1024 it seems.
3734 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3735 gives compiler warnings in some environments. I've now done it "properly",
3738 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3739 (because of problems with the built-in one) was declared to return uschar *
3740 instead of char *, causing compiler failure.
3742 25. Fixed a file descriptor leak when processing alias/forward files.
3744 26. Fixed a minor format string issue in dbfn.c.
3746 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3748 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3749 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3752 29. When a sender address is verified, it is cached, to save repeating the test
3753 when there is more than one recipient in a message. However, when the
3754 verification involves a callout, it is possible for different callout
3755 options to be set for different recipients. It is too complicated to keep
3756 track of this in the cache, so now Exim always runs a verification when a
3757 callout is required, relying on the callout cache for the optimization.
3758 The overhead is duplication of the address routing, but this should not be
3761 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3762 to be verified with callout=postmaster, and the main callout worked but the
3763 postmaster check failed, the verification correctly failed. However, if a
3764 subsequent RCPT command asked for sender verification *without* the
3765 postmaster check, incorrect caching caused this verification also to fail,
3768 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3769 it was not caching the DNS options (qualify_single, search_parents) that
3770 were used when the lookup failed. A subsequent lookup with different
3771 options therefore always gave the same answer, though there were cases
3772 where it should not have. (Example: a "domains = !$mx_any" option on a
3773 dnslookup router: the "domains" option is always processed without any
3774 widening, but the router might have qualify_single set.) Now Exim uses the
3775 cached value only when the same options are set.
3777 32. Added John Jetmore's "exipick" utility to the distribution.
3779 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3780 than a timeout (e.g. a certificate is required, and is not provided), an
3781 Exim server now closes the connection immediately. Previously it waited for
3782 the client to close - but if the client is SSL, it seems that they each
3783 wait for each other, leading to a delay before one of them times out.
3785 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3786 maintained 0.8.x compatibility because I don't think many are using it, and
3787 it is clearly obsolete.
3789 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3792 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3793 client certificate was expired. A simple patch fixes this, though I don't
3794 understand the full logic of why the verify callback is called multiple
3797 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3798 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3799 which causes problems with some clients (such as the Certicom SSL Plus
3800 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3801 disables the coutermeasure allowing Eudora to connect."
3803 38. Exim was not checking that a write() to a log file succeeded. This could
3804 lead to Bad Things if a log got too big, in particular if it hit a file
3805 size limit. Exim now panics and dies if it cannot write to a log file, just
3806 as it does if it cannot open a log file.
3808 39. Modified OS/Makefile-Linux so that it now contains
3810 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3812 The two -D definitions ensure that Exim is compiled with large file
3813 support, which makes it possible to handle log files that are bigger than
3816 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3817 instance) a domain was checked against a named list that involved a lookup,
3818 causing $domain_data to be set, then another domain was checked against the
3819 same list, then the first domain was re-checked, the value of $domain_data
3820 after the final check could be wrong. In particular, if the second check
3821 failed, it could be set empty. This bug probably also applied to
3824 41. The strip_trailing_dot option was not being applied to the address given
3825 with the -f command-line option.
3827 42. The code for reading a message's header from the spool was incrementing
3828 $received_count, but never initializing it. This meant that the value was
3829 incorrect (doubled) while delivering a message in the same process in which
3830 it was received. In the most common configuration of Exim, this never
3831 happens - a fresh exec is done - but it can happen when
3832 deliver_drop_privilege is set.
3834 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3835 it now includes up to 150 characters of the unexpected data in the log
3838 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3839 and building data strings. The size of both of these buffers was 10 000
3840 bytes - far larger than anybody would *ever* want, thought I. Needless to
3841 say, somebody hit the limit. I have increased the maximum line length to
3842 20 000 and the maximum data length of concatenated lines to 100 000. I have
3843 also fixed two bugs, because there was no checking on these buffers. Tsk,
3844 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3845 buffer is too small.
3847 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3848 lsearch lookups. Now it does.
3850 46. When parsing a route_list item in a manualroute router, a fixed-length
3851 buffer was used for the list of hosts. I made this 1024 bytes long,
3852 thinking that nobody would ever have a list of hosts that long. Wrong.
3853 Somebody had a whole pile of complicated expansion conditions, and the
3854 string was silently truncated, leading to an expansion error. It turns out
3855 that it is easier to change to an unlimited length (owing to other changes
3856 that have happened since this code was originally written) than to build
3857 structure for giving a limitation error. The length of the item that
3858 expands into the list of hosts is now unlimited.
3860 47. The lsearch lookup could not handle data where the length of text line was
3861 more than 4095 characters. Such lines were truncated, leading to shortened
3862 data being returned. It should now handle lines of any length.
3864 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3865 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3868 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3869 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3870 no longer happens in comment lines. A list of replacements is now placed
3871 at the head of all of the source files, except those whose only change is
3872 to replace PERL_COMMAND in the very first #! line.
3874 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3875 messages on the queue, with a bottom-up merge sort, using code contributed
3876 by Michael Haardt. This should make operations like -bp somewhat faster on
3877 large queues. It won't affect queue runners, except when queue_run_in_order
3880 51. Installed eximstats 1.31 in the distribution.
3882 52. Added support for SRV lookups to the dnslookup router.
3884 53. If an ACL referred to $message_body or $message_body_end, the value was not
3885 reset for any messages that followed in the same SMTP session.
3887 54. The store-handling optimization for building very long strings was not
3888 differentiating between the different store pools. I don't think this
3889 actually made any difference in practice, but I've tidied it.
3891 55. While running the routers to verify a sender address, $sender_address
3892 was still set to the sender address. This is wrong, because when routing to
3893 send a bounce to the sender, it would be empty. Therefore, I have changed
3894 it so that, while verifying a sender address, $sender_address is set to <>.
3895 (There is no change to what happens when verifying a recipient address.)
3897 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3898 target A or AAAA records (if not already returned) without resetting the
3899 qualify_single or search_parents options of the DNS resolver. These are
3900 inappropriate in this case because the targets of MX and SRV records must
3901 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3902 target that, when qualified, matched something in the local domain. These
3903 two options are now turned off when doing these lookups.
3905 57. It seems that at least some releases of Reiserfs (which does not have the
3906 concept of a fixed number of inodes) returns zero and not -1 for the
3907 number of available inodes. This interacted badly with check_spool_inodes,
3908 which assumed that -1 was the "no such thing" setting. What I have done is
3909 to check that the total number of inodes is greater than zero before doing
3910 the test of how many are available.
3912 58. When a "warn" ACL statement has a log_message modifier, the message is
3913 remembered, and not repeated. This is to avoid a lot of repetition when a
3914 message has many recipients that cause the same warning to be written.
3915 Howewer, Exim was preserving the list of already written lines for an
3916 entire SMTP session, which doesn't seem right. The memory is now reset if a
3917 new message is started.
3919 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3920 debugging output unless log_rewrite was also set.
3922 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3923 of (int)(handle) when we know that handle contains (void *)(-1).
3925 61. The Exim daemon panic-logs an error return when it closes the incoming
3926 connection. However "connection reset by peer" seems to be common, and
3927 isn't really an error worthy of noting specially, so that particular error
3930 62. When Exim is trying to find all the local interfaces, it used to panic and
3931 die if the ioctl to get the interface flags failed. However, it seems that
3932 on at least one OS (Solaris 9) it is possible to have an interface that is
3933 included in the list of interfaces, but for which you get a failure error
3934 for this call. This happens when the interface is not "plumbed" into a
3935 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3936 failure of the "get flags" call assumes that the interface is down.
3938 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3939 makes life easier for people who are doing arithmetic on fields extracted
3940 from dates, where you often get leading zeros that should not be
3941 interpreted as octal.
3943 64. Added qualify_domain to the redirect router, to override the global
3946 65. If a pathologically long header line contained very many addresses (the
3947 report of this problem mentioned 10 000) and each of them was rewritten,
3948 Exim could use up a very large amount of memory. (It kept on making new
3949 copies of the header line as it rewrote, and never released the old ones.)
3950 At the expense of a bit more processing, the header rewriting function has
3951 been changed so that it no longer eats memory in this way.
3953 66. The generation of the Received: header has been moved from the time that a
3954 message starts to be received, to the time that it finishes. The timestamp
3955 in the Received: header should now be very close to that of the <= log
3956 line. There are two side-effects of this change:
3958 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3959 logged header lines no longer include the local Received: line, because
3960 it has not yet been created. The same applies to a copy of the message
3961 that is returned to a non-SMTP sender when a message is rejected.
3963 (b) When a filter file is tested using -bf, no additional Received: header
3964 is added to the test message. After some thought, I decided that this
3967 This change does not affect the value of $received_for. It is still set
3968 after address rewriting, but before local_scan() is called.
3970 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3972 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3973 gave an unhelpful panic error message, and a defer error. I have managed to
3974 change this behaviour so that it now rejects any supplied certificate,
3975 which seems right, as the list of acceptable certificates is empty.
3977 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3978 gave an unhelpful defer error. I have not managed to make this reject any
3979 supplied certificates, but the error message it gives is "no certificate
3980 supplied", which is not helpful.
3982 70. exigrep's output now also includes lines that are not associated with any
3983 message, but which match the given pattern. Implemented by a patch from
3984 Martin Sluka, which also tidied up the Perl a bit.
3986 71. Recipient callout verification, like sender verification, was using <> in
3987 the MAIL FROM command. This isn't really the right thing, since the actual
3988 sender may affect whether the remote host accepts the recipient or not. I
3989 have changed it to use the actual sender in the callout; this means that
3990 the cache record is now keyed on a recipient/sender pair, not just the
3991 recipient address. There doesn't seem to be a real danger of callout loops,
3992 since a callout by the remote host to check the sender would use <>.
3993 [SEE ABOVE: changed after hitting problems.]
3995 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3996 temporary errors. However, in the case of such a code being given after
3997 the end of a data transmission (i.e. after ".") Exim was failing to write
3998 a retry record for the message. (Yes, there was some broken host that was
3999 actually sending 8xx at this point.)
4001 73. An unknown lookup type in a host list could cause Exim to panic-die when
4002 the list was checked. (An example that provoked this was putting <; in the
4003 middle of a list instead of at the start.) If this happened during a DATA
4004 ACL check, a -D file could be left lying around. This kind of configuration
4005 error no longer causes Exim to die; instead it causes a defer errror. The
4006 incident is still logged to the main and panic logs.
4008 74. Buglet left over from Exim 3 conversion. The message "too many messages
4009 in one connection" was written to the rejectlog but not the mainlog, except
4010 when address rewriting (yes!) was being logged.
4012 75. Added write_rejectlog option.
4014 76. When a system filter was run not as root (that is, when system_filter_user
4015 was set), the values of the $n variables were not being returned to the
4016 main process; thus, they were not subsequently available in the $sn
4019 77. Added +return_path_on_delivery log selector.
4021 78. A connection timeout was being treated differently from recipients deferred
4022 when testing hosts_max_try with a message that was older than the host's
4023 retry timeout. (The host should not be counted, thus allowing all hosts to
4024 be tried at least once before bouncing.) This may have been the cause of an
4025 occasionally reported bug whereby a message would remain on the queue
4026 longer than the retry timeout, but would be bounced if a delivery was
4027 forced. I say "may" because I never totally pinned down the problem;
4028 setting up timeout/retry tests is difficult. See also the next item.
4030 79. The ultimate address timeout was not being applied to errors that involved
4031 a combination of host plus message (for example, a timeout on a MAIL
4032 command). When an address resolved to a number of possible hosts, and they
4033 were not all tried for each delivery (e.g. because of hosts_max_try), a
4034 message could remain on the queue longer than the retry timeout.
4036 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4039 81. Fixed an obscure SMTP outgoing bug which required at least the following
4040 conditions: (a) there was another message waiting for the same server;
4041 (b) the server returned 5xx to all RCPT commands in the first message so
4042 that the message was not completed; (c) the server dropped the connection
4043 or gave a negative response to the RSET that Exim sends to abort the
4044 transaction. The observed case was a dropped connection after DATA that had
4045 been sent in pipelining mode. That is, the server had advertised PIPELINING
4046 but was not implementing it correctly. The effect of the bug was incorrect
4047 behaviour, such as trying another host, and this could lead to a crash.
4053 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4054 and daemon.c were passed as pointers to ints; they should have been
4055 pointers to socklen_t variables (which are typically unsigned ints).
4057 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4060 3. Fixed a really odd bug that affected only the testing scheme; patching a
4061 certain fixed string in the binary changed the value of another string that
4062 happened to be identical to the end of the original first string.
4064 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4065 name", it returns that address as the IP address. On some operating
4066 systems (e.g. Solaris), it also passes back the IP address string as the
4067 "host name". However, on others (e.g. Linux), it passes back an empty
4068 string. Exim wasn't checking for this, and was changing the host name to an
4069 empty string, assuming it had been canonicized.
4071 5. Although rare, it is permitted to have more than one PTR record for a given
4072 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4073 all the names associated with an address, because they do in Solaris.
4074 However, it seems that they do not in Linux for data that comes from the
4075 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4076 I found this out when I moved to a new Linux workstation and tried to run
4077 the Exim test suite.
4079 To get round this problem I have changed the code so that it now does its
4080 own call to the DNS to look up PTR records when searching for a host name.
4081 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4082 addresses that are only in /etc/hosts are still found.
4084 This behaviour is, however, controlled by an option called host_lookup_
4085 order, which defaults to "bydns:byaddr". If people want to use the other
4086 order, or indeed, just use one or the other means of lookup, they can
4087 specify it in this variable.
4089 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4090 some operating systems, this comes back from gethostbyaddr() as an empty
4091 string, and this is what Exim used to test for. However, it seems that in
4092 other systems, "." is yielded. Exim now tests for this case too.
4094 7. The values of check_spool_space and check_log_space are now held internally
4095 as a number of kilobytes instead of an absolute number of bytes. If a
4096 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4097 kilobyte. This means that much larger values can be stored.
4099 8. Exim monitor: an attempt to get the action menu when not actually pointing
4100 at a message produces an empty menu entitled "No message selected". This
4101 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4102 no entries in it ("Shell widget menu has zero width and/or height"). So I
4103 have added a single, blank menu entry in this case.
4105 9. Added ${quote_local_part.
4107 10. MIME decoding is now applied to the contents of Subject: header lines when
4110 11. Now that a reference to $sender_host_address automatically causes a reverse
4111 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4112 host lookup before query-style lookups in lists that might use this
4113 variable. This has therefore been abolished, and the "net-" prefix is no
4114 longer necessary for query-style lookups.
4116 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4117 have been a typo for LFLAGS, so it has been changed.
4119 13. The install script calls Exim with "-C /dev/null" in order to find the
4120 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4121 to be output. Howeve, since Exim outputs its version number before the
4122 error, it didn't break the script. It just looked ugly. I fixed this by
4123 always allowing "-C /dev/null" if the caller is root.
4125 14. Ignore overlarge ACL variable number when reading spool file - insurance
4126 against a later release with more variables having written the file.
4128 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4129 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4130 this, as well as the form without the "IPv6" on the front.
4132 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4133 OS/Makefile-Darwin file.
4135 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4136 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4138 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4139 variable. However, some OS use ints for inodes. I've added cast to long int
4140 to get rid of the compiler warning.
4142 19. I had forgotten to lock out "/../" in configuration file names when
4143 ALT_CONFIG_PREFIX was set.
4145 20. Routers used for verification do not need to specify transports. However,
4146 if such a router generated a host list, and callout was configured, Exim
4147 crashed, because it could not find a port number from the (non-existent)
4148 transport. It now assumes port 25 in this circumstance.
4150 21. Added the -t option to exigrep.
4152 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4153 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4154 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4155 them is set, LOOKUP_LSEARCH is forced.
4157 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4159 24. Added sender and host information to the "rejected by local_scan()" log
4160 line; previously there was no indication of these.
4162 25. Added .include_if_exists.
4164 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4165 Linux. It turns out that not all file systems support this. Apparently some
4166 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4167 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4168 not supported on the file descriptor, is now ignored when Exim is trying to
4169 sync a directory. This applies only to Linux.
4171 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4173 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4174 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4177 29. Added support for maildirsize files from supplied patch (modified a bit).
4179 30. The use of :fail: followed by an empty string could lead Exim to respond to
4180 sender verification failures with (e.g.):
4182 550 Verification failed for <xxx>
4183 550 Sender verify failed
4185 where the first response line was missing the '-' that indicates it is not
4186 the final line of the response.
4188 31. The loop for finding the name of the user that called Exim had a hardwired
4189 limit of 10; it now uses the value of finduser_retries, which is used for
4190 all other user lookups.
4192 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4195 33. Exim was neglecting to zero errno before one call of strtol() when
4196 expanding a string and expecting an integer value. On some systems this
4197 resulted in spurious "integer overflow" errors. Also, it was casting the
4198 result into an int without checking.
4200 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4201 did not work. The code looks as if it has *never* worked, though it appears
4202 to have been documented since at least releast 1.62. I have made it work.
4204 35. The "timeout_DNS" error in retry rules, also documented since at least
4205 1.62, also never worked. As it isn't clear exactly what this means, and
4206 clearly it isn't a major issue, I have abolished the feature by treating it
4207 as "timeout", and writing a warning to the main and panic logs.
4209 36. The display of retry rules for -brt wasn't always showing the error code
4212 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4213 timeout_connect_A, timeout_connect_MX.
4215 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4216 to the empty sender.
4218 39. The daemon was not analysing the content of -oX till after it had closed
4219 stderr and disconnected from the controlling terminal. This meant that any
4220 syntax errors were only noted on the panic log, and the return code from
4221 the command was 0. By re-arranging the code a little, I've made the
4222 decoding happen first, so such errors now appear on stderr, and the return
4223 code is 1. However, the actual setting up of the sockets still happens in
4224 the disconnected process, so errors there are still only recorded on the
4227 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4228 connections (as happens on some IP stacks) was logged at start up time as
4229 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4230 it from "IPv6 and IPv4", which means that two separate sockets are being
4233 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4234 says whether AF_INET or AF_INET6 was passed as an argument.
4236 42. Exiwhat output was messed up when time zones were included in log
4239 43. Exiwhat now gives more information about the daemon's listening ports,
4240 and whether -tls-on-connect was used.
4242 44. The "port" option of the smtp transport is now expanded.
4244 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4245 silently ignored. Now an error message is written to the main and panic
4248 46. There's a new ACL modifier called "logwrite" which writes to a log file
4249 as soon as it is encountered.
4251 47. Added $local_user_uid and $local_user_gid at routing time.
4253 48. Exim crashed when trying to verify a sender address that was being
4256 49. Exim was recognizing only a space character after ".include". It now also
4257 recognizes a tab character.
4259 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4260 extracting the relevant information from the specification. The man page no
4261 longer contains scrambled data for the -d option, and I've added a section
4262 at the front about calling Exim under different names.
4264 51. Added "extra_headers" argument to the "mail" command in filter files.
4266 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4269 53. Installed eximstats 1.29.
4271 54. Added transport_filter_timeout as a generic transport option.
4273 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4274 Cc: header lines. This was required by RFC 822, but it not required by RFC
4277 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4278 incoming messages that did not have them. Now it does so only if the
4279 message originates locally, that is, if there is no associated remote host
4280 address. When Resent- header lines are present, this applies to the Resent-
4281 lines rather than the non-Resent- lines.
4283 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4284 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4286 58. Messages for configuration errors now include the name of the main
4287 configuration file - useful now that there may be more than one file in a
4288 list (.included file names were always shown).
4290 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4291 for those rare installations that do not start the daemon as root or run it
4292 setuid root. I've cut out the call to initgroups() if the daemon is not
4295 60. The Exim user and group can now be bound into the binary as text strings
4296 that are looked up at the start of Exim's processing.
4298 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4300 62. Added $mailstore_basename variable.
4302 63. Installed patch to sieve.c from Michael Haardt.
4304 64. When Exim failed to open the panic log after failing to open the main log,
4305 the original message it was trying to log was written to stderr and debug
4306 output, but if they were not available (the usual case in production), it
4307 was lost. Now it is written to syslog before the two lines that record the
4308 failures to open the logs.
4310 65. Users' Exim filters run in subprocesses under the user's uid. It is
4311 possible for a "deliver" command or an alias in a "personal" command to
4312 provoke an address rewrite. If logging of address rewriting is configured,
4313 this fails because the process is not running as root or exim. There may be
4314 a better way of dealing with this, but for the moment (because 4.30 needs
4315 to be released), I have disabled address rewrite logging when running a
4316 filter in a non-root, non-exim process.
4322 1. The buildconfig auxiliary program wasn't quoting the value set for
4323 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4324 not defined. This bug was masked in 4.22 by the effect that was fixed in
4327 2. Some messages that were rejected after a message id was allocated were
4328 shown as "incomplete" by exigrep. It no longer does this for messages that
4329 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4331 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4332 have allow_domain_literals set, the ID did not get logged in the <= line.
4333 Domain literals are now always recognized in Message-ID: header lines.
4335 4. The first argument for a ${extract expansion item is the key name or field
4336 number. Leading and trailing spaces in this item were not being ignored,
4337 causing some misleading effects.
4339 5. When deliver_drop_privilege was set, single queue runner processes started
4340 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4341 same command in the process it spins off) were not dropping privilege.
4343 6. When the daemon running as "exim" started a queue runner, it always
4344 re-executed Exim in the spun-off process. This is a waste of effort when
4345 deliver_drop_privilege is set. The new process now just calls the
4346 queue-runner function directly.
4352 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4355 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4356 ignored. Though the use of -oP was forcing the writing of a pid file, it
4357 was always written to the default place.
4359 3. If the message "no IP address found for host xxxx" is generated during
4360 incoming verification, it is now followed by identification of the incoming
4361 connection (so you can more easily find what provoked it).
4363 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4365 5. Added some features to "harden" Exim a bit more against certain attacks:
4367 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4368 be put in Local/Makefile. This is like the never_users runtime option,
4369 but it cannot be overridden. The default setting is "root".
4371 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4372 prefix string with which any file named in a -C command line option
4375 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4376 is retained for -C and -D only if the caller of Exim is root. Without
4377 it, the exim user may also use -C and -D and retain privilege.
4379 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4380 command line option is disabled.
4382 6. Macro names set by the -D option must start with an upper case letter, just
4383 like macro names defined in the configuration file.
4385 7. Added "dereference=" facility to LDAP.
4387 8. Two instances of the typo "uknown" in the source files are fixed.
4389 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4390 the Configure-Makefile script screwed up while processing it.
4392 10. Incorporated PCRE 4.4.
4394 11. The SMTP synchronization check was not operating right at the start of an
4395 SMTP session. For example, it could not catch a HELO sent before the client
4396 waited for the greeting. There is now a check for outstanding input at the
4397 point when the greeting is written. Because of the duplex, asynchronous
4398 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4399 way, but not yet received, when the check is performed.
4401 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4402 on TCP/IP sockets, because this apparently causes some broken clients to
4405 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4406 unchanged) from the Cygwin maintainer.
4408 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4409 was supported instead of testing for mbx. Effectively a typo.
4411 15. The spa authenticator server code was not checking that the input it
4412 received was valid base64.
4414 16. The debug output line for the "set" modifier in ACLs was not showing the
4415 name of the variable that was being set.
4417 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4419 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4420 lookup on its own. Something else was needed to trigger the lookup. For
4421 example, a match in host_lookup or the need for a host name in a host list.
4422 Now, if $sender_host_name is referenced and the host name has not yet been
4423 looked up, a lookup is performed. If the lookup fails, the variable remains
4424 empty, and $host_lookup_failed is set to "1".
4426 19. Added "eqi" as a case-independent comparison operator.
4428 20. The saslauthd authentication condition could segfault if neither service
4429 nor realm was specified.
4431 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4432 error message that was logged was misleading, and incoming SMTP
4433 connections were dropped. The message is now more accurate, and temporary
4434 errors are given to SMTP connections.
4436 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4437 (except RSET) with a 421 error, until QUIT is received. However, it was
4438 failing to send a response to QUIT.
4440 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4441 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4442 ACL is now run after the helo_try_verify_hosts code.
4444 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4445 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4446 case-independent, so other case variants are also recognized). Apparently
4447 some systems use these upper case variants.
4449 25. If more than two messages were waiting for the same host, and a transport
4450 filter was specified for the transport, Exim sent two messages over the
4451 same TCP/IP connection, and then failed with "socket operation on non-
4452 socket" when it tried to send the third.
4454 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4456 27. The extern definition of crypt16() in expand.c was not being excluded when
4457 the OS had its own crypt16() function.
4459 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4460 as a preferred synonym for return_size_limit, both as an option and as an
4463 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4465 30. Changed the default configuration ACL to relax the local part checking rule
4466 for addresses that are not in any local domains. For these addresses,
4467 slashes and pipe symbols are allowed within local parts, but the sequence
4468 /../ is explicitly forbidden.
4470 31. SPA server authentication was not clearing the challenge buffer before
4473 32. log_message in a "warn" ACL statement was writing to the reject log as
4474 well as to the main log, which contradicts the documentation and doesn't
4475 seem right (because no rejection is happening). So I have stopped it.
4477 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4478 However, I am unable to do any testing of this.
4480 34. Fixed an infelicity in the appendfile transport. When checking directories
4481 for a mailbox, to see if any needed to be created, it was accidentally
4482 using path names with one or more superfluous leading slashes; tracing
4483 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4485 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4486 contents are added to the log line that is written for every discarded
4487 recipient. (Previously a log_message setting was ignored.)
4489 36. The ${quote: operator now quotes the string if it is empty.
4491 37. The install script runs exim in order to find its version number. If for
4492 some reason other than non-existence or emptiness, which it checks, it
4493 could not run './exim', it was installing it with an empty version number,
4494 i.e. as "exim-". This error state is now caught, and the installation is
4497 38. An argument was missing from the function that creates an error message
4498 when Exim fails to connect to the socket for saslauthd authentication.
4499 This could cause Exim to crash, or give a corrupted message.
4501 39. Added isip, isip4, and isip6 to ${if conditions.
4503 40. The ACL variables $acl_xx are now saved with the message, and can be
4504 accessed later in routers, transports, and filters.
4506 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4507 strings in the file are not string-expanded.
4509 42. If a MAIL command specified a SIZE value that was too large to fit into an
4510 int variable, the check against message_size_limit failed. Such values are
4511 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4512 day this will have to be increased, but I don't think I want to be around
4513 when emails are that large.
4520 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4521 iconv() is not standard in FreeBSD.
4523 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4524 IPv6 enabled. The observed symptom was a segmentation fault on return from
4525 the function os_common_find_running_interfaces() in src/os.c.
4527 3. In the check_special_case() function in daemon.c I had used "errno" as an
4528 argument name, which causes warnings on some systems. This was basically a
4529 typo, since it was named "eno" in the comments!
4531 4. The code that waits for the clock to tick (at a resolution of some fraction
4532 of a second) so as to ensure message-id uniqueness was always waiting for
4533 at least one whole tick, when it could have waited for less. [This is
4534 almost certainly not relevant at current processor speeds, where it is
4535 unlikely to ever wait at all. But we try to future-proof.]
4537 5. The function that sleeps for a time interval that includes fractions of a
4538 second contained a race. It did not block SIGALRM between setting the
4539 timer, and suspending (a couple of lines later). If the interval was short
4540 and the sigsuspend() was delayed until after it had expired, the suspension
4541 never ended. On busy systems this could lead to processes getting stuck for
4544 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4545 process, before it forks any delivery processes. The open lookup caching
4546 mechanism meant that the open file or database connection was passed into
4547 the delivery process. The problem was that delivery processes always tidy
4548 up cached lookup data. This could cause a problem for the next delivery
4549 process started by the queue runner, because the external queue runner
4550 process does not know about the closure. So the next delivery process
4551 still has data in the lookup cache. In the case of a file lookup, there was
4552 no problem because closing a file descriptor in a subprocess doesn't affect
4553 the parent. However, if the lookup was caching a connection to a database,
4554 the connection was closed, and the second delivery process was likely to
4555 see errors such as "PGSQL: query failed: server closed the connection
4556 unexpectedly". The problem has been fixed by closing all cached lookups
4557 in a queue runner before running a delivery process.
4559 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4560 seem to have the "const" qualifier which it has on other OS. I've
4563 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4564 *of the same type* (client or server) with the same public name that an
4565 error should be diagnosed.
4567 9. When Exim looked up a host name for an IP address, but failed to find the
4568 original IP address when looking up the host name (a safety check), it
4569 output the message "<ip address> does not match any IP for NULL", which was
4570 confusing, to say the least. The bug was that the host name should have
4571 appeared instead of "NULL".
4573 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4574 user that is built into the binary, and the -C or -D options is used, root
4575 privilege is dropped before the configuration file is read. In addition,
4576 logging is switched to stderr instead of the normal log files. If the
4577 configuration then re-defines the Exim user, the unprivileged environment
4578 is probably not what is expected, so Exim logs a panic warning message (but
4581 However, if deliver_drop_privilege is set, the unprivileged state may well
4582 be exactly what is intended, so the warning has been cut out in that case,
4583 and Exim is allowed to try to write to its normal log files.
4589 1. smtp_return_error_details was not giving details for temporary sender
4590 or receiver verification errors.
4592 2. Diagnose a configuration error if two authenticators have the same public
4595 3. Exim used not to create the message log file for a message until the first
4596 delivery attempt. This could be confusing when incoming messages were held
4597 for policy or load reasons. The message log file is now created at the time
4598 the message is received, and an initial "Received" line is written to it.
4600 4. The automatically generated man page for command line options had a minor
4601 bug that caused no ill effects; however, a more serious problem was that
4602 the procedure for building the man page automatically didn't always
4603 operate. Consequently, release 4.20 contains an out-of-date version. This
4604 shouldn't happen again.
4606 5. When building Exim with embedded Perl support, the script that builds the
4607 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4608 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4610 6. The freeze_tell option was not being used for messages that were frozen on
4611 arrival, either by an ACL or by local_scan().
4613 7. Added the smtp_incomplete_transaction log selector.
4615 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4616 was accepting AUTH without a new EHLO.
4618 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4619 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4622 10. Logging of TCP/IP connections (when configured) now happens in the main
4623 daemon process instead of the child process, so that the TCP/IP connection
4624 count is more accurate (but it can never be perfect).
4626 11. The use of "drop" in a nested ACL was not being handled correctly in the
4627 outer ACL. Now, if condition failure induced by the nested "drop" causes
4628 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4629 or "require"), the connection is dropped.
4631 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4632 that yield "discard" can now be used with an "accept" or a "discard" verb,
4633 but an error is generated for any others (because I can't see a useful way
4634 to define what should happen).
4636 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4637 are now processed in the same way as lines in the Exim configuration file.
4638 In particular, continuation lines are supported.
4640 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4642 15. Added -ti meaning -t -i.
4644 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4645 domains, and warn by logging if others are found.
4647 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4648 returned from GIFCONF when Exim is trying to find the list of interfaces on
4649 a host. The code in os.c has been modified to copy each ifreq to an aligned
4650 structure in all cases.
4652 Also, in some cases, the returned ifreq's were being copied to a 'struct
4653 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4654 means the last couple of bytes of an IPv6 address could be chopped if the
4655 ifreq contained only a normal sockaddr (14 bytes storage).
4657 18. Named domain lists were not supported in the hosts_treat_as_local option.
4658 An entry such as +xxxx was not recognized, and was treated as a literal
4661 19. Ensure that header lines added by a DATA ACL are included in the reject log
4662 if the ACL subsequently rejects the message.
4664 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4665 MD5 (which is deprecated).
4667 21. When testing a filter file using -bf, Exim was writing a message when it
4668 took the sender from a "From " line in the message, but it was not doing so
4669 when it took $return_path from a Return-Path: header line. It now does.
4671 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4672 with a valid header line field name (a series of printing characters
4673 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4675 23. Changed "disc" in the source to "disk" to conform to the documentation and
4676 the book and for uniformity.
4678 24. Ignore Sendmail's -Ooption=value command line item.
4680 25. When execve() failed while trying to run a command in a pipe transport,
4681 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4682 could be confused with a return value of 69 from the command itself. This
4683 has been changed to 127, the value the shell returns if it is asked to run
4684 a non-existent command. The wording for the related log line suggests a
4685 non-existent command as the problem.
4687 26. If received_header_text expands to an empty string, do not add a Received:
4688 header line to the message. (Well, it adds a token one on the spool, but
4689 marks it "old" so that it doesn't get used or transmitted.)
4691 27. Installed eximstats 1.28 (addition of -nt option).
4693 28. There was no check for failure on the call to getsockname() in the daemon
4694 code. This can fail if there is a shortage of resources on the system, with
4695 ENOMEM, for example. A temporary error is now given on failure.
4697 29. Contrary to the C standard, it seems that in some environments, the
4698 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4699 program. Exim now does this explicitly; it affects the formatting of
4700 timestamps using strftime().
4702 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4703 complaints. I've now initialized all the variables, to avoid this.
4705 32. Header lines added by a system filter were not being "seen" during
4706 transport-time rewrites.
4708 33. The info_callback() function passed to OpenSSL is set up with type void
4709 (*)(SSL *, int, int), as described somewhere. However, when calling the
4710 function (actually a macro) that sets it up, the type void(*)() is
4711 expected. I've put in a cast to prevent warnings from picky compilers.
4713 34. If a DNS black list lookup found a CNAME record, but there were no A
4714 records associated with the domain it pointed at, Exim crashed.
4716 35. If a DNS black list lookup returned more than one A record, Exim ignored
4717 all but the first. It now scans all returned addresses if a particular IP
4718 value is being sought. In this situation, the contents of the
4719 $dnslist_value variable are a list of all the addresses, separated by a
4722 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4723 to accept a host name and all its aliases if the forward lookup for any of
4724 them yielded the IP address of the incoming connection. Now it accepts only
4725 those names whose forward lookup yields the correct IP address. Any other
4726 names are discarded. This closes a loophole whereby a rogue DNS
4727 administrator could create reverse DNS records to break through a
4728 wildcarded host restriction in an ACL.
4730 37. If a user filter or a system filter that ran in a subprocess used any of
4731 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4732 the wrong values were passed to the pipe command ($thisaddress had the
4733 value of $0, $0 had the value of $1, etc). This bug was introduced by
4734 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4736 38. Improved the line breaking for long SMTP error messages from ACLs.
4737 Previously, if there was no break point between 40 and 75 characters, Exim
4738 left the rest of the message alone. Two changes have been made: (a) I've
4739 reduced the minimum length to 35 characters; (b) if it can't find a break
4740 point between 35 and 75 characters, it looks ahead and uses the first one
4741 that it finds. This may give the occasional overlong line, but at least the
4742 remaining text gets split now.
4744 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4745 file descriptors might be low, and that setting 1000 would always raise it.
4746 It turns out that in some environments, the limit is already over 1000 and
4747 that lowering it causes trouble. So now Exim takes care not to decrease it.
4749 40. When delivering a message, the value of $return_path is set to $sender_
4750 address at the start of routing (routers may change the value). By an
4751 oversight, this default was not being set up when an address was tested by
4752 -bt or -bv, which affected the outcome if any router or filter referred to
4755 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4756 log only when "message" or "log_message" are set. However, if one of the
4757 conditions was an address verification, or a call to a nested ACL, the
4758 messages generated by the underlying test were being passed through. This
4759 no longer happens. The underlying message is available in $acl_verify_
4760 message for both "message" and "log_message" expansions, so it can be
4761 passed through if needed.
4763 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4764 new expansion $bh_ to give the encoded byte string without charset
4765 translation. Translation happens only if iconv() is available; HAVE_ICONV
4766 indicates this at build time. HEADERS_CHARSET gives the charset to
4767 translate to; headers_charset can change it in the configuration, and
4768 "headers charset" can change it in an individual filter file.
4770 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4771 that creates RFC 2047 encoded "words" labels them as that charset instead
4772 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4773 expansion operator; (ii) when Exim creates a From: line for a local
4774 message; (iii) when a header line is rewritten to include a "phrase" part.
4776 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4777 buggy, causing it to skip the first lines of messages whose message ID
4778 ended in 'D'. This would not have bitten before Exim release 4.14, because
4779 message IDs were unlikely to end in 'D' before then. The effect was to have
4780 incorrect size information for certain domains.
4782 45. #include "config.h" was missing at the start of the crypt16.c module. This
4783 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4786 46. If there was a timeout during a "random" callout check, Exim treated it as
4787 a failure of the random address, and carried on sending RSET and the real
4788 address. If the delay was just some slowness somewhere, the response to the
4789 original RCPT would be taken as a response to RSET and so on, causing
4790 mayhem of various kinds.
4792 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4793 when I implemented it. It didn't allow for the fact that some option values
4794 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4795 the right test for positive values.
4797 48. Domain names in DNS records are case-independent. Exim always looks them up
4798 in lower case. Some resolvers return domain names in exactly the case they
4799 appear in the zone file, that is, they may contain uppercase letters. Not
4800 all resolvers do this - some return always lower case. Exim was treating a
4801 change of case by a resolver as a change of domain, similar to a widening
4802 of a domain abbreviation. This triggered its re-routing code and so it was
4803 trying to route what was effectively the same domain again. This normally
4804 caused routing to fail (because the router wouldn't handle the domain
4805 twice). Now Exim checks for this case specially, and just changes the
4806 casing of the domain that it ultimately uses when it transmits the message
4809 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4812 50. If a filter generated a file delivery with a non-absolute name (possible if
4813 no home directory exists for the router), the forbid_file option was not
4816 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4817 the existing equality matching.
4819 52. Exim was using ints instead of ino_t variables in some places where it was
4820 dealing with inode numbers.
4822 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4823 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4824 called TMPDIR, and if it finds it is different, it changes its value.
4826 54. The smtp_printf() function is now made available to local_scan() so
4827 additional output lines can be written before returning. There is also an
4828 smtp_fflush() function to enable the detection of a dropped connection.
4829 The variables smtp_input and smtp_batched_input are exported to
4832 55. Changed the default runtime configuration: the message "Unknown user"
4833 has been removed from the ACL, and instead placed on the localuser router,
4834 using the cannot_route_message feature. This means that any verification
4835 failures that generate their own messages won't get overridden. Similarly,
4836 the "Unrouteable address" message that was in the ACL for unverifiable
4837 relay addresses has also been removed.
4839 56. Added hosts_avoid_esmtp to the smtp transport.
4841 57. The exicyclog script was not checking for the esoteric option
4842 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4843 will work only if exicyclog is run under the appropriate euid.
4845 58. Following a discussion on the list, the rules by which Exim recognises line
4846 endings on incoming messages have been changed. The -dropcr and drop_cr
4847 options are now no-ops, retained only for backwards compatibility. The
4848 following line terminators are recognized: LF CRLF CR. However, special
4849 processing applies to CR:
4851 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4852 nor a local message in the state where . is a terminator.
4854 (ii) If a bare CR is encountered in a header line, an extra space is added
4855 after the line terminator so as not to end the header. The reasoning
4856 behind this is that bare CRs in header lines are most likely either
4857 to be mistakes, or people trying to play silly games.
4859 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4860 was being incorrectly given as 18 bytes larger than it should have been.
4861 This is a VOB (very old bug).
4863 60. This may never have affected anything current, but just in case it has:
4864 When the local host is found other than at the start of a list of hosts,
4865 the local host, those with the same MX, and any that follow, are discarded.
4866 When the list in question was part of a longer list of hosts, the following
4867 hosts (not currently being processed) were also being discarded. This no
4868 longer happens. I'm not sure if this situation could ever has previously
4871 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4874 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4875 header. This is something that is recommended in a new Internet Draft, and
4876 is something that is documented as being done by Sendmail. There are two
4877 possible values. For messages generated by the autoreply transport, Exim
4880 Auto-Submitted: auto-replied
4882 whereas for all other generated messages (e.g. bounces) it adds
4884 Auto-Submitted: auto-generated
4886 63. The "personal" condition in filters now includes a test for the
4887 Auto-Submitted: header. If it contains the string "auto-" the message it
4888 not considered personal.
4890 64. Added rcpt_include_affixes as a generic transport option.
4892 65. Added queue_only_override (default true).
4894 66. Added the syslog_duplication option.
4896 67. If what should have been the first header line of a message consisted of
4897 a space followed by a colon, Exim was mis-interpreting it as a header line.
4898 It isn't of course - it is syntactically invalid and should therefore be
4899 treated as the start of the message body. The misbehaviour could have
4900 caused a number of strange effects, including loss of data in subsequent
4901 header lines, and spool format errors.
4903 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4904 client host had authenticated. This control can now be exercised by an ACL
4905 for more flexibility.
4907 69. By default, callouts do not happen when testing with -bh. There is now a
4908 variant, -bhc, which does actually run the callout code, including
4909 consulting and updating the callout cache.
4911 70. Added support for saslauthd authentication, courtesy of Alexander
4914 71. If statvfs() failed on the spool or log directories while checking their
4915 size for availability, Exim confusingly gave the error "space shortage".
4916 Furthermore, in debugging mode it crashed with a floating point exception.
4917 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4918 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4919 serious problem, Exim now writes to the main and panic logs when this
4920 happens, with details of the failure. It then refuses to accept the
4921 incoming message, giving the message "spool directory problem" or "log
4922 directory problem" with a 421 code for SMTP messages.
4924 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4925 0, 1, and 2 exist, because some OS complain for execs without them (see
4926 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4927 descriptors. However, the code omitted to check that the open succeeded,
4928 causing mysterious errors if for some reason the permissions on /dev/null
4929 got screwed. Now Exim writes a message to the main and panic logs, and
4930 bombs out if it can't open /dev/null.
4932 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4933 interact so that it is all more flexible. It is supposed to remain
4934 backwards compatible. Also added extra_local_interfaces.
4936 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4937 to bomb out with an assertion failure - to the client this appears as a
4938 connection drop. This problem occurs in the part of the code that was taken
4939 from the Samba project. Fortunately, the assertion is in a very simple
4940 function, so I have fixed this by reproducing the function inline in the
4941 one place where it is called, and arranging for authentication to fail
4942 instead of killing the process with assert().
4944 75. The SPA client code was not working when the server requested OEM rather
4945 than Unicode encoding.
4947 76. Added code to make require_files with a specific uid setting more usable in
4948 the case where statting the file as root fails - usually a non-root-mounted
4949 NFS file system. When this happens and the failure is EACCES, Exim now
4950 forks a subprocess and does the per-uid checking as the relevant uid.
4952 77. Added process_log_path.
4954 78. If log_file_path was not explicitly set, a setting of check_log_space or
4955 check_log_inodes was ignored.
4957 79. If a space check for the spool or log partitions fails, the incident is now
4958 logged. Of course, in the latter case the data may get lost...
4960 80. Added the %p formatting code to string_format() so that it can be used to
4961 print addresses in debug_print(). Adjusted all the address printing in the
4962 debugging in store.c to use %p rather than %d.
4964 81. There was a concern that a line of code in smtp_in.c could overflow a
4965 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4966 initially expressed, the concern was not well-founded, because trailing
4967 spaces are removed early. However, if the trailing spaces were followed by
4968 a NULL, they did not get removed, so the overflow was possible. Two fixes
4971 (a) I re-wrote the offending code in a cleaner fashion.
4972 (b) If an incoming SMTP command contains a NULL character, it is rejected
4975 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4976 runs initgroups(), so that if the Exim user is in any additional groups,
4977 they will be used during message reception.
4983 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob