1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.576 2009/10/19 12:42:13 nm4 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TK/01 Added patch by Johannes Berg that expands the main option
10 "spamd_servers" if it starts with a dollar sign.
12 TK/02 Write list of recipients to X-Envelope-Sender header when building
13 the mbox-format spool file for content scanning (suggested by Jakob
16 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
17 (http://www.dcc-servers.net/) support via dccifd. Activated by
18 setting EXPERIMENTAL_DCC=yes in Local/Makefile. Check out
19 experimental_spec.txt for more documentation.
21 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
22 by Mark Daniel Reidel <mr@df.eu>.
24 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
25 When building exim an external PCRE library is now needed -
26 PCRE is a system library on the majority of modern systems.
27 See entry on PCRE_LIBS in EDITME file.
29 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
30 conversation. Added nologin parameter to request.
31 Patch contributed by Kirill Miazine
33 TF/01 Do not log submission mode rewrites if they do not change the address.
35 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
37 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
38 log files in place. Contributed by Roberto Lima
40 NM/04 Bugzilla 667: close socket used by dovecot authenticator
42 TF/03 Bugzilla 615: When checking the local_parts router precondition
43 after a local_part_suffix or local_part_prefix option, Exim now
44 does not use the address's named list lookup cache, since this
45 contains cached lookups for the whole local part.
47 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
48 Robert Millan. Documentation is in experimental-spec.txt
50 TF/04 Bugzilla 668: Fix parallel build (make -j).
52 NM/05 Bugzilla 437: Prevent Maildix aux files being created with mode 000
54 NM/05 Bugzilla 598: Improvement to Dovecot authenticator handling.
55 Patch provided by Jan Srzednicki
57 TF/05 Leading white space used to be stripped from $spam_report which
58 wrecked the formatting. Now it is preserved.
60 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
61 that they are available at delivery time.
63 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
65 TF/08 TLS error reporting now respects the incoming_interface and
66 incoming_port log selectors.
68 TF/09 Produce a more useful error message if an SMTP transport's hosts
69 setting expands to an empty string.
71 NM/06 Bugzilla 744: EXPN did not work under TLS.
72 Patch provided by Phil Pennock
74 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
75 Patch provided by Richard Godbee
77 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
78 acl_smtp_notquit, added index entry.
80 NM/09 Bugzilla 787: Potential buffer overflow in string_format
81 Patch provided by Eugene Bujak
83 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to accept()
84 Patch provided by Maxim Dounin
86 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
87 Patch provided by Phil Pennock
89 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
91 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
92 Patch provided by Brad "anomie" Jorsch
94 NM/14 Bugzilla 622: Added timeout setting to transport filter.
95 Patch provided by Dean Brooks
97 TK/05 Add native DKIM support (does not depend on external libraries).
99 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
100 Patch provided by Graeme Fowler
102 NM/16 Bugzilla 851: Documentation example syntax fix.
104 NM/17 Changed NOTICE file to remove references to embedded PCRE.
106 NM/18 Bugzilla 894: Fix issue with very long lines including comments in lsearch
108 NM/19 Bugzilla 745: TLS version reporting
109 Patch provided by Phil Pennock
111 NM/20 Bugzilla 167: bool: condition support
112 Patch provided by Phil Pennock
114 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken clients
115 Patch provided by Phil Pennock
117 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date
118 Patch provided by Brad "anomie" Jorsch
120 NM/23 Bugzilla 687: Fix misparses in eximstats
121 Patch provided by Heiko Schlittermann
123 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid
124 Patch provided by Heiko Schlittermann
126 NM/24 Bugzilla 727: Use transport mode as default mode for maildirsize file
132 TK/01 Add preliminary DKIM support. Currently requires a forked version of
133 ALT-N's libdkim that I have put here:
134 http://duncanthrax.net/exim-experimental/
136 Note to Michael Haardt: I had to rename some vars in sieve.c. They
137 were called 'true' and it seems that C99 defines that as a reserved
138 keyword to be used with 'bool' variable types. That means you could
139 not include C99-style headers which use bools without triggering
140 build errors in sieve.c.
142 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
143 as mailq or other aliases. Changed the --help handling significantly
144 to do whats expected. exim_usage() emits usage/help information.
146 SC/01 Added the -bylocaldomain option to eximstats.
148 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
150 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
152 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
158 PH/01 Another patch from the Sieve maintainer.
160 PH/02 When an IPv6 address is converted to a string for single-key lookup
161 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
162 dots are used instead of colons so that keys in lsearch files need not
163 contain colons. This was done some time before quoting was made available
164 in lsearch files. However, iplsearch files do require colons in IPv6 keys
165 (notated using the quote facility) so as to distinguish them from IPv4
166 keys. This meant that lookups for IP addresses in host lists did not work
167 for iplsearch lookups.
169 This has been fixed by arranging for IPv6 addresses to be expressed with
170 colons if the lookup type is iplsearch. This is not incompatible, because
171 previously such lookups could never work.
173 The situation is now rather anomolous, since one *can* have colons in
174 ordinary lsearch keys. However, making the change in all cases is
175 incompatible and would probably break a number of configurations.
177 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
180 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
181 conversion specification without a maximum field width, thereby enabling
182 a rogue spamd server to cause a buffer overflow. While nobody in their
183 right mind would setup Exim to query an untrusted spamd server, an
184 attacker that gains access to a server running spamd could potentially
185 exploit this vulnerability to run arbitrary code as the Exim user.
187 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
188 $primary_hostname instead of what libspf2 thinks the hosts name is.
190 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
191 a directory entry by the name of the lookup key. Previously, if a
192 symlink pointed to a non-existing file or a file in a directory that
193 Exim lacked permissions to read, a lookup for a key matching that
194 symlink would fail. Now it is enough that a matching directory entry
195 exists, symlink or not. (Bugzilla 503.)
197 PH/03 The body_linecount and body_zerocount variables are now exported in the
200 PH/04 Added the $dnslist_matched variable.
202 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
203 This means they are set thereafter only if the connection becomes
206 PH/06 Added the client_condition to authenticators so that some can be skipped
207 by clients under certain conditions.
209 PH/07 The error message for a badly-placed control=no_multiline_responses left
210 "_responses" off the end of the name.
212 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
214 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
215 (without spaces) instead of just copying the configuration text.
217 PH/10 Added the /noupdate option to the ratelimit ACL condition.
219 PH/11 Added $max_received_linelength.
221 PH/12 Added +ignore_defer and +include_defer to host lists.
223 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
224 way in which PCRE > 7.0 is built.
226 PH/14 Implemented queue_only_load_latch.
228 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
229 MAIL command. The effect was to mangle the value on 64-bit systems.
231 PH/16 Another patch from the Sieve maintainer.
233 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
235 PH/18 If a system quota error occurred while trying to create the file for
236 a maildir delivery, the message "Mailbox is full" was not appended to the
237 bounce if the delivery eventually timed out. Change 4.67/27 below applied
238 only to a quota excession during the actual writing of the file.
240 PH/19 It seems that peer DN values may contain newlines (and other non-printing
241 characters?) which causes problems in log lines. The DN values are now
242 passed through string_printing() before being added to log lines.
244 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
245 and InterBase are left for another time.)
247 PH/21 Added message_body_newlines option.
249 PH/22 Guard against possible overflow in moan_check_errorcopy().
251 PH/23 POSIX allows open() to be a macro; guard against that.
253 PH/24 If the recipient of an error message contained an @ in the local part
254 (suitably quoted, of course), incorrect values were put in $domain and
255 $local_part during the evaluation of errors_copy.
261 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
262 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
265 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
266 issue a MAIL command.
268 PH/02 In an ACL statement such as
270 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
272 if a client was not listed at all, or was listed with a value other than
273 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
274 the condition was not true (as it should be), so access was not denied.
275 The bug was that the ! inversion was incorrectly passed on to the second
276 item. This has been fixed.
278 PH/03 Added additional dnslists conditions == and =& which are different from
279 = and & when the dns lookup returns more than one IP address.
281 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
282 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
284 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
285 FSYNC, which compiles an option called disable_fsync that allows for
286 bypassing fsync(). The documentation is heavily laced with warnings.
288 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
290 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
291 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
292 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
293 including adding "make clean"; (3) Added -fPIC when compiling the test
294 dynamically loaded module, to get rid of a warning.
296 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
297 message fails, move_frozen_messages = true and ignore_bounce_errors_after
298 = 0s. The bug is otherwise harmless.
300 PH/07 There was a bug in the dovecot authenticator such that the value of
301 $auth1 could be overwritten, and so not correctly preserved, after a
302 successful authentication. This usually meant that the value preserved by
303 the server_setid option was incorrect.
305 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
307 PH/09 Installed PCRE release 7.0.
309 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
310 run for batched SMTP input. It is now run at the start of every message
311 in the batch. While fixing this I discovered that the process information
312 (output by running exiwhat) was not always getting set for -bs and -bS
313 input. This is fixed, and it now also says "batched" for BSMTP.
315 PH/11 Added control=no_pipelining.
317 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
318 patch, slightly modified), and move the expansion of helo_data till after
319 the connection is made in the smtp transport (so it can use these
322 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
324 PH/14 Added log_selector = +pid.
326 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
328 PH/16 Add ${if forany and ${if forall.
330 PH/17 Added dsn_from option to vary the From: line in DSNs.
332 PH/18 Flush SMTP output before performing a callout, unless control =
333 no_callout_flush is set.
335 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
336 was true (the default) a successful delivery failed to delete the retry
337 item, thus causing premature timeout of the address. The bug is now
340 PH/20 Added hosts_avoid_pipelining to the smtp transport.
342 PH/21 Long custom messages for fakedefer and fakereject are now split up
343 into multiline reponses in the same way that messages for "deny" and
344 other ACL rejections are.
346 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
347 with slight modification.
349 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
350 draft, changing the syntax and factoring some duplicate code".
352 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
353 for deliveries of the second and subsequent messages over the same SMTP
356 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
357 ${reduce, with only minor "tidies".
359 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
361 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
362 expansion side effects.
364 PH/27 When a message times out after an over-quota error from an Exim-imposed
365 quota, the bounce message says "mailbox is full". This message was not
366 being given when it was a system quota that was exceeded. It now should
369 MH/03 Made $recipients available in local_scan(). local_scan() already has
370 better access to the recipient list through recipients_list[], but
371 $recipients can be useful in postmaster-provided expansion strings.
373 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
374 in the case of a MAIL command with additional options following the
375 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
376 were accidentally chopped off.
378 PH/29 SMTP synchronization checks are implemented when a command is read -
379 there is a check that no more input is waiting when there shouldn't be
380 any. However, for some commands, a delay in an ACL can mean that it is
381 some time before the response is written. In this time, more input might
382 arrive, invalidly. So now there are extra checks after an ACL has run for
383 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
384 pipelining has not been advertised.
386 PH/30 MH's patch to allow iscntrl() characters to be list separators.
388 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
389 returned in the SMTP response when smtp_return_error_details was false.
392 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
393 instead of the C I/O that was originally supplied, because problems were
396 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
397 Exim which did not show up earlier: it was assuming that a call to
398 SSL_CTX_set_info_callback() might give an error value. In fact, there is
399 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
400 was a macro that became an assignment, so it seemed to work. This has
401 changed to a proper function call with a void return, hence the compile
402 error. Exim's code has been fixed.
404 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
407 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
409 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
411 PH/37 If a message is not accepted after it has had an id assigned (e.g.
412 because it turns out to be too big or there is a timeout) there is no
413 "Completed" line in the log. When some messages of this type were
414 selected by exigrep, they were listed as "not completed". Others were
415 picked up by some special patterns. I have improved the selection
416 criteria to be more general.
418 PH/38 The host_find_failed option in the manualroute router can now be set
419 to "ignore", to completely ignore a host whose IP address cannot be
420 found. If all hosts are ignored, the behaviour is controlled by the new
421 host_all_ignored option.
423 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
424 homing or because of multiple MX records with /mx) generated more than
425 one IP address, and the following item turned out to be the local host,
426 all the secondary addresses of the first item were incorrectly removed
427 from the list, along with the local host and any following hosts (which
428 is what is supposed to happen).
430 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
431 whoever called Exim into the -H file. In the case of the daemon it was
432 behaving confusingly. When first started, it used values for whoever
433 started the daemon, but after a SIGHUP it used the Exim user (because it
434 calls itself on a restart). I have changed the code so that it now always
437 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
438 message are rejected with the same error (e.g. no authentication or bad
439 sender address), and a DATA command is nevertheless sent (as can happen
440 with PIPELINING or a stupid MUA), the error message that was given to the
441 RCPT commands is included in the rejection of the DATA command. This is
442 intended to be helpful for MUAs that show only the final error to their
445 PH/42 Another patch from the Sieve maintainer.
447 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
448 Eximstats - Fixed some broken HTML links and added missing column headers
450 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
451 columns for Rejects, Temp Rejects, Ham, and Spam rows.
453 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
455 PH/43 Yet another patch from the Sieve maintainer.
457 PH/44 I found a way to check for a TCP/IP connection going away before sending
458 the response to the final '.' that terminates a message, but only in the
459 case where the client has not sent further data following the '.'
460 (unfortunately, this is allowed). However, in many cases there won't be
461 any further data because there won't be any more messages to send. A call
462 to select() can be used: if it shows that the input is "ready", there is
463 either input waiting, or the socket has been closed. An attempt to read
464 the next input character can distinguish the two cases. Previously, Exim
465 would have sent an OK response which the client would never have see.
466 This could lead to message repetition. This fix should cure that, at
467 least in a lot of common cases.
469 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
470 advertised in response to EHLO.
476 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
477 fixed by 4.65/MH/01 (is this a record?) are fixed:
479 (i) An empty string was always treated as zero by the numeric comparison
480 operators. This behaviour has been restored.
482 (ii) It is documented that the numeric comparison operators always treat
483 their arguments as decimal numbers. This was broken in that numbers
484 starting with 0 were being interpreted as octal.
486 While fixing these problems I realized that there was another issue that
487 hadn't been noticed. Values of message_size_limit (both the global option
488 and the transport option) were treated as octal if they started with 0.
489 The documentation was vague. These values are now always treated as
490 decimal, and I will make that clear in the documentation.
496 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
497 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
500 MH/01 Don't check that the operands of numeric comparison operators are
501 integers when their expansion is in "skipping" mode (fixes bug
502 introduced by 4.64-PH/07).
504 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
505 child addresses, Exim now panics and dies. Previously, because the count
506 is held in a short int, deliveries were likely to be lost. As such a
507 large number of recipients for a single message is ridiculous
508 (performance will be very, very poor), I have chosen to impose a limit
509 rather than extend the field.
515 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
516 leftover -K file (the existence of which was triggered by #402).
517 While we were at it, introduced process PID as part of the -K
518 filename. This should rule out race conditions when creating
521 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
522 processing considerably. Previous code took too long for large mails,
523 triggering a timeout which in turn triggers #401.
525 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
526 in the DK code in transports.c. sendfile() is not really portable,
527 hence the _LINUX specificness.
529 TF/01 In the add_headers option to the mail command in an Exim filter,
530 there was a bug that Exim would claim a syntax error in any
531 header after the first one which had an odd number of characters
534 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
535 callout verification, Exim cached a "reject" for the entire domain. This
536 is correct for most verifications, but it is not correct for a recipient
537 verification with use_sender or use_postmaster set, because in that case
538 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
539 case of MAIL FROM:<> rejection from other early rejections (e.g.
540 rejection of HELO). When verifying a recipient using a non-null MAIL
541 address, the cache is ignored if it shows MAIL FROM:<> rejection.
542 Whatever the result of the callout, the value of the domain cache is
543 left unchanged (for any other kind of callout, getting as far as trying
544 RCPT means that the domain itself is ok).
546 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
549 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
550 manifest itself as EPIPE rather than ECONNECT. When tidying away a
551 session, the daemon ignores ECONNECT errors and logs others; it now
552 ignores EPIPE as well.
554 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
555 (quoted-printable decoding).
557 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
558 later the small subsequent patch to fix an introduced bug.
560 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
562 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
564 PH/08 An error is now given if message_size_limit is specified negative.
566 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
567 to be given (somewhat) arbitrary names.
569 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
572 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
573 miscellaneous code fixes
575 PH/10 Added the log_reject_target ACL modifier to specify where to log
578 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
579 hostname. This is wrong, because it relates to the incoming message (and
580 probably the interface on which it is arriving) and not to the outgoing
581 callout (which could be using a different interface). This has been
582 changed to use the value of the helo_data option from the smtp transport
583 instead - this is what is used when a message is actually being sent. If
584 there is no remote transport (possible with a router that sets up host
585 addresses), $smtp_active_hostname is used.
587 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
588 tweaks were necessary in order to get it to work (see also 21 below):
589 (a) The code assumed that strncpy() returns a negative number on buffer
590 overflow, which isn't the case. Replaced with Exim's string_format()
592 (b) There were several signed/unsigned issues. I just did the minimum
593 hacking in of casts. There is scope for a larger refactoring.
594 (c) The code used strcasecmp() which is not a standard C function.
595 Replaced with Exim's strcmpic() function.
596 (d) The code set only $1; it now sets $auth1 as well.
597 (e) A simple test gave the error "authentication client didn't specify
598 service in request". It would seem that Dovecot has changed its
599 interface. Fortunately there's a specification; I followed it and
600 changed what the client sends and it appears to be working now.
602 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
605 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
606 address A is aliased to B and C, where B exists and C does not. Without
607 -v the output is "A verified" because verification stops after a
608 successful redirection if more than one address is generated. However,
609 with -v the child addresses are also verified. Exim was outputting "A
610 failed to verify" and then showing the successful verification for C,
611 with its parentage. It now outputs "B failed to verify", showing B's
612 parentage before showing the successful verification of C.
614 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
615 look up a TXT record in a specific list after matching in a combined
618 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
619 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
620 they consult the DNS. I had assumed they would set it the way they
621 wanted; and indeed my experiments on Linux seem to show that in some
622 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
623 To be on the safe side, however, I have now made the interface to
624 host_find_byname() similar to host_find_bydns(), with an argument
625 containing the DNS resolver options. The host_find_byname() function now
626 sets these options at its start, just as host_find_bydns() does. The smtp
627 transport options dns_qualify_single and dns_search_parents are passed to
628 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
629 of host_find_byname() use the default settings of RES_DEFNAMES
630 (qualify_single) but not RES_DNSRCH (search_parents).
632 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
633 spool_read_header() do less string testing, by means of a preliminary
634 switch on the second character of optional "-foo" lines. (This is
635 overdue, caused by the large number of possibilities that now exist.
636 Originally there were few.) While I was there, I also converted the
637 str(n)cmp tests so they don't re-test the leading "-" and the first
638 character, in the hope this might squeeze out yet more improvement.
640 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
641 flag allowing group syntax was set by the header_syntax check but not
642 turned off, possible causing trouble later; (2) The flag was not being
643 set at all for the header_verify test, causing "group"-style headers to
644 be rejected. I have now set it in this case, and also caused header_
645 verify to ignore an empty address taken from a group. While doing this, I
646 came across some other cases where the code for allowing group syntax
647 while scanning a header line wasn't quite right (mostly, not resetting
648 the flag correctly in the right place). These bugs could have caused
649 trouble for malformed header lines. I hope it is now all correct.
651 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
652 with the "reply" argument non-NULL. The code, however (which originally
653 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
654 but it didn't always do it. This confused somebody who was copying the
655 code for some other use. I have removed all the tests.
657 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
658 feature that was used to support insecure browsers during the U.S. crypto
659 embargo. It requires special client support, and Exim is probably the
660 only MTA that supported it -- and would never use it because real RSA is
661 always available. This code has been removed, because it had the bad
662 effect of slowing Exim down by computing (never used) parameters for the
663 RSA_EXPORT functionality.
665 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
666 authenticator to fail if there's a tab character in the incoming data
667 (there should never be unless someone is messing about, as it's supposed
668 to be base64-encoded). Also added, on Timo's advice, the "secured" option
669 if the connection is using TLS or if the remote IP is the same as the
670 local IP, and the "valid-client-cert option" if a client certificate has
673 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
674 authenticators. This can be used for authorization after authentication
675 succeeds. (In the case of plaintext, it servers for both authentication
678 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
679 if any retry times were supplied.
681 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
682 connection, where there is no client IP address to check. In this
683 situation, the verify now always succeeds.
685 PH/25 Applied John Jetmore's -Mset patch.
687 PH/26 Added -bem to be like -Mset, but loading a message from a file.
689 PH/27 In a string expansion for a processed (not raw) header when multiple
690 headers of the same name were present, leading whitespace was being
691 removed from all of them, but trailing whitespace was being removed only
692 from the last one. Now trailing whitespace is removed from each header
693 before concatenation. Completely empty headers in a concatenation (as
696 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
697 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
699 PH/29 [Removed. This was a change that I later backed out, and forgot to
700 correct the ChangeLog entry (that I had efficiently created) before
701 committing the later change.]
703 PH/30 Exim was sometimes attempting to deliver messages that had suffered
704 address errors (4xx response to RCPT) over the same connection as other
705 messages routed to the same hosts. Such deliveries are always "forced",
706 so retry times are not inspected. This resulted in far too many retries
707 for the affected addresses. The effect occurred only when there were more
708 hosts than the hosts_max_try setting in the smtp transport when it had
709 the 4xx errors. Those hosts that it had tried were not added to the list
710 of hosts for which the message was waiting, so if all were tried, there
711 was no problem. Two fixes have been applied:
713 (i) If there are any address or message errors in an SMTP delivery, none
714 of the hosts (tried or untried) are now added to the list of hosts
715 for which the message is waiting, so the message should not be a
716 candidate for sending over the same connection that was used for a
717 successful delivery of some other message. This seems entirely
718 reasonable: after all the message is NOT "waiting for some host".
719 This is so "obvious" that I'm not sure why it wasn't done
720 previously. Hope I haven't missed anything, but it can't do any
721 harm, as the worst effect is to miss an optimization.
723 (ii) If, despite (i), such a delivery is accidentally attempted, the
724 routing retry time is respected, so at least it doesn't keep
725 hammering the server.
727 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
728 in ${readsocket because some servers need this prod.
730 PH/32 Added some extra debug output when updating a wait-xxx database.
732 PH/33 The hint "could be header name not terminated by colon", which has been
733 given for certain expansion errors for a long time, was not being given
734 for the ${if def:h_colon_omitted{... case.
736 PH/34 The spec says: "With one important exception, whenever a domain list is
737 being scanned, $domain contains the subject domain." There was at least
738 one case where this was not true.
740 PH/35 The error "getsockname() failed: connection reset by peer" was being
741 written to the panic log as well as the main log, but it isn't really
742 panic-worthy as it just means the connection died rather early on. I have
743 removed the panic log writing for the ECONNRESET error when getsockname()
746 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
747 runs only) independently of the message's sender address. This meant
748 that, if the 4xx error was in fact related to the sender, a different
749 message to the same recipient with a different sender could confuse
750 things. In particualar, this can happen when sending to a greylisting
751 server, but other circumstances could also provoke similar problems.
752 I have changed the default so that the retry time for these errors is now
753 based a combination of the sender and recipient addresses. This change
754 can be overridden by setting address_retry_include_sender=false in the
757 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
758 remote server are returned as part of bounce messages. This was not
759 happening for LMTP over a pipe (the lmtp transport), but now it is the
760 same for both kinds of LMTP.
762 PH/38 Despite being documented as not happening, Exim was rewriting addresses
763 in header lines that were in fact CNAMEs. This is no longer the case.
765 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
766 and queue runs started by the daemon processed all messages. This has
767 been fixed so that -R and -S can now usefully be given with -q<time>.
769 PH/40 Import PCRE release 6.7 (fixes some bugs).
771 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
773 PH/42 Give an error if -q is specified more than once.
775 PH/43 Renamed the variables $interface_address and $interface_port as
776 $received_ip_address and $received_port, to make it clear that these
777 values apply to message reception, and not to the outgoing interface when
778 a message is delivered. (The old names remain recognized, of course.)
780 PH/44 There was no timeout on the connect() call when using a Unix domain
781 socket in the ${readsocket expansion. There now is.
783 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
784 be meaningful with "accept".
786 SC/01 Eximstats V1.43
787 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
789 SC/02 Eximstats V1.44
790 Use a glob alias rather than an array ref in the generated
791 parser. This improves both readability and performance.
793 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
794 Collect SpamAssassin and rejection statistics.
795 Don't display local sender or destination tables unless
796 there is data to show.
797 Added average volumes into the top table text output.
799 SC/04 Eximstats V1.46
800 Collect data on the number of addresses (recipients)
801 as well as the number of messages.
803 SC/05 Eximstats V1.47
804 Added 'Message too big' to the list of mail rejection
805 reasons (thanks to Marco Gaiarin).
807 SC/06 Eximstats V1.48
808 Mainlog lines which have GMT offsets and are too short to
809 have a flag are now skipped.
811 SC/07 Eximstats V1.49 (Alain Williams)
812 Added the -emptyok flag.
814 SC/08 Eximstats V1.50
815 Fixes for obtaining the IP address from reject messages.
817 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
818 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
819 whitesspace changes from 4.64-PH/27
821 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
824 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
825 are found, allow negative numbers in numeric criteria)
827 JJ/06 exipick.20061117.2, added new $message_body_missing variable
829 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
830 to match changes made in 4.64-PH/43
832 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
834 PH/47 Put in an explicit test for a DNS lookup of an address record where the
835 "domain" is actually an IP address, and force a failure. This locks out
836 those revolvers/nameservers that support "A-for-A" lookups, in
837 contravention of the specifications.
839 PH/48 When a host name was looked up from an IP address, and the subsequent
840 forward lookup of the name timed out, the host name was left in
841 $sender_host_name, contrary to the specification.
843 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
844 restricted to single-key lookups, Exim was not diagnosing an error if
845 * or *@ was used with a query-style lookup.
847 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
849 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
850 long ago, but noone interested enough thought of it. Let's just say that
851 the "1.1" means that there are some new functions that weren't there at
852 some point in the past.
854 PH/51 Error processing for expansion failure of helo_data from an smtp
855 transport during callout processing was broken.
857 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
858 tested/used via the -bh/-bhc/-bs options.
860 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
861 bug, fixed in subsequent PCRE releases).
863 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
864 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
866 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
872 SC/01 Use a glob alias rather than an array ref in eximstats generated
873 parser. This improves both readability and performance.
875 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
876 Don't display local sender or destination tables in eximstats unless
877 there is data to show.
878 Added average volumes into the eximstats top table text output.
880 SC/03 Collect data on the number of addresses (recipients) as well
881 as the number of messages in eximstats.
883 TF/01 Correct an error in the documentation for the redirect router. Exim
884 does (usually) call initgroups() when daemonizing.
886 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
887 with consistent privilege compared to when running as a daemon.
889 TF/03 Note in the spec that $authenticated_id is not set for local
890 submissions from trusted users.
892 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
893 Thanks to Dean Brooks <dean@iglou.com> for the patch.
895 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
896 by adding some example configuration directives to the default
897 configuration file. A little bit of work is required to uncomment the
898 directives and define how usernames and passwords are checked, but
899 there is now a framework to start from.
901 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
902 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
903 without this. I don't know how relevant this is to other LDAP libraries.
905 PH/02 Add the verb name to the "unknown ACL verb" error.
907 PH/03 Magnus Holmgren's patch for filter_prepend_home.
909 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
911 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
912 directory not expanded when it should be if an expanded home directory
913 was set for the address (which is overridden by the transport).
915 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
918 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
919 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
920 because it is too late at that time, and has no effect.
922 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
923 security issue with \' (bugzilla #107). I could not use the
924 PQescapeStringConn() function, because it needs a PGconn value as one of
927 PH/08 When testing addresses using -bt, indicate those final addresses that
928 are duplicates that would not cause an additional delivery. At least one
929 person was confused, thinking that -bt output corresponded to deliveries.
930 (Suppressing duplicates isn't a good idea as you lose the information
931 about possibly different redirections that led to the duplicates.)
933 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
934 systems where poll() doesn't work, in particular OS X.
936 PH/10 Added more information to debugging output for retry time not reached.
938 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
939 operations in malware.c.
941 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
944 PH/13 If write_rejectlog was set false when logging was sent to syslog with
945 syslog_duplication set false, log lines that would normally be written
946 both the the main log and to the reject log were not written to syslog at
949 PH/14 In the default configuration, change the use of "message" in ACL warn
950 statements to "add_header".
952 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
953 not followed by a command (e.g. "seen endif").
955 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
956 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
959 PH/17 Added extra conditions to the default value of delay_warning_condition
963 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
964 { match{$h_precedence:}{(?i)bulk|list|junk} } \
965 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
968 The Auto-Submitted: and various List- headers are standardised, whereas I
969 don't think Precedence: ever was.
971 PH/18 Refactored debugging code in route_finduser() to show more information,
972 in particular, the error code if getpwnam() issues one.
974 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
975 This is apparently needed in addition to the PH/07 change above to avoid
976 any possible encoding problems.
978 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
979 but not after initializing Perl.
981 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
982 output them only if debugging. By default they are written stderr,
983 apparently, which is not desirable.
985 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
988 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
991 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
993 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
994 authenticated or an ident call has been made. Suppress the default
995 values for $authenticated_id and $authenticated_sender (but permit -oMai
996 and -oMas) when testing with -bh.
998 PH/24 Re-jigged the order of the tests in the default configuration so that the
999 tests for valid domains and recipients precede the DNS black list and CSA
1000 tests, on the grounds that those ones are more expensive.
1002 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1003 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1004 This bug exists in every version of Exim that I still have, right back to
1007 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1008 However, an attempt to turn on case-sensitivity in a regex key by
1009 including (?-i) didn't work because the subject string was already
1010 lowercased, and the effects were non-intuitive. It turns out that a
1011 one-line patch can be used to allow (?-i) to work as expected.
1017 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1018 other effects) broke the use of negated acl sub-conditions.
1020 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1023 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1024 "Deny" causes Exim to reject the incoming connection with a 554 error.
1025 Unfortunately, if there is a major crisis, such as a disk failure,
1026 tcp-wrappers gives "deny", whereas what one would like would be some
1027 kind of temporary error. A kludge has been added to help with this.
1028 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1029 554 error is used if errno is still zero or contains ENOENT (which occurs
1030 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1033 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1035 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1036 errors. Otherwise a message that provokes a temporary error (when other
1037 messages do not) can cause a whole host to time out.
1039 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1040 addresses were routed directly to files or pipes from a redirect router.
1041 File deliveries just didn't batch; pipe deliveries might have suffered
1044 PH/06 A failure to get a lock for a hints database would erroneously always say
1045 "Failed to get write lock", even when it was really a read lock.
1047 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1048 of 0600. This has been changed to use the value of the lockfile_mode
1049 option (which defaults to 0600).
1051 PH/08 Applied small patch from the Sieve maintainer.
1053 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1054 folder from quota calculations, a direct delivery into this folder messed
1055 up the contents of the maildirsize file. This was because the regex was
1056 used only to exclude .Trash (or whatever) when the size of the mailbox
1057 was calculated. There was no check that a delivery was happening into an
1058 excluded directory. This bug has been fixed by ignoring all quota
1059 processing for deliveries into excluded directories.
1061 PH/10 Added the maildirfolder_create_regex option to appendfile.
1067 PH/01 The code for finding all the local interface addresses on a FreeBSD
1068 system running IPv6 was broken. This may well have applied to all BSD
1069 systems, as well as to others that have similar system calls. The broken
1070 code found IPv4 interfaces correctly, but gave incorrect values for the
1071 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1072 that it would not match correctly against @[] and not recognize the IPv6
1075 PH/02 The ipliteral router was not recognizing addresses of the form user@
1076 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1078 PH/03 Added disable_ipv6.
1080 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1081 decoded form, because it is most often used to construct To: headers
1082 lines in autoreplies, and the decoded form may well be syntactically
1083 invalid. However, $reply_address has leading white space removed, and all
1084 newlines turned into spaces so that the autoreply transport does not
1087 PH/05 If group was specified without a user on a router, and no group or user
1088 was specified on a transport, the group from the router was ignored.
1090 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1091 for visible compile-time settings that can be used to change these
1092 numbers, for those that want even more. Backwards compatibility with old
1093 spool files has been maintained. However, going back to a previous Exim
1094 release will lost any variables that are in spool files.
1096 PH/07 Two small changes when running in the test harness: increase delay when
1097 passing a TCP/IP connection to a new process, in case the original
1098 process has to generate a bounce, and remove special handling of
1099 127.0.0.2 (sic), which is no longer necessary.
1101 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1102 be the same on different OS.
1104 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1107 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1108 whether --show-vars was specified or not
1110 JJ/02 exipick: Added support for new ACL variable spool format introduced
1113 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1114 syntactically invalid From: or Reply-to: line, and a filter used this to
1115 generate an autoreply, and therefore failed to obtain an address for the
1116 autoreply, Exim could try to deliver to a non-existent relative file
1117 name, causing unrelated and misleading errors. What now happens is that
1118 it logs this as a hard delivery error, but does not attempt to create a
1121 PH/11 The exinext utility has a -C option for testing purposes, but although
1122 the given file was scanned by exinext itself; it wasn't being passed on
1123 when Exim was called.
1125 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1126 an end-of-file indication when reading a command response.
1128 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1129 compiled. In many other places in Exim, IPv6 addresses are always
1130 recognized, so I have changed this. It also means that IPv4 domain
1131 literals of the form [IPV4:n.n.n.n] are now always recognized.
1133 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1134 used if the router is not running as root, for example, when verifying at
1135 ACL time, or when using -bh. The debugging output from this situation was
1136 non-existent - all you got was a failure to exec. I have made two
1139 (a) Failures to set uid/gid, the current directory, or a process leader
1140 in a subprocess such as that created by queryprogram now generate
1141 suitable debugging ouput when -d is set.
1143 (b) The queryprogram router detects when it is not running as root,
1144 outputs suitable debugging information if -d is set, and then runs
1145 the subprocess without attempting to change uid/gid.
1147 PH/15 Minor change to Makefile for building test_host (undocumented testing
1150 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1151 additional section of a DNS packet that returns MX or SRV records.
1152 Instead, it always explicitly searches for A/AAAA records. This avoids
1153 major problems that occur when a DNS server includes only records of one
1154 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1155 fixed another bug: if SRV records were looked up and the corresponding
1156 address records were *not* found in the additional section, the port
1157 values from the SRV records were lost.
1159 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1160 using the correct key (the original address) when searching the retry
1161 rules in order to find which one to use for generating the retry hint.
1163 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1164 adding the default one. Similarly, if it contains a Reply-To: header, the
1165 errors_reply_to option, if set, is not used.
1167 PH/19 When calculating a retry time, Exim used to measure the "time since
1168 failure" by looking at the "first failed" field in the retry record. Now
1169 it does not use this if it is later than than the arrival time of the
1170 message. Instead it uses the arrival time. This makes for better
1171 behaviour in cases where some deliveries succeed, thus re-setting the
1172 "first failed" field. An example is a quota failure for a huge message
1173 when small messages continue to be delivered. Without this change, the
1174 "time since failure" will always be short, possible causing more frequent
1175 delivery attempts for the huge message than are intended.
1176 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1178 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1179 $1, $2, $3) because the numerical variables can be reset during some
1180 expansion items (e.g. "match"), thereby losing the authentication data.
1182 PH/21 Make -bV show the size of off_t variables so that the test suite can
1183 decide whether to run tests for quotas > 2G.
1185 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1186 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1187 filecount value is greater than 2G or if a quota value is greater than 2G
1188 on a system where the size of off_t is not greater than 4, a panic error
1191 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1192 never match. The debug and -bh output now contains an explicit error
1193 message indicating a malformed IPv4 address or mask.
1195 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1196 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1197 PH/23 above applies.
1199 PH/25 Do not write to syslog when running in the test harness. The only
1200 occasion when this arises is a failure to open the main or panic logs
1201 (for which there is an explicit test).
1203 PH/26 Added the /no_tell option to "control=freeze".
1205 PH/27 If a host name lookup failed very early in a connection, for example, if
1206 the IP address matched host_lookup and the reverse lookup yielded a name
1207 that did not have a forward lookup, an error message of the form "no IP
1208 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1209 could be logged. Now it outputs the IP address instead of "NULL".
1211 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1212 allows the sender and the authenticated sender to be set when
1213 submitting a message from within Exim. Since child_open_exim() is
1214 documented for local_scan(), the new function should be too.
1216 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1217 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1218 results in an empty string is now treated as unset.
1220 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1222 PH/31 Added sender_verify_fail logging option.
1224 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1225 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1226 tidied the source and removed it altogether.
1228 PH/33 When a queue run was abandoned because the load average was too high, a
1229 log line was always written; now it is written only if the queue_run log
1230 selector is set. In addition, the log line for abandonment now contains
1231 information about the queue run such as the pid. This is always present
1232 in "start" and "stop" lines but was omitted from the "abandon" line.
1234 PH/34 Omit spaces between a header name and the colon in the error message that
1235 is given when verify = headers_syntax fails (if there are lots of them,
1236 the message gets confusing).
1238 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1239 names, as there are now some PTR records that contain slashes. This check
1240 is only to protect against broken name servers that fall over on strange
1241 characters, so the fact that it applies to all lookups doesn't matter.
1243 PH/36 Now that the new test suite is complete, we can remove some of the
1244 special code in Exim that was needed for the old test suite. For example,
1245 sorting DNS records because real resolvers return them in an arbitrary
1246 order. The new test suite's fake resolver always returns records in the
1249 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1250 bounces) except when queue_only is set, to avoid logging races between
1251 the different processes.
1253 PH/38 Panic-die if .include specifies a non-absolute path.
1255 PH/39 A tweak to the "H" retry rule from its user.
1257 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1258 a label. They prevented compilation on older perls.
1260 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1261 a warning to be raised on newish perls.
1263 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1264 on queue. Changes to match documented behaviour of showing count of
1265 messages matching specified criteria.
1267 PH/40 Changed the default ident timeout from 30s to 5s.
1269 PH/41 Added support for the use of login_cap features, on those BSD systems
1270 that have them, for controlling the resources used by pipe deliveries.
1272 PH/42 The content-scanning code uses fopen() to create files in which to put
1273 message data. Previously it was not paying any attention to the mode of
1274 the files. Exim runs with umask(0) because the rest of the code creates
1275 files with open(), and sets the required mode explicitly. Thus, these
1276 files were ending up world-writeable. This was not a big issue, because,
1277 being within the spool directory, they were not world-accessible. I have
1278 created a function called modefopen, which takes an additional mode
1279 argument. It sets umask(777), creates the file, chmods it to the required
1280 mode, then resets the umask. All the relevant calls to fopen() in the
1281 content scanning code have been changed to use this function.
1283 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1284 to 24 hours. This avoids potential overflow problems when processing G
1285 and H retry rules. I suspect nobody ever tinkers with this value.
1287 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1289 PH/45 When the plaintext authenticator is running as a client, the server's
1290 challenges are checked to ensure they are valid base64 strings. By
1291 default, the authentication attempt is cancelled if an invalid string is
1292 received. Setting client_ignore_invalid_base64 true ignores these errors.
1293 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1294 they are received. Thus, the responses can be made to depend on the
1295 challenges. If an invalid string is ignored, an empty string is placed in
1298 PH/46 Messages that are created by the autoreply transport now contains a
1299 References: header, in accordance with RFCs 2822 and 3834.
1301 PH/47 Added authenticated_sender_force to the smtp transport.
1303 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1305 PH/49 Installed latest patch from the Sieve maintainer.
1307 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1308 was also set, the appendfile transport was unnecessarily scanning a
1309 directory of message files (e.g. for maildir delivery) to find the count
1310 of files (along with the size), even though it did not need this
1311 information. It now does the scan only if it needs to find either the
1312 size of the count of files.
1314 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1316 PH/52 Two bugs concerned with error handling when the smtp transport is
1319 (i) Exim was not creating retry information for temporary errors given
1320 for individual recipients after the DATA command when the smtp transport
1321 was used in LMTP mode. This meant that they could be retried too
1322 frequently, and not timed out correctly.
1324 (ii) Exim was setting the flag that allows error details to be returned
1325 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1326 recipients that were returned after the DATA command.
1328 PH/53 This is related to PH/52, but is more general: for any failing address,
1329 when detailed error information was permitted to be returned to the
1330 sender, but the error was temporary, then after the final timeout, only
1331 "retry timeout exceeded" was returned. Now it returns the full error as
1332 well as "retry timeout exceeded".
1334 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1335 do this, and (what is worse) MTAs that accept it.
1337 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1338 will now be deprecated.
1340 PH/56 New os.c-cygwin from the Cygwin maintainer.
1342 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1343 formats (previously only available in exim formats via -bpr, -bpru,
1344 and -bpra. Now also available in native and exiqgrep formats)
1346 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1347 with very large, slow to parse queues
1349 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1351 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1353 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1354 responses to authentication challenges, though it was showing the
1355 challenges; (ii) I've removed the CR characters from the debug output for
1358 PH/58 Allow for the insertion of a newline as well as a space when a string
1359 is turned into more than one encoded-word during RFC 2047 encoding. The
1360 Sieve code now uses this.
1362 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1363 data_4xx, lost_connection, tls_required.
1365 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1366 message was being sent as an SMTP response.
1368 PH/61 Add -l and -k options to exicyclog.
1370 PH/62 When verifying, if an address was redirected to one new address, so that
1371 verification continued, and the new address failed or deferred after
1372 having set something in $address_data, the value of $address_data was not
1373 passed back to the ACL. This was different to the case when no
1374 redirection occurred. The value is now passed back in both cases.
1376 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1377 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1378 use login_cap.h, so on its own it isn't the distinguishing feature. The
1379 new name refers directly to the setclassresources() function.
1381 PH/65 Added configuration files for NetBSD3.
1383 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1385 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1386 is preferred over IPv4.
1388 PH/68 The bounce_return_message and bounce_return_body options were not being
1389 honoured for bounces generated during the reception of non-SMTP messages.
1390 In particular, this applied to messages rejected by the ACL. This bug has
1391 been fixed. However, if bounce_return_message is true and bounce_return_
1392 body is false, the headers that are returned for a non-SMTP message
1393 include only those that have been read before the error was detected.
1394 (In the case of an ACL rejection, they have all been read.)
1396 PH/69 The HTML version of the specification is now built in a directory called
1397 spec_html instead of spec.html, because the latter looks like a path with
1398 a MIME-type, and this confuses some software.
1400 PH/70 Catch two compiler warnings in sieve.c.
1402 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1403 function verify_get_ident() calls ip_connect() to connect a socket, but
1404 if the "connect()" function timed out, ip_connect() used to close the
1405 socket. However, verify_get_ident() also closes the socket later, and in
1406 between Exim writes to the log, which may get opened at this point. When
1407 the socket was closed in ip_connect(), the log could get the same file
1408 descriptor number as the socket. This naturally causes chaos. The fix is
1409 not to close the socket in ip_connect(); the socket should be closed by
1410 the function that creates it. There was only one place in the code where
1411 this was missing, in the iplookup router, which I don't think anybody now
1412 uses, but I've fixed it anyway.
1414 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1415 well as to direct DNS lookups. Otherwise the handling of names in host
1416 lists is inconsistent and therefore confusing.
1422 PH/01 Two changes to the default runtime configuration:
1424 (1) Move the checks for relay_from_hosts and authenticated clients from
1425 after to before the (commented out) DNS black list checks.
1427 (2) Add control=submission to the relay_from_hosts and authenticated
1428 clients checks, on the grounds that messages accepted by these
1429 statements are most likely to be submissions.
1431 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1433 (1) Generate an error if the third argument for the ${prvs expansion is
1436 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1439 (3) Reset the variables that are obtained from the first argument of
1440 ${prvscheck and used in the second argument before leaving the code,
1441 because their memory is reclaimed, so using them afterwards may do
1444 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1445 one (it's much easier than Tom thought :-).
1447 (5) Because of (4), we can now allow for the use of $prvscheck_result
1448 inside the third argument.
1450 PH/03 For some reason, the default setting of PATH when running a command from
1451 a pipe transport was just "/usr/bin". I have changed it to
1454 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1455 anything to be listed in the output from -bV.
1457 PH/05 When a filter generated an autoreply, the entire To: header line was
1458 quoted in the delivery log line, like this:
1460 => >A.N.Other <ano@some.domain> <original@ddress> ...
1462 This has been changed so that it extracts the operative address. There
1463 may be more than one such address. If so, they are comma-separated, like
1466 => >ano@some.domain,ona@other.domain <original@ddress> ...
1468 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1469 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1470 not being looked up in the rDNS to get a host name, Exim was showing the
1471 IP address twice in Received: lines, even though the IP addresses were
1472 identical. For example:
1474 Received: from [1.2.3.4] (helo=[1.2.3.4])
1476 However, if the real host name was known, it was omitting the HELO data
1477 if it matched the actual IP address. This has been tidied up so that it
1478 doesn't show the same IP address twice.
1480 PH/07 When both +timestamp and +memory debugging was on, the value given by
1481 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1482 called by the debug printing, thereby overwriting the timestamp buffer.
1483 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1485 PH/08 When the original message was included in an autoreply transport, it
1486 always said "this is a copy of the message, including all the headers",
1487 even if body_only or headers_only was set. It now gives an appropriate
1490 PH/09 Applied a patch from the Sieve maintainer which:
1492 o fixes some comments
1493 o adds the (disabled) notify extension core
1494 o adds some debug output for the result of if/elsif tests
1495 o points to the current vacation draft in the documentation
1496 and documents the missing references header update
1500 o fixes a bug in processing the envelope test (when testing
1501 multiple envelope elements, the last element determinted the
1504 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1505 Electronic Mail") by including:
1507 Auto-submitted: auto-generated
1509 in the messages that it generates (bounce messages and others, such as
1510 warnings). In the case of bounce messages for non-SMTP mesages, there was
1511 also a typo: it was using "Auto_submitted" (underscore instead of
1512 hyphen). Since every message generated by Exim is necessarily in response
1513 to another message, thes have all been changed to:
1515 Auto-Submitted: auto-replied
1517 in accordance with these statements in the RFC:
1519 The auto-replied keyword:
1521 - SHOULD be used on messages sent in direct response to another
1522 message by an automatic process,
1524 - MUST NOT be used on manually-generated messages,
1526 - MAY be used on Delivery Status Notifications (DSNs) and Message
1527 Disposition Notifications (MDNs),
1529 - MUST NOT be used on messages generated by automatic or periodic
1530 processes, except for messages which are automatic responses to
1533 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1534 to the default Received: header definition.
1536 PH/12 Added log selector acl_warn_skipped (default on).
1538 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1539 variables because (a) they are in the wrong storage pool and (b) even if
1540 they were copied, it wouldn't work properly because of the caching.
1542 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1543 checking when decoding. Apparently there are clients that generate
1544 overlong encoded strings. Why am I not surprised?
1546 PH/15 If the first argument of "${if match_address" was not empty, but did not
1547 contain an "@" character, Exim crashed. Now it writes a panic log message
1548 and treats the condition as false.
1550 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1552 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1553 extension "envelope-auth". The code is finished and in agreement with
1554 other implementations, but there is no documentation so far and in fact,
1555 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1556 not changing the active code.
1558 Print executed "if" and "elsif" statements when debugging is used. This
1559 helps a great deal to understand what a filter does.
1561 Document more things not specified clearly in RFC3028. I had all this
1562 sorted out, when out of a sudden new issues came to my mind. Oops."
1564 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1567 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1568 canonical form (as documented). However, after a host name lookup from
1569 the IP address, check_host() was doing a simple string comparison with
1570 addresses acquired from the DNS when checking that the found name did
1571 have the original IP as one of its addresses. Since any found IPv6
1572 addresses are likely to be in abbreviated form, the comparison could
1573 fail. Luckily, there already exists a function for doing the comparison
1574 by converting both addresses to binary, so now that is used instead of
1575 the text comparison.
1577 PH/20 There was another similar case to PH/19, when a complete host name was
1578 given in a host list; looking up its IP address could give an abbreviated
1579 form, whereas the current host's name might or might not be abbreviated.
1580 The same fix has been applied.
1586 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1587 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1590 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1591 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1593 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1595 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1596 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1597 Also, for "Auto-Submitted", treat anything other than "no" as
1598 non-personal, in accordance with RFC 3834. (Previously it treated
1599 anything starting "auto-" as non-personal.)
1601 TF/01 The control=submission/name=... option had a problem with syntax
1602 errors if the name included a slash character. The /name= option
1603 now slurps the rest of the string, so it can include any characters
1604 but it must come last in the list of options (after /sender_retain
1607 PH/05 Some modifications to the interface to the fake nameserver for the new
1615 TK/01 Added the "success_on_redirect" address verification option. See
1616 NewStuff for rationale and an example.
1618 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1620 PH/02 Patch to exigrep to allow it to work on syslog lines.
1622 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1623 fread() to skip over the body file's header line, because in Cygwin the
1624 header line is locked and is inaccessible.
1626 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1627 co-exist for some time) to make it clear that it is the Exim ID that is
1628 referenced, not the Message-ID: header line.
1630 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1631 string_format() function, because snprintf() does not exist on all
1634 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1635 ${stat: expansion item.
1637 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1638 protocol synchronization error", to keep the pedants happy.
1640 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1641 well as for IRIX systems, when gcc is being used. See the host.c source
1644 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1646 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1649 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1650 transport and to the smtp transport in LMTP mode.
1652 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1654 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1655 run a filter in a subprocess. This could lead to confusion in subsequent
1656 lookups in the parent process. There should also be a search_tidyup() at
1657 the end of the subprocess.
1659 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1660 only if the host matched helo_try_verify_hosts, which caused the
1661 verification to occur when the EHLO/HELO command was issued. The ACL just
1662 tested the remembered result. Now, if a previous verification attempt has
1663 not happened, "verify = helo" does it there and then.
1665 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1667 TK/03 Fix log output including CR from clamd.
1669 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1670 exist provoked a memory error which could cause a segfault.
1672 PH/15 Installed PCRE 6.2
1674 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1676 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1677 of the problem. Specifically, suggested +O2 rather than +O1 for the
1680 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1682 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1683 record, the port was not correctly logged when the outgoing_port log
1684 selector was set (it logged the transort's default port).
1686 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1687 fallback_hosts, and "hosts" in the smtp transport.
1689 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1690 host errors such as "Connection refused".
1692 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1693 authentication with radiusclient 0.4.9:
1695 - Error returned from rc_read_config was caught wrongly
1696 - Username/password not passed on to radius server due to wrong length.
1698 The presumption is that some radiusclient API changes for 4.51/PH/17
1699 were not taken care of correctly. The code is still untested by me (my
1700 Linux distribution still has 0.3.2 of radiusclient), but it was
1701 contributed by a Radius user.
1703 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1704 expanding the "port" option of the smtp transport.
1706 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1707 while reading a MIME header. Thanks to Tom Hughes for a patch.
1709 PH/24 Include config.h inside local_scan.h so that configuration settings are
1712 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1713 that in an ACL for RCPT (for example), you can examine exactly what was
1716 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1717 commands, but it was not correctly comparing the address with the actual
1718 client host address. Thus, it would show the EHLO address in Received:
1719 header lines when this was not necessary.
1721 PH/27 Added the % operator to ${eval:}.
1723 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1724 it should be ignoring failures (because with -C, for example, it has lost
1725 privilege). It wasn't ignoring creation failures other than "already
1728 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1731 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1732 because an input line was too long, either on its own, or by virtue of
1733 too many continuations, the temporary file was not being removed, and the
1734 return code was incorrect.
1736 PH/31 Missing "BOOL" in function definition in filtertest.c.
1738 PH/32 Applied Sieve patches from the maintainer.
1740 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1742 PH/33 Added "verify = not_blind".
1744 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1745 Local/Makefile (with some defaults set). These are used in built scripts
1746 such as exicyclog, but they have never been used in the exim_install
1747 script (though there are many overriding facilities there). I have
1748 arranged that the exim_install script now takes note of these two
1751 PH/35 Installed configuration files for Dragonfly.
1753 PH/36 When a locally submitted message by a trusted user did not contain a
1754 From: header, and the sender address was obtained from -f or from an SMTP
1755 MAIL command, and the trusted user did not use -F to supply a sender
1756 name, $originator_name was incorrectly used when constructing a From:
1757 header. Furthermore, $originator_name was used for submission mode
1758 messages from external hosts without From: headers in a similar way,
1759 which is clearly wrong.
1761 PH/37 Added control=suppress_local_fixups.
1763 PH/38 When log_selector = +received_sender was set, and the addition of the
1764 sender made the log line's construction buffer exactly full, or one byte
1765 less than full, an overflow happened when the terminating "\n" was
1768 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1769 when the result of a list match is failure because a DNS lookup failed.
1771 PH/40 RM_COMMAND is now used in the building process.
1773 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1774 the "build-* directories that it finds.
1776 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1777 address in a domain literal was a prefix of an interface address.
1779 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1780 when verifying a sender address, unless rewrite_headers is false.
1782 PH/44 Wrote a long comment about why errors_to addresses are verified as
1783 recipients, not senders.
1785 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1786 the ratelimit ACL was added.
1788 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1790 PH/46 Added extra information about PostgreSQL errors to the error string.
1792 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1793 avoiding the need to install special zones in a real server. This is
1794 backwards compatible; if it can't find the fake resolver, it drops back.
1795 Thus, both old and new test suites can be run.
1797 TF/02 Added util/ratelimit.pl
1799 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1800 clock is set back in time.
1802 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1803 Candler <B.Candler@pobox.com>.
1805 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1806 OK for addresses that are the result of redirections.
1808 PH/48 A number of further additions for the benefit of the new test suite,
1809 including a fake gethostbyname() that interfaces to the fake DNS resolver
1812 TF/06 The fix for widen_domains has also been applied to qualify_single and
1813 search_parents which are the other dnslookup options that can cause
1816 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1819 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1821 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1822 service" if it encountered an MX record with an empty target hostname.
1823 The message is now "an MX or SRV record indicated no SMTP service".
1825 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1826 if the DNS of the sending site is misconfigured. This is quite a
1827 common situation. This change restores the behaviour of treating a
1828 helo verification defer as a failure.
1830 PH/51 If self=fail was set on a router, the bounce message did not include the
1831 actual error message.
1837 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1839 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1840 command itself ended in error, the underlying message about the transport
1841 filter timeout was being overwritten with the pipe command error. Now the
1842 underlying error message should be appended to the second error message.
1844 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1846 PH/02 Reduce the amount of output that "make" produces by default. Full output
1847 can still be requested.
1849 PH/03 The warning log line about a condition test deferring for a "warn" verb
1850 was being output only once per connection, rather than after each
1851 occurrence (because it was using the same function as for successful
1852 "warn" verbs). This seems wrong, so I have changed it.
1854 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1855 it should not have, which might have caused a crash in the right
1856 circumstances, but probably never did.
1858 PH/04 Installed a modified version of Tony Finch's patch to make submission
1859 mode fix the return path as well as the Sender: header line, and to
1860 add a /name= option so that you can make the user's friendly name appear
1863 TF/03 Added the control = fakedefer ACL modifier.
1865 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1866 Mark Lowes for thorough testing.
1868 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1870 TK/03 Merged latest SRS patch from Miles Wilton.
1872 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1873 with the definition in sysexits.h (which is #included earlier).
1874 Fortunately, Exim does not actually use EX_OK. The code used to try to
1875 preserve the sysexits.h value, by assumimg that macro definitions were
1876 scanned for macro replacements. I have been disabused of this notion,
1877 so now the code just undefines EX_OK before #including unistd.h.
1879 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1880 in the smtp transport. When a block could not be written in a single
1881 write() function, the timeout was being re-applied to each part-write.
1882 This seems wrong - if the receiver was accepting one byte at a time it
1883 would take for ever. The timeout is now adjusted when this happens. It
1884 doesn't have to be particularly precise.
1886 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1887 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1889 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1890 without a domain if the check to <postmaster@domain> fails.
1892 SC/01 Eximstats: added -xls and the ability to specify output files
1893 (patch written by Frank Heydlauf).
1895 SC/02 Eximstats: use FileHandles for outputing results.
1897 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1899 SC/04 Eximstats: fixed display of large numbers with -nvr option
1901 SC/05 Eximstats: fixed merging of reports with empty tables.
1903 SC/06 Eximstats: added the -include_original_destination flag
1905 SC/07 Eximstats: removed tabs and trailing whitespace.
1907 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1909 TK/06 MBOX spool code: Add real "From " MBOX separator line
1910 so the .eml file is really in mbox format (even though
1911 most programs do not really care). Patch from Alex Miller.
1913 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1914 The latter is generated from $received_to and is only set if the
1915 message has one envelope recipient. SA can use these headers,
1916 obviously out-of-the-box. Patch from Alex Miller.
1918 PH/08 The ${def test on a variable was returning false if the variable's
1919 value was "0", contrary to what the specification has always said!
1920 The result should be true unless the variable is empty.
1922 PH/09 The syntax error of a character other than { following "${if
1923 def:variable_name" (after optional whitespace) was not being diagnosed.
1924 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1925 accidental colon was present, for example, could give incorrect results.
1927 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1928 result is used (not including appendfile, where other changes are about
1931 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1932 This involved changing a lot of size variables from int to off_t. It
1933 should work with maildirs and everything.
1935 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1936 spamd dying while we are connected to it.
1938 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1941 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1942 function for BDB 4.3.
1944 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1946 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1947 the sources for PCRE have been re-arranged and more files are now
1950 PH/15 The code I had for printing potentially long long variables in PH/11
1951 above was not the best (it lost precision). The length of off_t variables
1952 is now inspected at build time, and an appropriate printing format (%ld
1953 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1954 to be "long long int" or "long int". This is needed for the internal
1955 formatting function string_vformat().
1957 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1958 the configuration file to be ":syslog", then the script "guesses" where
1959 the logs files are, rather than using the compiled in default. In our
1960 case the guess is not the same as the compiled default, so the script
1961 suddenly stopped working when I started to use syslog. The patch checks
1962 to see if log_file_path is "". If so, it attempts to read it from exim
1963 with no configuration file to get the compiled in version, before it
1964 falls back to the previous guessing code."
1966 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1967 implementing BATV in an Exim configuration. See NewStuff for the gory
1970 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1971 Makefile that are specific to HP-UX.
1973 PH/18 If the "use_postmaster" option was set for a recipient callout together
1974 with the "random" option, the postmaster address was used as the MAIL
1975 FROM address for the random test, but not for the subsequent recipient
1976 test. It is now used for both.
1978 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1979 patch removes a few documentation additions to RFC 3028, because the
1980 latest draft now contains them. It adds the new en;ascii-case comparator
1981 and a new error check for 8bit text in MIME parts. Comparator and
1982 require names are now matched exactly. I enabled the subaddress
1983 extension, but it is not well tested yet (read: it works for me)."
1985 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1986 rework some of the code of TK/09 above to avoid the hardwired use of
1987 "%lld" and "long long". Replaced the call to snprintf() with a call to
1990 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1991 records point to non-existent hosts", "retry timeout exceeded", and
1992 "retry time not reached for any host after a long failure period".
1994 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1995 experimental DomainKeys support:
1997 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1998 (2) On an error such as an illegally used "control", the wrong name for
1999 the control was given.
2001 These problems did NOT occur unless DomainKeys support was compiled.
2003 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2005 PH/24 Added ${if match_ip condition.
2007 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2008 databases so that it will be absolutely obvious if a crash occurs in the
2009 DB library. This is a regular occurrence (often caused by mis-matched
2012 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2013 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2014 picked up on a user's system that detects such things. There doesn't seem
2015 to be a gcc warning option for this - only an attribute that has to be
2016 put on the function's prototype. It seems that in Fedora Core 4 they have
2017 set this on a number of new functions. No doubt there will be more in due
2020 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2021 specify a transport. However, if an address that was verified by such a
2022 router was the subject of a callout, Exim crashed because it tried to
2023 read the rcpt_include_affixes from the non-existent transport. Now it
2024 just assumes that the setting of that option is false. This bug was
2025 introduced by 4.51/PH/31.
2027 PH/28 Changed -d+all to exclude +memory, because that information is very
2028 rarely of interest, but it makes the output a lot bigger. People tend to
2029 do -d+all out of habit.
2031 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2032 code in os-type was giving problems when libc.so lives in lib64, like on
2035 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2036 aren't the modern standard, and it seems that some systems' include files
2037 don't always have them. Exim was already checking for some of the newer
2038 ones like T_AAAA, and defining it itself. I've added checks for all the
2039 record types that Exim uses.
2041 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2042 not automatically generating a new one, as it is supposed to. This
2043 prevented TLS from working. If the file did exist, but contained invalid
2044 data, a new version was generated, as expected. It was only the case of a
2045 non-existent file that was broken.
2047 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2048 with a change in libdomainkeys > 0.64.
2050 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2051 from DNS. If the selector record carries the flag, it now has
2052 precedence over the domain-wide flag.
2054 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2056 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2057 the use of a port name that isn't defined in /etc/services) occurred, the
2058 message was deferred as in a normal delivery, and thus remained on the
2059 spool, instead of being failed because of the mua_wrapper setting. This
2060 is now fixed, and I tidied up some of the mua_wrapper messages at the
2063 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2064 the messages in a hash of arrays rather than using individual hashes.
2065 This is a bit cleaner and results in dramatic memory savings, albeit
2066 at a slight CPU cost.
2068 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2069 as requested by Marc Sherman.
2071 SC/10 Eximstats: added histograms for user specified patterns as requested
2074 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2076 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2077 fopen() in the content-scanning modules that did not already have it.
2083 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2084 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2086 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2088 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2090 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2091 to test Sieve filters that use "vacation".
2093 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2094 that changes the way the GnuTLS parameters are stored in the cache file.
2095 The new format can be generated externally. For backward compatibility,
2096 if the data in the cache doesn't make sense, Exim assumes it has read an
2097 old-format file, and it generates new data and writes a new file. This
2098 means that you can't go back to an older release without removing the
2101 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2102 work if there are any delivery delays because "one_time" forces the
2103 parent to be marked "delivered", so its unseen clone is never tried
2104 again. For this reason, Exim now forbids the simultaneous setting of
2107 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2108 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2109 entry if you want to know the details. The fix, however, neglected to
2110 consider the case when local delivery batching is involved. The test for
2111 "previously delivered" was not happening when checking to see if an
2112 address could be batched with a previous (undelivered) one; under
2113 certain circumstances this could lead to multiple deliveries to the same
2116 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2117 in its include files, and this causes problems building Exim.
2119 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2120 header_syntax) but Exim was just ignoring anything given after a slash.
2121 In particular, this caused confusion with an attempt to use "verify =
2122 reverse_host_lookup/defer_ok". An error is now given when options are
2123 supplied for verify items that do not have them. (Maybe reverse_host_
2124 lookup should have a defer_ok option, but that's a different point.)
2126 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2127 defined by RFC 821) to 2048, because there were problems with some AUTH
2128 commands, and RFC 1869 says the size should be increased for extended
2129 SMTP commands that take arguments.
2131 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2134 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2135 "unknown" error; now it says that the functionality isn't in the binary.
2137 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2138 an address' error message when a string expansion fails (syntax or
2139 whatever). Otherwise the password may appear in the log. Following change
2140 PH/42 below, there is no longer a chance of it appearing in a bounce
2143 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2145 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2146 of its addresses was ever tried. (Bugzilla bug #2.)
2148 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2149 the result incorrectly in the debug output. (It correctly added a newline
2150 to what was transported.)
2152 TF/01 Added $received_time.
2154 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2155 commented out examples of how to interface to a virus scanner and to
2156 SpamAssassin. Also added commented examples of av_scanner and
2157 spamd_address settings.
2159 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2160 and controls are allowed in which ACLs. There were a couple of minor
2161 errors. Some of the entries in the conditions table (which is a table of
2162 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2163 negation of where the condition IS allowed.
2165 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2167 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2168 header file does not have a version number, so I've had to invent a new
2169 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2170 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2171 radiusclient), but it was contributed by a Radius user.
2173 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2174 files or directories, for OpenSSL.
2176 PH/19 When an Exim process that is running as root has to create an Exim log
2177 file, it does so in a subprocess that runs as exim:exim so as to get the
2178 ownership right at creation (otherwise, other Exim processes might see
2179 the file with the wrong ownership). There was no test for failure of this
2180 fork() call, which would lead to the process getting stuck as it waited
2181 for a non-existent subprocess. Forks do occasionally fail when resources
2182 run out. I reviewed all the other calls to fork(); they all seem to check
2185 PH/20 When checking for unexpected SMTP input at connect time (before writing
2186 the banner), Exim was not dealing correctly with a non-positive return
2187 from the read() function. If the client had disconnected by this time,
2188 the result was a log entry for a synchronization error with an empty
2189 string after "input=" when read() returned zero. If read() returned -1
2190 (an event I could not check), uninitialized data bytes were printed.
2191 There were reports of junk text (parts of files, etc) appearing after
2194 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2196 PH/22 Added support for macro redefinition, and (re)definition in between
2197 driver and ACL definitions.
2199 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2200 forgetting to use the resulting value; it was using the unexpanded value.
2202 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2203 hadn't been configured. The fix is from Juergen Kreileder, who
2204 understands it better than I do:
2206 "Here's what I see happening with three configured cyrus_sasl
2207 authenticators configured (plain, login, cram-md5):
2209 On startup auth_cyrus_sasl_init() gets called for each of these.
2210 This means three calls to sasl_listmech() without a specified mech_list.
2211 => SASL tests which mechs of all available mechs actually work
2212 => three warnings about OTP not working
2213 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2215 With the patch, sasl_listmech() also gets called three times. But now
2216 SASL's mech_list option is set to the server_mech specified in the the
2217 authenticator. Or in other words, the answer from sasl_listmech()
2218 gets limited to just the mech you're testing for (which is different
2220 => the return list contains just 'plain' or 'login', 'cram-md5' or
2221 nothing depending on the value of ob->server_mech.
2223 I've just tested the patch: Authentication still works fine,
2224 unavailable mechs specified in the exim configuration are still
2225 caught, and the auth.log warnings about OTP are gone."
2227 PH/25 When debugging is enabled, the contents of the command line are added
2228 to the debugging output, even when log_selector=+arguments is not
2231 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2232 answer is "GNU", and only if the return is "GNU/something" is the answer
2235 PH/27 $acl_verify_message is now set immediately after the failure of a
2236 verification in an ACL, and so is available in subsequent modifiers. In
2237 particular, the message can be preserved by coding like this:
2239 warn !verify = sender
2240 set acl_m0 = $acl_verify_message
2242 Previously, $acl_verify_message was set only while expanding "message"
2243 and "log_message" when a very denied access.
2245 PH/28 Modified OS/os.c-Linux with
2247 -#ifndef OS_LOAD_AVERAGE
2248 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2250 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2251 nomenclature these days.)
2253 PH/29 Installed patch from the Sieve maintainer that adds the options
2254 sieve_useraddress and sieve_subaddress to the redirect router.
2256 PH/30 In these circumstances:
2257 . Two addresses routed to the same list of hosts;
2258 . First host does not offer TLS;
2259 . First host accepts first address;
2260 . First host gives temporary error to second address;
2261 . Second host offers TLS and a TLS session is established;
2262 . Second host accepts second address.
2263 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2264 and peerdn, if requested) that were in fact used only for the second
2267 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2268 attention to any local part prefix or suffix that was matched by the
2269 router that accepted the address. It now behaves in the same way as it
2270 does for delivery: the affixes are removed from the local part unless
2271 rcpt_include_affixes is set on the transport.
2273 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2274 timeout during the DATA phase of an incoming message.
2276 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2277 applied a patch sanctioned by the Sieve maintainer.
2279 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2280 the uid or gid is negative. A case of a negative gid caused this to be
2281 noticed. The fix allows for either to be negative.
2283 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2284 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2287 PH/36 The change PH/12 above was broken. Fixed it.
2289 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2290 the grounds that routing the same address twice would always produce the
2291 same answer. This might have been true once, but it is certainly no
2292 longer true now. Routing a child address may depend on the previous
2293 routing that produced that child. Some complicated redirection strategies
2294 went wrong when messages had multiple recipients, and made Exim's
2295 behaviour dependent on the order in which the addresses were given.
2297 I have moved the duplicate checking until after the routing is complete.
2298 Exim scans the addresses that are assigned to local and remote
2299 transports, and removes any duplicates. This means that more work will be
2300 done, as duplicates will always all be routed, but duplicates are
2301 presumably rare, so I don't expect this is of any significance.
2303 For deliveries to pipes, files, and autoreplies, the duplicate checking
2304 still happens during the routing process, since they are not going to be
2307 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2308 It corrects a timeout issue with spamd. This is Ian's comment: "The
2309 background is that sometimes spamd either never reads data from a
2310 connection it has accepted, or it never writes response data. The exiscan
2311 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2312 blindly assumes that writes won't block so it may never time out."
2314 PH/39 Allow G after quota size as well as K and M.
2316 PH/40 The value set for $authenticated_id in an authenticator may not contain
2317 binary zeroes or newlines because the value is written to log lines and
2318 to spool files. There was no check on this. Now the value is run through
2319 the string_printing() function so that such characters are converted to
2320 printable escape sequences.
2322 PH/41 $message_linecount is a new variable that contains the total number of
2323 lines in the message. Compare $body_linecount, which is the count for the
2326 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2327 bounce and delay warning messages, except in certain special cases, which
2330 (a) An SMTP error message from a remote host;
2331 (b) A message specified in a :fail: redirection;
2332 (c) A message specified in a "fail" command in a system filter;
2333 (d) A message specified in a FAIL return from the queryprogram router;
2334 (e) A message specified by the cannot_route_message router option.
2336 In these cases only, Exim does include the error details in bounce and
2337 warning messages. There are also a few cases where bland messages such
2338 as "unrouteable address" or "local delivery error" are given.
2340 PH/43 $value is now also set for the "else" part of a ${run expansion.
2342 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2343 being worked on, but at least Exim now implements the latest version to
2346 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2347 process to complete was treated as a delivery failure, a timeout while
2348 writing the message to the pipe was logged, but erroneously treated as a
2349 successful delivery. Such timeouts include transport filter timeouts. For
2350 consistency with the overall process timeout, these timeouts are now
2351 treated as errors, giving rise to delivery failures by default. However,
2352 there is now a new Boolean option for the pipe transport called
2353 timeout_defer, which, if set TRUE, converts the failures into defers for
2354 both kinds of timeout. A transport filter timeout is now identified in
2357 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2358 systems where "make" and "gmake" are different, calling "gmake" at top
2359 level broke things. I've arranged for the value of $(MAKE) to be passed
2360 from the Makefile to this script so that it can call the same version of
2364 A note about Exim versions 4.44 and 4.50
2365 ----------------------------------------
2367 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2368 changes of various kinds. As a consequence, a big documentation update was
2369 needed. This delayed the release for rather longer than seemed good, especially
2370 in the light of a couple of (minor) security issues. Therefore, the changes
2371 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2372 release. So 4.44 and 4.50 are in effect two different branches that both start
2375 I have left the 4.50 change log unchanged; it contains all the changes since
2376 4.43. The change log for 4.44 is below; many of its items are identical to
2377 those for 4.50. This seems to be the most sensible way to preserve the
2378 historical information.
2384 1. Minor wording change to the doc/README.SIEVE file.
2386 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2387 computation of the current number of files was incorrect.
2389 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2390 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2393 4. Give more explanation in the error message when the command for a transport
2394 filter fails to execute.
2396 5. There are several places where Exim runs a non-Exim command in a
2397 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2398 was being done only for the command run by the queryprogram router. It is
2399 now done for all such subprocesses. The other cases are: ${run, transport
2400 filters, and the commands run by the lmtp and pipe transports.
2402 6. Added CONFIGURE_GROUP build-time option.
2404 7. Some older OS have a limit of 256 on the maximum number of file
2405 descriptors. Exim was using setrlimit() to set 1000 as a large value
2406 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2407 systems. I've change it so that if it can't get 1000, it tries for 256.
2409 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2410 was an oversight, and furthermore, ever since the addition of extra
2411 controls (e.g. 4.43/32), the checks on when to allow different forms of
2412 "control" were broken. There should now be diagnostics for all cases when a
2413 control that does not make sense is encountered.
2415 9. Added the /retain_sender option to "control=submission".
2417 10. $recipients is now available in the predata ACL (oversight).
2419 11. Tidy the search cache before the fork to do a delivery from a message
2420 received from the command line. Otherwise the child will trigger a lookup
2421 failure and thereby defer the delivery if it tries to use (for example) a
2422 cached ldap connection that the parent has called unbind on.
2424 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2425 of $address_data from the recipient verification was clobbered by the
2426 sender verification.
2428 13. The value of address_data from a sender verification is now available in
2429 $sender_address_data in subsequent conditions in the ACL statement.
2431 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2433 15. Added a new option "connect=<time>" to callout options, to set a different
2436 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2437 was its contents. (It was OK if the option was not defined at all.)
2439 17. A "Completed" log line is now written for messages that are removed from
2440 the spool by the -Mrm option.
2442 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2443 information about exactly what failed.
2445 19. Added -dd to debug only the daemon process.
2447 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2448 handles timeouts, both on the server side and network timeouts. Renamed the
2449 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2451 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2452 It is now set to "smtps".
2454 22. $host_address is now set to the target address during the checking of
2455 ignore_target_hosts.
2457 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2458 being passed; this would have caused $sender_host_name to have been used if
2459 matching the list had actually called for a host name (not very likely,
2460 since this list is usually IP addresses). A host name is now passed as
2463 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2464 code that specifies a non-restarting handler (typically sigaction() in
2465 modern systems) in an attempt to fix a rare and obscure crash bug.
2467 25. Narrowed the window for a race in the daemon that could cause it to ignore
2468 SIGCHLD signals. This is not a major problem, because they are used only to
2469 wake it up if nothing else does.
2471 26. A malformed maildirsize file could cause Exim to calculate negative values
2472 for the mailbox size or file count. Odd effects could occur as a result.
2473 The maildirsize information is now recalculated if the size or filecount
2476 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2477 support for a long time. Removed HAVE_SYS_VFS_H.
2479 28. Installed the latest version of exipick from John Jetmore.
2481 29. In an address list, if the pattern was not a regular expression, an empty
2482 subject address (from a bounce message) matched only if the pattern was an
2483 empty string. Non-empty patterns were not even tested. This was the wrong
2484 because it is perfectly reasonable to use an empty address as part of a
2485 database query. An empty address is now tested by patterns that are
2486 lookups. However, all the other forms of pattern expect the subject to
2487 contain a local part and a domain, and therefore, for them, an empty
2488 address still always fails if the pattern is not itself empty.
2490 30. Exim went into a mad DNS loop when attempting to do a callout where the
2491 host was specified on an smtp transport, and looking it up yielded more
2492 than one IP address.
2494 31. Re-factored the code for checking spool and log partition space into a
2495 function that finds that data and another that does the check. The former
2496 is then used to implement four new variables: $spool_space, $log_space,
2497 $spool_inodes, and $log_inodes.
2499 32. The RFC2047 encoding function was originally intended for short strings
2500 such as real names; it was not keeping to the 75-character limit for
2501 encoded words that the RFC imposes. It now respects the limit, and
2502 generates multiple encoded words if necessary. To be on the safe side, I
2503 have increased the buffer size for the ${rfc2047: expansion operator from
2506 33. It is now permitted to omit both strings after an "if" condition; if the
2507 condition is true, the result is "true". As before, when the second string
2508 is omitted, a false condition yields an empty string. This makes it less
2509 cumbersome to write custom ACL and router conditions.
2511 34. Failure to deliver a bounce message always caused it to be frozen, even if
2512 there was an errors_to setting on the router. The errors_to setting is now
2515 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2516 canonical form (fully expanded) before being placed in
2517 $sender_host_address.
2519 36. The table in the code that translates DNS record types into text (T_A to
2520 "A" for instance) was missing entries for NS and CNAME. It is just possible
2521 that this could have caused confusion if both these types were looked up
2522 for the same domain, because the text type is used as part of Exim's
2523 per-process caching. But the chance of anyone hitting this buglet seems
2526 37. The dnsdb lookup has been extended in a number of ways.
2528 (1) There is a new type, "zns", which walks up the domain tree until it
2529 finds some nameserver records. It should be used with care.
2531 (2) There is a new type, "mxh", which is like "mx" except that it returns
2532 just the host names, not the priorities.
2534 (3) It is now possible to give a list of domains (or IP addresses) to be
2535 looked up. The behaviour when one of the lookups defers can be
2536 controlled by a keyword.
2538 (4) It is now possible to specify the separator character for use when
2539 multiple records are returned.
2541 38. The dnslists ACL condition has been extended: it is now possible to supply
2542 a list of IP addresses and/or domains to be looked up in a particular DNS
2545 39. Added log_selector=+queue_time_overall.
2547 40. When running the queue in the test harness, wait just a tad after forking a
2548 delivery process, to get repeatability of debugging output.
2550 41. Include certificate and key file names in error message when GnuTLS fails
2551 to set them up, because the GnuTLS error message doesn't include the name
2552 of the failing file when there is a problem reading it.
2554 42. Allow both -bf and -bF in the same test run.
2556 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2558 44. The "Exiscan patch" is now merged into the mainline Exim source.
2560 45. Sometimes the final signoff response after QUIT could fail to get
2561 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2562 < 0 before doing a fflush(). This bug looks as though it goes back to the
2563 introduction of TLS in release 3.20, but "sometimes" must have been rare
2564 because the tests only now provoked it.
2566 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2567 (this can affect the format of dates).
2569 47. exim_tidydb, when checking for the continued existence of a message for
2570 which it has found a message-specific retry record, was not finding
2571 messages that were in split spool directories. Consequently, it was
2572 deleting retry records that should have stayed in existence.
2574 48. Steve fixed some bugs in eximstats.
2576 49. The SPA authentication driver was not abandoning authentication and moving
2577 on to the next authenticator when an expansion was forced to fail,
2578 contradicting the general specification for all authenticators. Instead it
2579 was generating a temporary error. It now behaves as specified.
2581 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2582 (the order specifies the preference for clients). The order is now AES256,
2583 AES128, 3DES, ARCFOUR128.
2585 51. Small patch to Sieve code - explicitly set From: when generating an
2588 52. Exim crashed if a remote delivery caused a very long error message to be
2589 recorded - for instance if somebody sent an entire SpamAssassin report back
2590 as a large number of 550 error lines. This bug was coincidentally fixed by
2591 increasing the size of one of Exim's internal buffers (big_buffer) that
2592 happened as part of the Exiscan merge. However, to be on the safe side, I
2593 have made the code more robust (and fixed the comments that describe what
2596 53. Now that there can be additional text after "Completed" in log lines (if
2597 the queue_time_overall log selector is set), a one-byte patch to exigrep
2598 was needed to allow it to recognize "Completed" as not the last thing in
2601 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2602 patch that reportedly fixes this has been added. I am not expert enough to
2603 create a test for it. This is what the patch creator wrote:
2605 "I found a little strange behaviour of ldap code when working with
2606 Windows 2003 AD Domain, where users was placed in more than one
2607 Organization Units. When I tried to give exim partial DN, the exit code
2608 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2609 But simultaneously result of request was absolutely normal ldap result,
2610 so I produce this patch..."
2612 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2613 REFERENCE, so I have modified the code to exclude the patch when that macro
2616 55. Some experimental protocols are using DNS PTR records for new purposes. The
2617 keys for these records are domain names, not reversed IP addresses. The
2618 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2619 leaves it alone. Component reversal etc. now happens only for IP addresses.
2622 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2624 57. Double the size of the debug message buffer (to 2048) so that more of very
2625 long debug lines gets shown.
2627 58. The exicyclog utility now does better if the number of log files to keep
2628 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2630 59. Two changes related to the smtp_active_hostname option:
2632 (1) $smtp_active_hostname is now available as a variable.
2633 (2) The default for smtp_banner uses $smtp_active_hostname instead
2634 of $primary_hostname.
2636 60. The host_aton() function is supposed to be passed a string that is known
2637 to be a valid IP address. However, in the case of IPv6 addresses, it was
2638 not checking this. This is a hostage to fortune. Exim now panics and dies
2639 if the condition is not met. A case was found where this could be provoked
2640 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2641 components; fortuitously, this particular loophole had already been fixed
2642 by change 4.50/55 above.
2644 If there are any other similar loopholes, the new check in host_aton()
2645 itself should stop them being exploited. The report I received stated that
2646 data on the command line could provoke the exploit when Exim was running as
2647 exim, but did not say which command line option was involved. All I could
2648 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2649 running as the user.
2652 61. There was a buffer overflow vulnerability in the SPA authentication code
2653 (which came originally from the Samba project). I have added a test to the
2654 spa_base64_to_bits() function which I hope fixes it.
2657 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2658 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2660 63. The daemon start-up calls getloadavg() while still root for those OS that
2661 need the first call to be done as root, but it missed one case: when
2662 deliver_queue_load_max is set with deliver_drop_privilege. This is
2663 necessary for the benefit of the queue runner, because there is no re-exec
2664 when deliver_drop_privilege is set.
2666 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2667 This has been fixed.
2669 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2670 was in use, was not putting the data itself into the right store pool;
2671 consequently, it could be overwritten for a subsequent message in the same
2672 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2675 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2677 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2678 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2679 the function were treating the return as a boolean value, which happened to
2680 work because 0=false and not-0=true, but is not correct code.
2682 68. The host_aton() function was not handling scoped IPv6 addresses (those
2683 with, for example, "%eth0" on the end) correctly.
2685 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2686 negated items (that is, ~something) in unsigned ints. Some compilers
2687 apparently mutter when there is no cast.
2689 70. If an address verification called from an ACL failed, and did not produce a
2690 user-specific message (i.e. there was only a "system" message), nothing was
2691 put in $acl_verify_message. In this situation, it now puts the system
2694 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2695 session; change 4.31/43 added the unwanted input to the log line - except
2696 that it did not do this in the start of session case. It now does.
2698 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2699 This is wrong and can cause the other end to generate a synchronization
2700 error if it is another Exim or anything else that does the synchronization
2701 check. A QUIT command is no longer sent after a timeout.
2703 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2704 during host lookups.
2706 74. The defer_ok option of callout verification was not working if it was used
2707 when verifying addresses in header lines, that is, for this case:
2709 verify = header_sender/callout=defer_ok
2711 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2712 those file descriptors could be used for SMTP connections. If anything
2713 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2714 could be sent to the SMTP client, causing chaos. The daemon now opens
2715 stdin, stdout, and stderr to /dev/null when it puts itself into the
2718 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2719 log by default. The user can override this with suitable Perl magic.
2721 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2722 the log message when discard triggers, was not working for the DATA ACL or
2723 for the non-SMTP ACL.
2725 78. Error message wording change in sieve.c.
2727 79. If smtp_accept_max_per_host was set, the number of connections could be
2728 restricted to fewer than expected, because the daemon was trying to set up
2729 a new connection before checking whether the processes handling previous
2730 connections had finished. The check for completed processes is now done
2731 earlier. On busy systems, this bug wouldn't be noticed because something
2732 else would have woken the daemon, and it would have reaped the completed
2735 80. If a message was submitted locally by a user whose login name contained one
2736 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2737 It caused a spool format error. I have fixed the spool reading code. A
2738 related problem was that the "from" clause in the Received: line became
2739 illegal because of the space(s). It is now covered by ${quote_local_part.
2741 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2744 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2746 83. Patch from Sieve maintainer to fix unterminated string problem in
2747 "vacation" handling.
2749 84. Some minor changes to the Linux configuration files to help with other
2750 OS variants using glibc.
2752 85. One more patch for Sieve to update vacation handling to latest spec.
2755 ----------------------------------------------------
2756 See the note above about the 4.44 and 4.50 releases.
2757 ----------------------------------------------------
2763 1. Change 4.43/35 introduced a bug that caused file counts to be
2764 incorrectly computed when quota_filecount was set in an appendfile
2767 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2768 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2771 3. Give more explanation in the error message when the command for a transport
2772 filter fails to execute.
2774 4. There are several places where Exim runs a non-Exim command in a
2775 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2776 was being done only for the command run by the queryprogram router. It is
2777 now done for all such subprocesses. The other cases are: ${run, transport
2778 filters, and the commands run by the lmtp and pipe transports.
2780 5. Some older OS have a limit of 256 on the maximum number of file
2781 descriptors. Exim was using setrlimit() to set 1000 as a large value
2782 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2783 systems. I've change it so that if it can't get 1000, it tries for 256.
2785 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2786 was an oversight, and furthermore, ever since the addition of extra
2787 controls (e.g. 4.43/32), the checks on when to allow different forms of
2788 "control" were broken. There should now be diagnostics for all cases when a
2789 control that does not make sense is encountered.
2791 7. $recipients is now available in the predata ACL (oversight).
2793 8. Tidy the search cache before the fork to do a delivery from a message
2794 received from the command line. Otherwise the child will trigger a lookup
2795 failure and thereby defer the delivery if it tries to use (for example) a
2796 cached ldap connection that the parent has called unbind on.
2798 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2799 of $address_data from the recipient verification was clobbered by the
2800 sender verification.
2802 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2803 was its contents. (It was OK if the option was not defined at all.)
2805 11. A "Completed" log line is now written for messages that are removed from
2806 the spool by the -Mrm option.
2808 12. $host_address is now set to the target address during the checking of
2809 ignore_target_hosts.
2811 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2812 being passed; this would have caused $sender_host_name to have been used if
2813 matching the list had actually called for a host name (not very likely,
2814 since this list is usually IP addresses). A host name is now passed as
2817 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2818 code that specifies a non-restarting handler (typically sigaction() in
2819 modern systems) in an attempt to fix a rare and obscure crash bug.
2821 15. Narrowed the window for a race in the daemon that could cause it to ignore
2822 SIGCHLD signals. This is not a major problem, because they are used only to
2823 wake it up if nothing else does.
2825 16. A malformed maildirsize file could cause Exim to calculate negative values
2826 for the mailbox size or file count. Odd effects could occur as a result.
2827 The maildirsize information is now recalculated if the size or filecount
2830 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2831 support for a long time. Removed HAVE_SYS_VFS_H.
2833 18. Updated exipick to current release from John Jetmore.
2835 19. Allow an empty sender to be matched against a lookup in an address list.
2836 Previously the only cases considered were a regular expression, or an
2839 20. Exim went into a mad DNS lookup loop when doing a callout where the
2840 host was specified on the transport, if the DNS lookup yielded more than
2843 21. The RFC2047 encoding function was originally intended for short strings
2844 such as real names; it was not keeping to the 75-character limit for
2845 encoded words that the RFC imposes. It now respects the limit, and
2846 generates multiple encoded words if necessary. To be on the safe side, I
2847 have increased the buffer size for the ${rfc2047: expansion operator from
2850 22. Failure to deliver a bounce message always caused it to be frozen, even if
2851 there was an errors_to setting on the router. The errors_to setting is now
2854 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2855 canonical form (fully expanded) before being placed in
2856 $sender_host_address.
2858 24. Updated eximstats to version 1.33
2860 25. Include certificate and key file names in error message when GnuTLS fails
2861 to set them up, because the GnuTLS error message doesn't include the name
2862 of the failing file when there is a problem reading it.
2864 26. Expand error message when OpenSSL has problems setting up cert/key files.
2867 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2868 (this can affect the format of dates).
2870 28. exim_tidydb, when checking for the continued existence of a message for
2871 which it has found a message-specific retry record, was not finding
2872 messages that were in split spool directories. Consequently, it was
2873 deleting retry records that should have stayed in existence.
2875 29. eximstats updated to version 1.35
2876 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2877 1.35 - bugfix such that pie charts by volume are generated correctly
2879 30. The SPA authentication driver was not abandoning authentication and moving
2880 on to the next authenticator when an expansion was forced to fail,
2881 contradicting the general specification for all authenticators. Instead it
2882 was generating a temporary error. It now behaves as specified.
2884 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2885 (the order specifies the preference for clients). The order is now AES256,
2886 AES128, 3DES, ARCFOUR128.
2888 31. Small patch to Sieve code - explicitly set From: when generating an
2891 32. Exim crashed if a remote delivery caused a very long error message to be
2892 recorded - for instance if somebody sent an entire SpamAssassin report back
2893 as a large number of 550 error lines. This bug was coincidentally fixed by
2894 increasing the size of one of Exim's internal buffers (big_buffer) that
2895 happened as part of the Exiscan merge. However, to be on the safe side, I
2896 have made the code more robust (and fixed the comments that describe what
2899 33. Some experimental protocols are using DNS PTR records for new purposes. The
2900 keys for these records are domain names, not reversed IP addresses. The
2901 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2902 leaves it alone. Component reversal etc. now happens only for IP addresses.
2905 34. The host_aton() function is supposed to be passed a string that is known
2906 to be a valid IP address. However, in the case of IPv6 addresses, it was
2907 not checking this. This is a hostage to fortune. Exim now panics and dies
2908 if the condition is not met. A case was found where this could be provoked
2909 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2910 components; fortuitously, this particular loophole had already been fixed
2911 by change 4.50/55 or 4.44/33 above.
2913 If there are any other similar loopholes, the new check in host_aton()
2914 itself should stop them being exploited. The report I received stated that
2915 data on the command line could provoke the exploit when Exim was running as
2916 exim, but did not say which command line option was involved. All I could
2917 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2918 running as the user.
2921 35. There was a buffer overflow vulnerability in the SPA authentication code
2922 (which came originally from the Samba project). I have added a test to the
2923 spa_base64_to_bits() function which I hope fixes it.
2926 36. The daemon start-up calls getloadavg() while still root for those OS that
2927 need the first call to be done as root, but it missed one case: when
2928 deliver_queue_load_max is set with deliver_drop_privilege. This is
2929 necessary for the benefit of the queue runner, because there is no re-exec
2930 when deliver_drop_privilege is set.
2932 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2933 was in use, was not putting the data itself into the right store pool;
2934 consequently, it could be overwritten for a subsequent message in the same
2935 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2938 38. Sometimes the final signoff response after QUIT could fail to get
2939 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2940 < 0 before doing a fflush(). This bug looks as though it goes back to the
2941 introduction of TLS in release 3.20, but "sometimes" must have been rare
2942 because the tests only now provoked it.
2948 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2949 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2950 Now it returns an integer. A number of calls were still expecting a T/F
2951 return. Fortuitously, in all cases, the tests worked in OK situations,
2952 which is the norm. However, things would have gone wrong on any write
2953 failures on the smtp file descriptor. This function is used when sending
2954 messages over SMTP and also when doing verify callouts.
2956 2. When Exim is called to do synchronous delivery of a locally submitted
2957 message (the -odf or -odi options), it no longer closes stderr before doing
2960 3. Implemented the mua_wrapper option.
2962 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2964 5. Implemented the functions header_remove(), header_testname(),
2965 header_add_at_position(), and receive_remove_recipient(), and exported them
2968 6. If an ACL "warn" statement specified the addition of headers, Exim already
2969 inserted X-ACL-Warn: at the start if there was no header name. However, it
2970 was not making this test for the second and subsequent header lines if
2971 there were newlines in the string. This meant that an invalid header could
2972 be inserted if Exim was badly configured.
2974 7. Allow an ACL "warn" statement to add header lines at the start or after all
2975 the Received: headers, as well as at the end.
2977 8. Added the rcpt_4xx retry error code.
2979 9. Added postmaster_mailfrom=xxx to callout verification option.
2981 10. Added mailfrom=xxxx to the callout verification option, for verify=
2984 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2985 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2987 12. Inserted some casts to stop certain compilers warning when using pointer
2988 differences as field lengths or precisions in printf-type calls (mostly
2989 affecting debugging statements).
2991 13. Added optional readline() support for -be (dynamically loaded).
2993 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2994 same clock tick as a message's arrival, so that its received time was the
2995 same as the "first fail" time on the retry record, and that message
2996 remained on the queue past the ultimate address timeout, every queue runner
2997 would try a delivery (because it was past the ultimate address timeout) but
2998 after another failure, the ultimate address timeout, which should have then
2999 bounced the address, did not kick in. This was a "< instead of <=" error;
3000 in most cases the first failure would have been in the next clock tick
3001 after the received time, and all would be well.
3003 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3004 being recognized when the domain list was tested by the match_domain
3005 condition in an expansion string.
3007 16. Added the ${str2b64: operator.
3009 17. Exim was always calling setrlimit() to set a large limit for the number of
3010 processes, without checking whether the existing limit was already
3011 adequate. (It did check for the limit on file descriptors.) Furthermore,
3012 errors from getrlimit() and setrlimit() were being ignored. Now they are
3013 logged to the main and panic logs, but Exim does carry on, to try to do its
3014 job under whatever limits there are.
3016 18. Imported PCRE 5.0.
3018 19. Trivial typo in log message " temporarily refused connection" (the leading
3021 20. If the log selector return_path_on_delivery was set and an address was
3022 redirected to /dev/null, the delivery process crashed because it assumed
3023 that a return path would always be set for a "successful" delivery. In this
3024 case, the whole delivery is bypassed as an optimization, and therefore no
3027 21. Internal re-arrangement: the function for sending a challenge and reading
3028 a response while authentication was assuming a zero-terminated challenge
3029 string. It's now changed to take a pointer and a length, to allow for
3030 binary data in such strings.
3032 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3034 23. Exim was not respecting finduser_retries when seeking the login of the
3035 uid under which it was called; it was always trying 10 times. (The default
3036 setting of finduser_retries is zero.) Also, it was sleeping after the final
3037 failure, which is pointless.
3039 24. Implemented tls_on_connect_ports.
3041 25. Implemented acl_smtp_predata.
3043 26. If the domain in control=submission is set empty, Exim assumes that the
3044 authenticated id is a complete email address when it generates From: or
3045 Sender: header lines.
3047 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3048 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3049 chown and chgrp in /bin and hostname in /usr/bin.
3051 28. Exim was keeping the "process log" file open after each use, just as it
3052 does for the main log. This opens the possibility of it remaining open for
3053 long periods when the USR1 signal hits a daemon. Occasional processlog
3054 errors were reported, that could have been caused by this. Anyway, it seems
3055 much more sensible not to leave this file open at all, so that is what now
3058 29. The long-running daemon process does not normally write to the log once it
3059 has entered its main loop, and it closes the log before doing so. This is
3060 so that log files can straightforwardly be renamed and moved. However,
3061 there are a couple of unusual error situations where the daemon does write
3062 log entries, and I had neglected to close the log afterwards.
3064 30. The text of an SMTP error response that was received during a remote
3065 delivery was being truncated at 512 bytes. This is too short for some of
3066 the long messages that one sometimes sees. I've increased the limit to
3069 31. It is now possible to make retry rules that apply only when a message has a
3070 specific sender, in particular, an empty sender.
3072 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3073 it possible to be selective about when SMTP synchronization is enforced.
3075 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3077 32. Implemented hosts_connection_nolog.
3079 33. Added an ACL for QUIT.
3081 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3084 35. Added mailbox_size and mailbox_filecount to appendfile.
3086 36. Added control = no_multiline_responses to ACLs.
3088 37. There was a bug in the logic of the code that waits for the clock to tick
3089 in the case where the clock went backwards by a substantial amount such
3090 that the microsecond fraction of "now" was more than the microsecond
3091 fraction of "then" (but the whole seconds number was less).
3093 38. Added support for the libradius Radius client library this is found on
3094 FreeBSD (previously only the radiusclient library was supported).
3100 1. When certain lookups returned multiple values in the form name=value, the
3101 quoting of the values was not always being done properly. Specifically:
3102 (a) If the value started with a double quote, but contained no whitespace,
3104 (b) If the value contained whitespace other than a space character (i.e.
3105 tabs or newlines or carriage returns) it was not quoted.
3106 This fix has been applied to the mysql and pgsql lookups by writing a
3107 separate quoting function and calling it from the lookup code. The fix
3108 should probably also be applied to nisplus, ibase and oracle lookups, but
3109 since I cannot test any of those, I have not disturbed their existing code.
3111 2. A hit in the callout cache for a specific address caused a log line with no
3112 reason for rejecting RCPT. Now it says "Previous (cached) callout
3113 verification failure".
3115 3. There was an off-by-one bug in the queryprogram router. An over-long
3116 return line was truncated at 256 instead of 255 characters, thereby
3117 overflowing its buffer with the terminating zero. As well as fixing this, I
3118 have increased the buffer size to 1024 (and made a note to document this).
3120 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3121 when Exim is waiting for an SMTP response from a remote server, Exim
3122 restarts its select() call on the socket, thereby resetting its timeout.
3123 This is not a problem when such interrupts are rare. Somebody set up a cron
3124 job to run exiwhat every 2 minutes, which is less than the normal select()
3125 timeout (5 or 10 minutes). This meant that the select() timeout never
3126 kicked in because it was always reset. I have fixed this by comparing the
3127 time when an interrupt arrives with the time at the start of the first call
3128 to select(). If more time than the timeout has elapsed, the interrupt is
3129 treated as a timeout.
3131 5. Some internal re-factoring in preparation for the addition of Sieve
3132 extensions (by MH). In particular, the "personal" test is moved to a
3133 separate function, and given an option for scanning Cc: and Bcc: (which is
3134 not set for Exim filters).
3136 6. When Exim created an email address using the login of the caller as the
3137 local part (e.g. when creating a From: or Sender: header line), it was not
3138 quoting the local part when it contained special characters such as @.
3140 7. Installed new OpenBSD configuration files.
3142 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3143 try to make them clearer.
3145 9. Callout options, other than the timeout value, were being ignored when
3146 verifying sender addresses in header lines. For example, when using
3148 verify = header_sender/callout=no_cache
3150 the cache was (incorrectly) being used.
3152 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3153 only the Cygwin environment.
3155 11. When return_path_on_delivery was set as a log selector, if different remote
3156 addresses in the same message used different return paths and parallel
3157 remote delivery occurred, the wrong values would sometimes be logged.
3158 (Whenever a remote delivery process finished, the return path value from
3159 the most recently started remote delivery process was logged.)
3161 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3162 lines when AUTH and/or TLS are in use. This is the "received protocol"
3163 field. Exim used to use "asmtp" for authenticated SMTP, without any
3164 indication (in the protocol name) for TLS use. Now it follows the RFC and
3165 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3166 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3167 names appear in log lines as well as in Received: header lines.
3169 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3170 extensions, and comparison tests, and to fix some bugs.
3172 14. Changes to the "personal" filter test:
3174 (1) The test was buggy in that it was just doing the equivalent of
3175 "contains" tests on header lines. For example, if a user's address was
3176 anne@some.where, the "personal" test would incorrectly be true for
3178 To: susanne@some.where
3180 This test is now done by extracting each address from the header in turn,
3181 and checking the entire address. Other tests that are part of "personal"
3182 are now done using regular expressions (for example, to check local parts
3183 of addresses in From: header lines).
3185 (2) The list of non-personal local parts in From: addresses has been
3186 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3187 taken from the Sieve specification recommendations.
3189 (3) If the message contains any header line starting with "List-" it is
3190 treated as non-personal.
3192 (4) The test for "circular" in the Subject: header line has been removed
3193 because it now seems ill-conceived.
3195 15. Minor typos in src/EDITME comments corrected.
3197 16. Installed latest exipick from John Jetmore.
3199 17. If headers_add on a router specified a text string that was too long for
3200 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3201 of string_sprintf() is now avoided.
3203 18. $message_body_size was not set (it was always zero) when running the DATA
3204 ACL and the local_scan() function.
3206 19. For the "mail" command in an Exim filter, no default was being set for
3207 the once_repeat time, causing a random time value to be used if "once" was
3208 specified. (If the value happened to be <= 0, no repeat happened.) The
3209 default is now 0s, meaning "never repeat". The "vacation" command was OK
3210 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3211 (I found it when inspecting the code).
3213 20. There is now an overall timeout for performing a callout verification. It
3214 defaults to 4 times the callout timeout, which applies to individual SMTP
3215 commands during the callout. The overall timeout applies when there is more
3216 than one host that can be tried. The timeout is checked before trying the
3217 next host. This prevents very long delays if there are a large number of
3218 hosts and all are timing out (e.g. when the network connections are timing
3219 out). The value of the overall timeout can be changed by specifying an
3220 additional sub-option for "callout", called "maxwait". For example:
3222 verify = sender/callout=5s,maxwait=20s
3224 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3225 to the end before writing, but this should make it even safer).
3227 22. Exim was forgetting that it had advertised PIPELINING for the second and
3228 subsequent messages on an SMTP connection. It was also not resetting its
3229 memory on STARTTLS and an internal HELO.
3231 23. When Exim logs an SMTP synchronization error within a session, it now
3232 records whether PIPELINING has been advertised or not.
3234 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3235 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3236 rather than long int.
3238 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3240 26. Added the never_mail option to autoreply.
3246 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3247 crash if the getsockname() call failed; this can happen if a connection is
3248 closed very soon after it is established. The problem was simply in the
3249 order in which certain operations were done, causing Exim to try to write
3250 to the SMTP stream before it had set up the file descriptor. The bug has
3251 been fixed by making things happen in the correct order.
3257 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3258 before the connection was closed, thus losing the rejection response.
3260 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3261 some early Solaris releases, but causes trouble in current releases where
3262 socklen_t is defined.
3264 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3267 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3270 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3271 that releasing the top of it at the end releases what was used for sub-
3272 expansions (unless the block got too big). However, discard this block if
3273 the first thing is a variable or header, so that we can use its block when
3274 it is dynamic (useful for very large $message_headers, for example).
3276 6. Lookups now cache *every* query, not just the most recent. A new, separate
3277 store pool is used for this. It can be recovered when all lookup caches are
3278 flushed. Lookups now release memory at the end of their result strings.
3279 This has involved some general refactoring of the lookup sources.
3281 7. Some code has been added to the store_xxx() functions to reduce the amount
3282 of flapping under certain conditions.
3284 8. log_incoming_interface used to affect only the <= reception log lines. Now
3285 it causes the local interface and port to be added to several more SMTP log
3286 lines, for example "SMTP connection from", and rejection lines.
3288 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3290 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3292 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3293 could be overwritten at the end of the current message (or the start of a
3294 new message if it was set in a HELO ACL). The value is now preserved for
3295 the duration of the SMTP connection.
3297 12. If a transport had a headers_rewrite setting, and a matching header line
3298 contained an unqualified address, that address was qualified, even if it
3299 did not match any rewriting rules. The underlying bug was that the values
3300 of the flags that permit the existence of unqualified sender and recipient
3301 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3302 non-local messages, and by -bnq for local messages) were not being
3303 preserved with the message after it was received.
3305 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3306 "next input=" as part of the text comprising the host identity instead of
3307 the correct text. The code was using the same buffer for two different
3308 strings. However, depending on which order the printing function evaluated
3309 its arguments, the bug did not always show up. Under Linux, for example, my
3310 test suite worked just fine.
3312 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3313 doesn't work with some older versions of Perl. It has been changed to "my",
3314 which in any case is probably the better facility to use.
3316 15. A really picky compiler found some instances of statements for creating
3317 error messages that either had too many or two few arguments for the format
3320 16. The size of the buffer for calls to the DNS resolver has been increased
3321 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3322 for addresses that have a lot of PTR records. This alleviates a problem; it
3323 does not fully solve it.
3325 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3326 buffer now truncates the list and logs the incident, which is the same
3327 action as happens when Exim is looking up a host name and its aliases.
3328 Previously in this situation something unpredictable would happen;
3329 sometimes it was "internal error: store_reset failed".
3331 18. If a server dropped the connection unexpectedly when an Exim client was
3332 using GnuTLS and trying to read a response, the client delivery process
3333 crashed while trying to generate an error log message.
3335 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3336 string, for example:
3338 warn message = H1: something\nH2: something
3340 the text was added as a single header line from Exim's point of view
3341 though it ended up OK in the delivered message. However, searching for the
3342 second and subsequent header lines using $h_h2: did not work. This has been
3343 fixed. Similarly, if a system filter added multiple headers in this way,
3344 the routers could not see them.
3346 20. Expanded the error message when iplsearch is called with an invalid key to
3347 suggest using net-iplsearch in a host list.
3349 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3350 ACLs are no longer actually imposed (and a message to that effect is
3353 22. If a "gecos" field in a passwd entry contained escaped characters, in
3354 particular, if it contained a \" sequence, Exim got it wrong when building
3355 a From: or a Sender: header from that name. A second bug also caused
3356 incorrect handling when an unquoted " was present following a character
3357 that needed quoting.
3359 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3360 was not being matched caselessly.
3362 24. Arranged for all hyphens in the exim.8 source to be escaped with
3365 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3366 properly. Recipient callout cache records were still being keyed to include
3367 the sender, even when use_sender was set false. This led to far more
3368 callouts that were necessary. The sender is no longer included in the key
3369 when use_sender is false.
3371 26. Added "control = submission" modifier to ACLs.
3373 27. Added the ${base62d: operator to decode base 62 numbers.
3375 28. dnsdb lookups can now access SRV records.
3377 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3378 the configuration file.
3380 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3381 (-v) mode. This makes the output for a verbose queue run more intelligible.
3383 31. Added a use_postmaster feature to recipient callouts.
3385 32. Added the $body_zerocount variable, containing the number of binary zero
3386 bytes in the message body.
3388 33. The time of last modification of the "new" subdirectory is now used as the
3389 "mailbox time last read" when there is a quota error for a maildir
3392 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3394 35. Added +ignore_unknown as a special item in host lists.
3396 36. Code for decoding IPv6 addresses in host lists is now included, even if
3397 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3398 address was recognized as an IP address, but was then not correctly decoded
3399 into binary, causing unexpected and incorrect effects when compared with
3406 1. Very minor rewording of debugging text in manualroute to say "list of
3407 hosts" instead of "hostlist".
3409 2. If verify=header_syntax was set, and a header line with an unqualified
3410 address (no domain) and a large number of spaces between the end of the
3411 name and the colon was received, the reception process suffered a buffer
3412 overflow, and (when I tested it) crashed. This was caused by some obsolete
3413 code that should have been removed. The fix is to remove it!
3415 3. When running in the test harness, delay a bit after writing a bounce
3416 message to get a bit more predictability in the log output.
3418 4. Added a call to search_tidyup() just before forking a reception process. In
3419 theory, someone could use a lookup in the expansion of smtp_accept_max_
3420 per_host which, without the tidyup, could leave open a database connection.
3422 5. Added the variables $recipient_data and $sender_data which get set from a
3423 lookup success in an ACL "recipients" or "senders" condition, or a router
3424 "senders" option, similar to $domain_data and $local_part_data.
3426 6. Moved the writing of debug_print from before to after the "senders" test
3429 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3430 problems for message scanning, either using a data ACL, or using
3431 local_scan() because the Received: header was not generated till after they
3432 were called (in order to set the time as the time of reception completion).
3433 I have revised the way this works. The header is now generated after the
3434 body is received, but before the ACL or local_scan() are called. After they
3435 are run, the timestamp in the header is updated.
3441 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3442 before starting a queue runner without re-exec. This happened only when
3443 deliver_drop_privilege was set or when the Exim user was set to root. The
3444 effect of the bug was that timeouts during subsequent deliveries caused
3445 crashes instead of being properly handled. The handler is now left at its
3446 default (and expected) setting.
3448 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3449 message, again when deliver_drop_privilege is set or Exim is run as root.
3450 The bug described in (1) was not present in this case, but the tidying up
3451 of the other signals was missing. I have made the two cases consistent.
3453 3. The ignore_target_hosts setting on a manualroute router was being ignored
3454 for hosts that were looked up using the /MX notation.
3456 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3459 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3460 operated on the sender address. After changing the $sender_address to <>
3461 for the sender address verify, Exim was re-instated it as the original
3462 (before rewriting) address, but remembering that it had rewritten it, so it
3463 wasn't rewriting it again. This bug also had the effect of breaking the
3464 sender address verification caching when the sender address was rewritten.
3466 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3467 This has been changed so that if the ip literal address matches
3468 ignore_target_hosts, the router declines.
3470 7. Added expansion conditions match_domain, match_address, and match_local_
3471 part (NOT match_host).
3473 8. The placeholder for the Received: header didn't have a length field set.
3475 9. Added code to Exim itself and to exim_lock to test for a specific race
3476 condition that could lead to file corruption when using MBX delivery. The
3477 issue is with the lockfile that is created in /tmp. If this file is removed
3478 after a process has opened it but before that process has acquired a lock,
3479 there is the potential for a second process to recreate the file and also
3480 acquire a lock. This could lead to two Exim processes writing to the file
3481 at the same time. The added code performs the same test as UW imapd; it
3482 checks after acquiring the lock that its file descriptor still refers to
3483 the same named file.
3485 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3486 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3489 11. Added the smtp_active_hostname option. If used, this will typically be made
3490 to depend on the incoming interface address. Because $interface_address is
3491 not set up until the daemon has forked a reception process, error responses
3492 that can happen earlier (such as "too many connections") no longer contain
3495 12. If an expansion in a condition on a "warn" statement fails because a lookup
3496 defers, the "warn" statement is abandoned, and the next ACL statement is
3497 processed. Previously this caused the whole ACL to be aborted.
3499 13. Added the iplsearch lookup type.
3501 14. Added ident_timeout as a log selector.
3503 15. Added tls_certificate_verified as a log selector.
3505 16. Added a global option tls_require_ciphers (compare the smtp transport
3506 option of the same name). This controls incoming TLS connections.
3508 17. I finally figured out how to make tls_require_ciphers do a similar thing
3509 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3510 before starting the TLS session.
3512 18. Tabs are now shown as \t in -bP output.
3514 19. If the log selector return_path_on_delivery was set, Exim crashed when
3515 bouncing a message because it had too many Received: header lines.
3517 20. If two routers both had headers_remove settings, and the first one included
3518 a superfluous trailing colon, the final name in the first list and the
3519 first name in the second list were incorrectly joined into one item (with a
3520 colon in the middle).
3526 1. Added -C and -D options to the exinext utility, mainly to make it easier
3527 to include in the automated testing, but these could be helpful when
3528 multiple configurations are in use.
3530 2. The exinext utility was not formatting the output nicely when there was
3531 an alternate port involved in the retry record key, nor when there was a
3532 message id as well (for retries that were specific to a specific message
3533 and a specific host). It was also confused by IPv6 addresses, because of
3534 the additional colons they contain. I have fixed the IPv4 problem, and
3535 patched it up to do a reasonable job for IPv6.
3537 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3538 delivery, the log line now contains "pipelined" if PIPELINING was used.
3540 4. An SMTP transport process used to panic and die if the bind() call to set
3541 an explicit outgoing interface failed. This has been changed; it is now
3542 treated in the same way as a connect() failure.
3544 5. A reference to $sender_host_name in the part of a conditional expansion
3545 that was being skipped was still causing a DNS lookup. This no longer
3548 6. The def: expansion condition was not recognizing references to header lines
3549 that used bh_ and bheader_.
3551 7. Added the _cache feature to named lists.
3553 8. The code for checking quota_filecount in the appendfile transport was
3554 allowing one more file than it should have been.
3556 9. For compatibility with Sendmail, the command line option
3564 and sets the incoming protocol and host name (for trusted callers). The
3565 host name and its colon can be omitted when only the protocol is to be set.
3566 Note the Exim already has two private options, -pd and -ps, that refer to
3567 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3568 "s", but I don't think that's a major issue.
3570 10. A number of refactoring changes to the code, none of which should affect
3573 (a) The number of logging options was getting close to filling up the
3574 32-bit word that was used as a bit map. I have split them into two classes:
3575 those that are passed in the argument to log_write(), and those that are
3576 only ever tested independently outside of that function. These are now in
3577 separate 32-bit words, so there is plenty of room for expansion again.
3578 There is no change in the user interface or the logging behaviour.
3580 (b) When building, for example, log lines, the code previously used a
3581 macro that called string_cat() twice, in order to add two strings. This is
3582 not really sufficiently general. Furthermore, there was one instance where
3583 it was actually wrong because one of the argument was used twice, and in
3584 one call a function was used. (As it happened, calling the function twice
3585 did not affect the overall behaviour.) The macro has been replaced by a
3586 function that can join an arbitrary number of extra strings onto a growing
3589 (c) The code for expansion conditions now uses a table and a binary chop
3590 instead of a serial search (which was left over from when there were very
3591 few conditions). Also, it now recognizes conditions like "pam" even when
3592 the relevant support is not compiled in: a suitably worded error message is
3593 given if an attempt is made to use such a condition.
3595 11. Added ${time_interval:xxxxx}.
3597 12. A bug was causing one of the ddress fields not to be passed back correctly
3598 from remote delivery subprocesses. The field in question was not being
3599 subsequently used, so this caused to problems in practice.
3601 13. Added new log selectors queue_time and deliver_time.
3603 14. Might have fixed a bug in maildirsizefile handling that threw up
3604 "unexpected character" debug warnings, and recalculated the data
3605 unnecessarily. In any case, I expanded the warning message to give more
3608 15. Added the message "Restricted characters in address" to the statements in
3609 the default ACL that block characters like @ and % in local parts.
3611 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3612 Three changes have been made:
3614 (a) There was a serious bug; a negative response to MAIL caused the whole
3615 recipient domain to be cached as invalid, thereby blocking all messages
3616 to all local parts at the same domain, from all senders. This bug has
3617 been fixed. The domain is no longer cached after a negative response to
3618 MAIL if the sender used is not empty.
3620 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3623 (c) A new callout option, "use_sender" has been added for people who want
3624 the modified behaviour.
3630 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3633 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3634 indeed breaks things for older releases.
3636 3. Added additional logging to the case where there is a problem reading data
3637 from a filter that is running in a subprocess using a pipe, in order to
3638 try to track down a specific problem.
3640 4. Testing facility fudge: when running in the test harness and attempting
3641 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3642 getting "No route to host". Convert this to a timeout.
3644 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3647 6. Some OS don't have socklen_t but use size_t instead. This affects the
3648 fifth argument of getsockopt() amongst other things. This is now
3649 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3650 can be set for individual OS. I have set it for SunOS5, OSF1, and
3651 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3652 some earlier ones do not.
3654 7. Change 4.30/15 was not doing the test caselessly.
3656 8. The standard form for an IPv6 address literal was being rejected by address
3657 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3658 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3659 this, as well as the form without the "IPv6" on the front (but only when
3660 address literals are enabled, of course).
3662 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3664 10. Exim crashed if a message with an empty sender address specified by -f
3665 encountered a router with an errors_to setting. This could be provoked only
3666 by a command such as
3670 where an empty string was supplied; "<>" did not hit this bug.
3672 11. Installed PCRE release 4.5.
3674 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3675 remained set. It is now erased.
3677 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3678 times from message ids (which are base 36 rather than the normal 62).
3680 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3681 were being counted as actual protocol errors, and logged if the log
3682 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3683 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3684 and DATA following a set of rejected RCPTs do not count as protocol errors.
3685 In other words, Exim assumes they were pipelined, though this may not
3686 actually be the case. Of course, in all cases the client gets an
3687 appropriate error code.
3689 15. If a lookup fails in an ACL condition, a message about the failure may
3690 be available; it is used if testing the ACL cannot continue, because most
3691 such messages specify what the cause of the deferral is. However, some
3692 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3693 that caused an old message to be retained and used if a later statement
3694 caused a defer, replacing the real cause of the deferral.
3696 16. If an IP address had so many PTR records that the DNS lookup buffer
3697 was not large enough to hold them, Exim could crash while trying to process
3698 the truncated data. It now detects and logs this case.
3700 17. Further to 4.21/58, another change has been made: if (and only if) the
3701 first line of a message (the first header line) ends with CRLF, a bare LF
3702 in a subsequent header line has a space inserted after it, so as not to
3703 terminate the header.
3705 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3706 unnecessarily, used atoi() instead of strtol(), and didn't check the
3707 termination when getting file sizes from file names by regex.
3709 19. Completely re-implemented the support for maildirsize files, in the light
3710 of a number of problems with the previous contributed implementation
3711 (4.30/29). In particular:
3713 . If the quota is zero, the maildirsize file is maintained, but no quota is
3716 . If the maildir directory does not exist, it is created before any attempt
3717 to write a maildirsize file.
3719 . The quota value in the file is just a cache; if the quota is changed in
3720 the transport, the new value overrides.
3722 . A regular expression is available for excluding directories from the
3725 20. The autoreply transport checks the characters in options that define the
3726 message's headers; it allows continued headers, but it was checking with
3727 isspace() after an embedded newline instead of explicitly looking for a
3730 21. If all the "regular" hosts to which an address was routed had passed their
3731 expiry times, and had not reached their retry times, the address was
3732 bounced, even if fallback hosts were defined. Now Exim should go on to try
3735 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3736 equivalent code in the SMTP transport. Some hosts send humungous responses
3737 to HELO/EHLO, more than 1024 it seems.
3739 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3740 gives compiler warnings in some environments. I've now done it "properly",
3743 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3744 (because of problems with the built-in one) was declared to return uschar *
3745 instead of char *, causing compiler failure.
3747 25. Fixed a file descriptor leak when processing alias/forward files.
3749 26. Fixed a minor format string issue in dbfn.c.
3751 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3753 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3754 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3757 29. When a sender address is verified, it is cached, to save repeating the test
3758 when there is more than one recipient in a message. However, when the
3759 verification involves a callout, it is possible for different callout
3760 options to be set for different recipients. It is too complicated to keep
3761 track of this in the cache, so now Exim always runs a verification when a
3762 callout is required, relying on the callout cache for the optimization.
3763 The overhead is duplication of the address routing, but this should not be
3766 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3767 to be verified with callout=postmaster, and the main callout worked but the
3768 postmaster check failed, the verification correctly failed. However, if a
3769 subsequent RCPT command asked for sender verification *without* the
3770 postmaster check, incorrect caching caused this verification also to fail,
3773 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3774 it was not caching the DNS options (qualify_single, search_parents) that
3775 were used when the lookup failed. A subsequent lookup with different
3776 options therefore always gave the same answer, though there were cases
3777 where it should not have. (Example: a "domains = !$mx_any" option on a
3778 dnslookup router: the "domains" option is always processed without any
3779 widening, but the router might have qualify_single set.) Now Exim uses the
3780 cached value only when the same options are set.
3782 32. Added John Jetmore's "exipick" utility to the distribution.
3784 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3785 than a timeout (e.g. a certificate is required, and is not provided), an
3786 Exim server now closes the connection immediately. Previously it waited for
3787 the client to close - but if the client is SSL, it seems that they each
3788 wait for each other, leading to a delay before one of them times out.
3790 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3791 maintained 0.8.x compatibility because I don't think many are using it, and
3792 it is clearly obsolete.
3794 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3797 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3798 client certificate was expired. A simple patch fixes this, though I don't
3799 understand the full logic of why the verify callback is called multiple
3802 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3803 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3804 which causes problems with some clients (such as the Certicom SSL Plus
3805 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3806 disables the coutermeasure allowing Eudora to connect."
3808 38. Exim was not checking that a write() to a log file succeeded. This could
3809 lead to Bad Things if a log got too big, in particular if it hit a file
3810 size limit. Exim now panics and dies if it cannot write to a log file, just
3811 as it does if it cannot open a log file.
3813 39. Modified OS/Makefile-Linux so that it now contains
3815 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3817 The two -D definitions ensure that Exim is compiled with large file
3818 support, which makes it possible to handle log files that are bigger than
3821 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3822 instance) a domain was checked against a named list that involved a lookup,
3823 causing $domain_data to be set, then another domain was checked against the
3824 same list, then the first domain was re-checked, the value of $domain_data
3825 after the final check could be wrong. In particular, if the second check
3826 failed, it could be set empty. This bug probably also applied to
3829 41. The strip_trailing_dot option was not being applied to the address given
3830 with the -f command-line option.
3832 42. The code for reading a message's header from the spool was incrementing
3833 $received_count, but never initializing it. This meant that the value was
3834 incorrect (doubled) while delivering a message in the same process in which
3835 it was received. In the most common configuration of Exim, this never
3836 happens - a fresh exec is done - but it can happen when
3837 deliver_drop_privilege is set.
3839 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3840 it now includes up to 150 characters of the unexpected data in the log
3843 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3844 and building data strings. The size of both of these buffers was 10 000
3845 bytes - far larger than anybody would *ever* want, thought I. Needless to
3846 say, somebody hit the limit. I have increased the maximum line length to
3847 20 000 and the maximum data length of concatenated lines to 100 000. I have
3848 also fixed two bugs, because there was no checking on these buffers. Tsk,
3849 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3850 buffer is too small.
3852 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3853 lsearch lookups. Now it does.
3855 46. When parsing a route_list item in a manualroute router, a fixed-length
3856 buffer was used for the list of hosts. I made this 1024 bytes long,
3857 thinking that nobody would ever have a list of hosts that long. Wrong.
3858 Somebody had a whole pile of complicated expansion conditions, and the
3859 string was silently truncated, leading to an expansion error. It turns out
3860 that it is easier to change to an unlimited length (owing to other changes
3861 that have happened since this code was originally written) than to build
3862 structure for giving a limitation error. The length of the item that
3863 expands into the list of hosts is now unlimited.
3865 47. The lsearch lookup could not handle data where the length of text line was
3866 more than 4095 characters. Such lines were truncated, leading to shortened
3867 data being returned. It should now handle lines of any length.
3869 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3870 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3873 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3874 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3875 no longer happens in comment lines. A list of replacements is now placed
3876 at the head of all of the source files, except those whose only change is
3877 to replace PERL_COMMAND in the very first #! line.
3879 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3880 messages on the queue, with a bottom-up merge sort, using code contributed
3881 by Michael Haardt. This should make operations like -bp somewhat faster on
3882 large queues. It won't affect queue runners, except when queue_run_in_order
3885 51. Installed eximstats 1.31 in the distribution.
3887 52. Added support for SRV lookups to the dnslookup router.
3889 53. If an ACL referred to $message_body or $message_body_end, the value was not
3890 reset for any messages that followed in the same SMTP session.
3892 54. The store-handling optimization for building very long strings was not
3893 differentiating between the different store pools. I don't think this
3894 actually made any difference in practice, but I've tidied it.
3896 55. While running the routers to verify a sender address, $sender_address
3897 was still set to the sender address. This is wrong, because when routing to
3898 send a bounce to the sender, it would be empty. Therefore, I have changed
3899 it so that, while verifying a sender address, $sender_address is set to <>.
3900 (There is no change to what happens when verifying a recipient address.)
3902 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3903 target A or AAAA records (if not already returned) without resetting the
3904 qualify_single or search_parents options of the DNS resolver. These are
3905 inappropriate in this case because the targets of MX and SRV records must
3906 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3907 target that, when qualified, matched something in the local domain. These
3908 two options are now turned off when doing these lookups.
3910 57. It seems that at least some releases of Reiserfs (which does not have the
3911 concept of a fixed number of inodes) returns zero and not -1 for the
3912 number of available inodes. This interacted badly with check_spool_inodes,
3913 which assumed that -1 was the "no such thing" setting. What I have done is
3914 to check that the total number of inodes is greater than zero before doing
3915 the test of how many are available.
3917 58. When a "warn" ACL statement has a log_message modifier, the message is
3918 remembered, and not repeated. This is to avoid a lot of repetition when a
3919 message has many recipients that cause the same warning to be written.
3920 Howewer, Exim was preserving the list of already written lines for an
3921 entire SMTP session, which doesn't seem right. The memory is now reset if a
3922 new message is started.
3924 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3925 debugging output unless log_rewrite was also set.
3927 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3928 of (int)(handle) when we know that handle contains (void *)(-1).
3930 61. The Exim daemon panic-logs an error return when it closes the incoming
3931 connection. However "connection reset by peer" seems to be common, and
3932 isn't really an error worthy of noting specially, so that particular error
3935 62. When Exim is trying to find all the local interfaces, it used to panic and
3936 die if the ioctl to get the interface flags failed. However, it seems that
3937 on at least one OS (Solaris 9) it is possible to have an interface that is
3938 included in the list of interfaces, but for which you get a failure error
3939 for this call. This happens when the interface is not "plumbed" into a
3940 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3941 failure of the "get flags" call assumes that the interface is down.
3943 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3944 makes life easier for people who are doing arithmetic on fields extracted
3945 from dates, where you often get leading zeros that should not be
3946 interpreted as octal.
3948 64. Added qualify_domain to the redirect router, to override the global
3951 65. If a pathologically long header line contained very many addresses (the
3952 report of this problem mentioned 10 000) and each of them was rewritten,
3953 Exim could use up a very large amount of memory. (It kept on making new
3954 copies of the header line as it rewrote, and never released the old ones.)
3955 At the expense of a bit more processing, the header rewriting function has
3956 been changed so that it no longer eats memory in this way.
3958 66. The generation of the Received: header has been moved from the time that a
3959 message starts to be received, to the time that it finishes. The timestamp
3960 in the Received: header should now be very close to that of the <= log
3961 line. There are two side-effects of this change:
3963 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3964 logged header lines no longer include the local Received: line, because
3965 it has not yet been created. The same applies to a copy of the message
3966 that is returned to a non-SMTP sender when a message is rejected.
3968 (b) When a filter file is tested using -bf, no additional Received: header
3969 is added to the test message. After some thought, I decided that this
3972 This change does not affect the value of $received_for. It is still set
3973 after address rewriting, but before local_scan() is called.
3975 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3977 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3978 gave an unhelpful panic error message, and a defer error. I have managed to
3979 change this behaviour so that it now rejects any supplied certificate,
3980 which seems right, as the list of acceptable certificates is empty.
3982 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3983 gave an unhelpful defer error. I have not managed to make this reject any
3984 supplied certificates, but the error message it gives is "no certificate
3985 supplied", which is not helpful.
3987 70. exigrep's output now also includes lines that are not associated with any
3988 message, but which match the given pattern. Implemented by a patch from
3989 Martin Sluka, which also tidied up the Perl a bit.
3991 71. Recipient callout verification, like sender verification, was using <> in
3992 the MAIL FROM command. This isn't really the right thing, since the actual
3993 sender may affect whether the remote host accepts the recipient or not. I
3994 have changed it to use the actual sender in the callout; this means that
3995 the cache record is now keyed on a recipient/sender pair, not just the
3996 recipient address. There doesn't seem to be a real danger of callout loops,
3997 since a callout by the remote host to check the sender would use <>.
3998 [SEE ABOVE: changed after hitting problems.]
4000 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4001 temporary errors. However, in the case of such a code being given after
4002 the end of a data transmission (i.e. after ".") Exim was failing to write
4003 a retry record for the message. (Yes, there was some broken host that was
4004 actually sending 8xx at this point.)
4006 73. An unknown lookup type in a host list could cause Exim to panic-die when
4007 the list was checked. (An example that provoked this was putting <; in the
4008 middle of a list instead of at the start.) If this happened during a DATA
4009 ACL check, a -D file could be left lying around. This kind of configuration
4010 error no longer causes Exim to die; instead it causes a defer errror. The
4011 incident is still logged to the main and panic logs.
4013 74. Buglet left over from Exim 3 conversion. The message "too many messages
4014 in one connection" was written to the rejectlog but not the mainlog, except
4015 when address rewriting (yes!) was being logged.
4017 75. Added write_rejectlog option.
4019 76. When a system filter was run not as root (that is, when system_filter_user
4020 was set), the values of the $n variables were not being returned to the
4021 main process; thus, they were not subsequently available in the $sn
4024 77. Added +return_path_on_delivery log selector.
4026 78. A connection timeout was being treated differently from recipients deferred
4027 when testing hosts_max_try with a message that was older than the host's
4028 retry timeout. (The host should not be counted, thus allowing all hosts to
4029 be tried at least once before bouncing.) This may have been the cause of an
4030 occasionally reported bug whereby a message would remain on the queue
4031 longer than the retry timeout, but would be bounced if a delivery was
4032 forced. I say "may" because I never totally pinned down the problem;
4033 setting up timeout/retry tests is difficult. See also the next item.
4035 79. The ultimate address timeout was not being applied to errors that involved
4036 a combination of host plus message (for example, a timeout on a MAIL
4037 command). When an address resolved to a number of possible hosts, and they
4038 were not all tried for each delivery (e.g. because of hosts_max_try), a
4039 message could remain on the queue longer than the retry timeout.
4041 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4044 81. Fixed an obscure SMTP outgoing bug which required at least the following
4045 conditions: (a) there was another message waiting for the same server;
4046 (b) the server returned 5xx to all RCPT commands in the first message so
4047 that the message was not completed; (c) the server dropped the connection
4048 or gave a negative response to the RSET that Exim sends to abort the
4049 transaction. The observed case was a dropped connection after DATA that had
4050 been sent in pipelining mode. That is, the server had advertised PIPELINING
4051 but was not implementing it correctly. The effect of the bug was incorrect
4052 behaviour, such as trying another host, and this could lead to a crash.
4058 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4059 and daemon.c were passed as pointers to ints; they should have been
4060 pointers to socklen_t variables (which are typically unsigned ints).
4062 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4065 3. Fixed a really odd bug that affected only the testing scheme; patching a
4066 certain fixed string in the binary changed the value of another string that
4067 happened to be identical to the end of the original first string.
4069 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4070 name", it returns that address as the IP address. On some operating
4071 systems (e.g. Solaris), it also passes back the IP address string as the
4072 "host name". However, on others (e.g. Linux), it passes back an empty
4073 string. Exim wasn't checking for this, and was changing the host name to an
4074 empty string, assuming it had been canonicized.
4076 5. Although rare, it is permitted to have more than one PTR record for a given
4077 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4078 all the names associated with an address, because they do in Solaris.
4079 However, it seems that they do not in Linux for data that comes from the
4080 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4081 I found this out when I moved to a new Linux workstation and tried to run
4082 the Exim test suite.
4084 To get round this problem I have changed the code so that it now does its
4085 own call to the DNS to look up PTR records when searching for a host name.
4086 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4087 addresses that are only in /etc/hosts are still found.
4089 This behaviour is, however, controlled by an option called host_lookup_
4090 order, which defaults to "bydns:byaddr". If people want to use the other
4091 order, or indeed, just use one or the other means of lookup, they can
4092 specify it in this variable.
4094 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4095 some operating systems, this comes back from gethostbyaddr() as an empty
4096 string, and this is what Exim used to test for. However, it seems that in
4097 other systems, "." is yielded. Exim now tests for this case too.
4099 7. The values of check_spool_space and check_log_space are now held internally
4100 as a number of kilobytes instead of an absolute number of bytes. If a
4101 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4102 kilobyte. This means that much larger values can be stored.
4104 8. Exim monitor: an attempt to get the action menu when not actually pointing
4105 at a message produces an empty menu entitled "No message selected". This
4106 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4107 no entries in it ("Shell widget menu has zero width and/or height"). So I
4108 have added a single, blank menu entry in this case.
4110 9. Added ${quote_local_part.
4112 10. MIME decoding is now applied to the contents of Subject: header lines when
4115 11. Now that a reference to $sender_host_address automatically causes a reverse
4116 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4117 host lookup before query-style lookups in lists that might use this
4118 variable. This has therefore been abolished, and the "net-" prefix is no
4119 longer necessary for query-style lookups.
4121 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4122 have been a typo for LFLAGS, so it has been changed.
4124 13. The install script calls Exim with "-C /dev/null" in order to find the
4125 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4126 to be output. Howeve, since Exim outputs its version number before the
4127 error, it didn't break the script. It just looked ugly. I fixed this by
4128 always allowing "-C /dev/null" if the caller is root.
4130 14. Ignore overlarge ACL variable number when reading spool file - insurance
4131 against a later release with more variables having written the file.
4133 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4134 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4135 this, as well as the form without the "IPv6" on the front.
4137 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4138 OS/Makefile-Darwin file.
4140 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4141 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4143 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4144 variable. However, some OS use ints for inodes. I've added cast to long int
4145 to get rid of the compiler warning.
4147 19. I had forgotten to lock out "/../" in configuration file names when
4148 ALT_CONFIG_PREFIX was set.
4150 20. Routers used for verification do not need to specify transports. However,
4151 if such a router generated a host list, and callout was configured, Exim
4152 crashed, because it could not find a port number from the (non-existent)
4153 transport. It now assumes port 25 in this circumstance.
4155 21. Added the -t option to exigrep.
4157 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4158 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4159 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4160 them is set, LOOKUP_LSEARCH is forced.
4162 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4164 24. Added sender and host information to the "rejected by local_scan()" log
4165 line; previously there was no indication of these.
4167 25. Added .include_if_exists.
4169 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4170 Linux. It turns out that not all file systems support this. Apparently some
4171 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4172 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4173 not supported on the file descriptor, is now ignored when Exim is trying to
4174 sync a directory. This applies only to Linux.
4176 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4178 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4179 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4182 29. Added support for maildirsize files from supplied patch (modified a bit).
4184 30. The use of :fail: followed by an empty string could lead Exim to respond to
4185 sender verification failures with (e.g.):
4187 550 Verification failed for <xxx>
4188 550 Sender verify failed
4190 where the first response line was missing the '-' that indicates it is not
4191 the final line of the response.
4193 31. The loop for finding the name of the user that called Exim had a hardwired
4194 limit of 10; it now uses the value of finduser_retries, which is used for
4195 all other user lookups.
4197 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4200 33. Exim was neglecting to zero errno before one call of strtol() when
4201 expanding a string and expecting an integer value. On some systems this
4202 resulted in spurious "integer overflow" errors. Also, it was casting the
4203 result into an int without checking.
4205 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4206 did not work. The code looks as if it has *never* worked, though it appears
4207 to have been documented since at least releast 1.62. I have made it work.
4209 35. The "timeout_DNS" error in retry rules, also documented since at least
4210 1.62, also never worked. As it isn't clear exactly what this means, and
4211 clearly it isn't a major issue, I have abolished the feature by treating it
4212 as "timeout", and writing a warning to the main and panic logs.
4214 36. The display of retry rules for -brt wasn't always showing the error code
4217 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4218 timeout_connect_A, timeout_connect_MX.
4220 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4221 to the empty sender.
4223 39. The daemon was not analysing the content of -oX till after it had closed
4224 stderr and disconnected from the controlling terminal. This meant that any
4225 syntax errors were only noted on the panic log, and the return code from
4226 the command was 0. By re-arranging the code a little, I've made the
4227 decoding happen first, so such errors now appear on stderr, and the return
4228 code is 1. However, the actual setting up of the sockets still happens in
4229 the disconnected process, so errors there are still only recorded on the
4232 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4233 connections (as happens on some IP stacks) was logged at start up time as
4234 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4235 it from "IPv6 and IPv4", which means that two separate sockets are being
4238 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4239 says whether AF_INET or AF_INET6 was passed as an argument.
4241 42. Exiwhat output was messed up when time zones were included in log
4244 43. Exiwhat now gives more information about the daemon's listening ports,
4245 and whether -tls-on-connect was used.
4247 44. The "port" option of the smtp transport is now expanded.
4249 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4250 silently ignored. Now an error message is written to the main and panic
4253 46. There's a new ACL modifier called "logwrite" which writes to a log file
4254 as soon as it is encountered.
4256 47. Added $local_user_uid and $local_user_gid at routing time.
4258 48. Exim crashed when trying to verify a sender address that was being
4261 49. Exim was recognizing only a space character after ".include". It now also
4262 recognizes a tab character.
4264 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4265 extracting the relevant information from the specification. The man page no
4266 longer contains scrambled data for the -d option, and I've added a section
4267 at the front about calling Exim under different names.
4269 51. Added "extra_headers" argument to the "mail" command in filter files.
4271 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4274 53. Installed eximstats 1.29.
4276 54. Added transport_filter_timeout as a generic transport option.
4278 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4279 Cc: header lines. This was required by RFC 822, but it not required by RFC
4282 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4283 incoming messages that did not have them. Now it does so only if the
4284 message originates locally, that is, if there is no associated remote host
4285 address. When Resent- header lines are present, this applies to the Resent-
4286 lines rather than the non-Resent- lines.
4288 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4289 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4291 58. Messages for configuration errors now include the name of the main
4292 configuration file - useful now that there may be more than one file in a
4293 list (.included file names were always shown).
4295 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4296 for those rare installations that do not start the daemon as root or run it
4297 setuid root. I've cut out the call to initgroups() if the daemon is not
4300 60. The Exim user and group can now be bound into the binary as text strings
4301 that are looked up at the start of Exim's processing.
4303 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4305 62. Added $mailstore_basename variable.
4307 63. Installed patch to sieve.c from Michael Haardt.
4309 64. When Exim failed to open the panic log after failing to open the main log,
4310 the original message it was trying to log was written to stderr and debug
4311 output, but if they were not available (the usual case in production), it
4312 was lost. Now it is written to syslog before the two lines that record the
4313 failures to open the logs.
4315 65. Users' Exim filters run in subprocesses under the user's uid. It is
4316 possible for a "deliver" command or an alias in a "personal" command to
4317 provoke an address rewrite. If logging of address rewriting is configured,
4318 this fails because the process is not running as root or exim. There may be
4319 a better way of dealing with this, but for the moment (because 4.30 needs
4320 to be released), I have disabled address rewrite logging when running a
4321 filter in a non-root, non-exim process.
4327 1. The buildconfig auxiliary program wasn't quoting the value set for
4328 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4329 not defined. This bug was masked in 4.22 by the effect that was fixed in
4332 2. Some messages that were rejected after a message id was allocated were
4333 shown as "incomplete" by exigrep. It no longer does this for messages that
4334 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4336 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4337 have allow_domain_literals set, the ID did not get logged in the <= line.
4338 Domain literals are now always recognized in Message-ID: header lines.
4340 4. The first argument for a ${extract expansion item is the key name or field
4341 number. Leading and trailing spaces in this item were not being ignored,
4342 causing some misleading effects.
4344 5. When deliver_drop_privilege was set, single queue runner processes started
4345 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4346 same command in the process it spins off) were not dropping privilege.
4348 6. When the daemon running as "exim" started a queue runner, it always
4349 re-executed Exim in the spun-off process. This is a waste of effort when
4350 deliver_drop_privilege is set. The new process now just calls the
4351 queue-runner function directly.
4357 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4360 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4361 ignored. Though the use of -oP was forcing the writing of a pid file, it
4362 was always written to the default place.
4364 3. If the message "no IP address found for host xxxx" is generated during
4365 incoming verification, it is now followed by identification of the incoming
4366 connection (so you can more easily find what provoked it).
4368 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4370 5. Added some features to "harden" Exim a bit more against certain attacks:
4372 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4373 be put in Local/Makefile. This is like the never_users runtime option,
4374 but it cannot be overridden. The default setting is "root".
4376 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4377 prefix string with which any file named in a -C command line option
4380 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4381 is retained for -C and -D only if the caller of Exim is root. Without
4382 it, the exim user may also use -C and -D and retain privilege.
4384 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4385 command line option is disabled.
4387 6. Macro names set by the -D option must start with an upper case letter, just
4388 like macro names defined in the configuration file.
4390 7. Added "dereference=" facility to LDAP.
4392 8. Two instances of the typo "uknown" in the source files are fixed.
4394 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4395 the Configure-Makefile script screwed up while processing it.
4397 10. Incorporated PCRE 4.4.
4399 11. The SMTP synchronization check was not operating right at the start of an
4400 SMTP session. For example, it could not catch a HELO sent before the client
4401 waited for the greeting. There is now a check for outstanding input at the
4402 point when the greeting is written. Because of the duplex, asynchronous
4403 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4404 way, but not yet received, when the check is performed.
4406 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4407 on TCP/IP sockets, because this apparently causes some broken clients to
4410 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4411 unchanged) from the Cygwin maintainer.
4413 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4414 was supported instead of testing for mbx. Effectively a typo.
4416 15. The spa authenticator server code was not checking that the input it
4417 received was valid base64.
4419 16. The debug output line for the "set" modifier in ACLs was not showing the
4420 name of the variable that was being set.
4422 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4424 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4425 lookup on its own. Something else was needed to trigger the lookup. For
4426 example, a match in host_lookup or the need for a host name in a host list.
4427 Now, if $sender_host_name is referenced and the host name has not yet been
4428 looked up, a lookup is performed. If the lookup fails, the variable remains
4429 empty, and $host_lookup_failed is set to "1".
4431 19. Added "eqi" as a case-independent comparison operator.
4433 20. The saslauthd authentication condition could segfault if neither service
4434 nor realm was specified.
4436 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4437 error message that was logged was misleading, and incoming SMTP
4438 connections were dropped. The message is now more accurate, and temporary
4439 errors are given to SMTP connections.
4441 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4442 (except RSET) with a 421 error, until QUIT is received. However, it was
4443 failing to send a response to QUIT.
4445 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4446 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4447 ACL is now run after the helo_try_verify_hosts code.
4449 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4450 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4451 case-independent, so other case variants are also recognized). Apparently
4452 some systems use these upper case variants.
4454 25. If more than two messages were waiting for the same host, and a transport
4455 filter was specified for the transport, Exim sent two messages over the
4456 same TCP/IP connection, and then failed with "socket operation on non-
4457 socket" when it tried to send the third.
4459 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4461 27. The extern definition of crypt16() in expand.c was not being excluded when
4462 the OS had its own crypt16() function.
4464 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4465 as a preferred synonym for return_size_limit, both as an option and as an
4468 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4470 30. Changed the default configuration ACL to relax the local part checking rule
4471 for addresses that are not in any local domains. For these addresses,
4472 slashes and pipe symbols are allowed within local parts, but the sequence
4473 /../ is explicitly forbidden.
4475 31. SPA server authentication was not clearing the challenge buffer before
4478 32. log_message in a "warn" ACL statement was writing to the reject log as
4479 well as to the main log, which contradicts the documentation and doesn't
4480 seem right (because no rejection is happening). So I have stopped it.
4482 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4483 However, I am unable to do any testing of this.
4485 34. Fixed an infelicity in the appendfile transport. When checking directories
4486 for a mailbox, to see if any needed to be created, it was accidentally
4487 using path names with one or more superfluous leading slashes; tracing
4488 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4490 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4491 contents are added to the log line that is written for every discarded
4492 recipient. (Previously a log_message setting was ignored.)
4494 36. The ${quote: operator now quotes the string if it is empty.
4496 37. The install script runs exim in order to find its version number. If for
4497 some reason other than non-existence or emptiness, which it checks, it
4498 could not run './exim', it was installing it with an empty version number,
4499 i.e. as "exim-". This error state is now caught, and the installation is
4502 38. An argument was missing from the function that creates an error message
4503 when Exim fails to connect to the socket for saslauthd authentication.
4504 This could cause Exim to crash, or give a corrupted message.
4506 39. Added isip, isip4, and isip6 to ${if conditions.
4508 40. The ACL variables $acl_xx are now saved with the message, and can be
4509 accessed later in routers, transports, and filters.
4511 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4512 strings in the file are not string-expanded.
4514 42. If a MAIL command specified a SIZE value that was too large to fit into an
4515 int variable, the check against message_size_limit failed. Such values are
4516 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4517 day this will have to be increased, but I don't think I want to be around
4518 when emails are that large.
4525 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4526 iconv() is not standard in FreeBSD.
4528 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4529 IPv6 enabled. The observed symptom was a segmentation fault on return from
4530 the function os_common_find_running_interfaces() in src/os.c.
4532 3. In the check_special_case() function in daemon.c I had used "errno" as an
4533 argument name, which causes warnings on some systems. This was basically a
4534 typo, since it was named "eno" in the comments!
4536 4. The code that waits for the clock to tick (at a resolution of some fraction
4537 of a second) so as to ensure message-id uniqueness was always waiting for
4538 at least one whole tick, when it could have waited for less. [This is
4539 almost certainly not relevant at current processor speeds, where it is
4540 unlikely to ever wait at all. But we try to future-proof.]
4542 5. The function that sleeps for a time interval that includes fractions of a
4543 second contained a race. It did not block SIGALRM between setting the
4544 timer, and suspending (a couple of lines later). If the interval was short
4545 and the sigsuspend() was delayed until after it had expired, the suspension
4546 never ended. On busy systems this could lead to processes getting stuck for
4549 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4550 process, before it forks any delivery processes. The open lookup caching
4551 mechanism meant that the open file or database connection was passed into
4552 the delivery process. The problem was that delivery processes always tidy
4553 up cached lookup data. This could cause a problem for the next delivery
4554 process started by the queue runner, because the external queue runner
4555 process does not know about the closure. So the next delivery process
4556 still has data in the lookup cache. In the case of a file lookup, there was
4557 no problem because closing a file descriptor in a subprocess doesn't affect
4558 the parent. However, if the lookup was caching a connection to a database,
4559 the connection was closed, and the second delivery process was likely to
4560 see errors such as "PGSQL: query failed: server closed the connection
4561 unexpectedly". The problem has been fixed by closing all cached lookups
4562 in a queue runner before running a delivery process.
4564 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4565 seem to have the "const" qualifier which it has on other OS. I've
4568 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4569 *of the same type* (client or server) with the same public name that an
4570 error should be diagnosed.
4572 9. When Exim looked up a host name for an IP address, but failed to find the
4573 original IP address when looking up the host name (a safety check), it
4574 output the message "<ip address> does not match any IP for NULL", which was
4575 confusing, to say the least. The bug was that the host name should have
4576 appeared instead of "NULL".
4578 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4579 user that is built into the binary, and the -C or -D options is used, root
4580 privilege is dropped before the configuration file is read. In addition,
4581 logging is switched to stderr instead of the normal log files. If the
4582 configuration then re-defines the Exim user, the unprivileged environment
4583 is probably not what is expected, so Exim logs a panic warning message (but
4586 However, if deliver_drop_privilege is set, the unprivileged state may well
4587 be exactly what is intended, so the warning has been cut out in that case,
4588 and Exim is allowed to try to write to its normal log files.
4594 1. smtp_return_error_details was not giving details for temporary sender
4595 or receiver verification errors.
4597 2. Diagnose a configuration error if two authenticators have the same public
4600 3. Exim used not to create the message log file for a message until the first
4601 delivery attempt. This could be confusing when incoming messages were held
4602 for policy or load reasons. The message log file is now created at the time
4603 the message is received, and an initial "Received" line is written to it.
4605 4. The automatically generated man page for command line options had a minor
4606 bug that caused no ill effects; however, a more serious problem was that
4607 the procedure for building the man page automatically didn't always
4608 operate. Consequently, release 4.20 contains an out-of-date version. This
4609 shouldn't happen again.
4611 5. When building Exim with embedded Perl support, the script that builds the
4612 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4613 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4615 6. The freeze_tell option was not being used for messages that were frozen on
4616 arrival, either by an ACL or by local_scan().
4618 7. Added the smtp_incomplete_transaction log selector.
4620 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4621 was accepting AUTH without a new EHLO.
4623 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4624 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4627 10. Logging of TCP/IP connections (when configured) now happens in the main
4628 daemon process instead of the child process, so that the TCP/IP connection
4629 count is more accurate (but it can never be perfect).
4631 11. The use of "drop" in a nested ACL was not being handled correctly in the
4632 outer ACL. Now, if condition failure induced by the nested "drop" causes
4633 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4634 or "require"), the connection is dropped.
4636 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4637 that yield "discard" can now be used with an "accept" or a "discard" verb,
4638 but an error is generated for any others (because I can't see a useful way
4639 to define what should happen).
4641 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4642 are now processed in the same way as lines in the Exim configuration file.
4643 In particular, continuation lines are supported.
4645 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4647 15. Added -ti meaning -t -i.
4649 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4650 domains, and warn by logging if others are found.
4652 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4653 returned from GIFCONF when Exim is trying to find the list of interfaces on
4654 a host. The code in os.c has been modified to copy each ifreq to an aligned
4655 structure in all cases.
4657 Also, in some cases, the returned ifreq's were being copied to a 'struct
4658 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4659 means the last couple of bytes of an IPv6 address could be chopped if the
4660 ifreq contained only a normal sockaddr (14 bytes storage).
4662 18. Named domain lists were not supported in the hosts_treat_as_local option.
4663 An entry such as +xxxx was not recognized, and was treated as a literal
4666 19. Ensure that header lines added by a DATA ACL are included in the reject log
4667 if the ACL subsequently rejects the message.
4669 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4670 MD5 (which is deprecated).
4672 21. When testing a filter file using -bf, Exim was writing a message when it
4673 took the sender from a "From " line in the message, but it was not doing so
4674 when it took $return_path from a Return-Path: header line. It now does.
4676 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4677 with a valid header line field name (a series of printing characters
4678 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4680 23. Changed "disc" in the source to "disk" to conform to the documentation and
4681 the book and for uniformity.
4683 24. Ignore Sendmail's -Ooption=value command line item.
4685 25. When execve() failed while trying to run a command in a pipe transport,
4686 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4687 could be confused with a return value of 69 from the command itself. This
4688 has been changed to 127, the value the shell returns if it is asked to run
4689 a non-existent command. The wording for the related log line suggests a
4690 non-existent command as the problem.
4692 26. If received_header_text expands to an empty string, do not add a Received:
4693 header line to the message. (Well, it adds a token one on the spool, but
4694 marks it "old" so that it doesn't get used or transmitted.)
4696 27. Installed eximstats 1.28 (addition of -nt option).
4698 28. There was no check for failure on the call to getsockname() in the daemon
4699 code. This can fail if there is a shortage of resources on the system, with
4700 ENOMEM, for example. A temporary error is now given on failure.
4702 29. Contrary to the C standard, it seems that in some environments, the
4703 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4704 program. Exim now does this explicitly; it affects the formatting of
4705 timestamps using strftime().
4707 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4708 complaints. I've now initialized all the variables, to avoid this.
4710 32. Header lines added by a system filter were not being "seen" during
4711 transport-time rewrites.
4713 33. The info_callback() function passed to OpenSSL is set up with type void
4714 (*)(SSL *, int, int), as described somewhere. However, when calling the
4715 function (actually a macro) that sets it up, the type void(*)() is
4716 expected. I've put in a cast to prevent warnings from picky compilers.
4718 34. If a DNS black list lookup found a CNAME record, but there were no A
4719 records associated with the domain it pointed at, Exim crashed.
4721 35. If a DNS black list lookup returned more than one A record, Exim ignored
4722 all but the first. It now scans all returned addresses if a particular IP
4723 value is being sought. In this situation, the contents of the
4724 $dnslist_value variable are a list of all the addresses, separated by a
4727 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4728 to accept a host name and all its aliases if the forward lookup for any of
4729 them yielded the IP address of the incoming connection. Now it accepts only
4730 those names whose forward lookup yields the correct IP address. Any other
4731 names are discarded. This closes a loophole whereby a rogue DNS
4732 administrator could create reverse DNS records to break through a
4733 wildcarded host restriction in an ACL.
4735 37. If a user filter or a system filter that ran in a subprocess used any of
4736 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4737 the wrong values were passed to the pipe command ($thisaddress had the
4738 value of $0, $0 had the value of $1, etc). This bug was introduced by
4739 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4741 38. Improved the line breaking for long SMTP error messages from ACLs.
4742 Previously, if there was no break point between 40 and 75 characters, Exim
4743 left the rest of the message alone. Two changes have been made: (a) I've
4744 reduced the minimum length to 35 characters; (b) if it can't find a break
4745 point between 35 and 75 characters, it looks ahead and uses the first one
4746 that it finds. This may give the occasional overlong line, but at least the
4747 remaining text gets split now.
4749 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4750 file descriptors might be low, and that setting 1000 would always raise it.
4751 It turns out that in some environments, the limit is already over 1000 and
4752 that lowering it causes trouble. So now Exim takes care not to decrease it.
4754 40. When delivering a message, the value of $return_path is set to $sender_
4755 address at the start of routing (routers may change the value). By an
4756 oversight, this default was not being set up when an address was tested by
4757 -bt or -bv, which affected the outcome if any router or filter referred to
4760 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4761 log only when "message" or "log_message" are set. However, if one of the
4762 conditions was an address verification, or a call to a nested ACL, the
4763 messages generated by the underlying test were being passed through. This
4764 no longer happens. The underlying message is available in $acl_verify_
4765 message for both "message" and "log_message" expansions, so it can be
4766 passed through if needed.
4768 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4769 new expansion $bh_ to give the encoded byte string without charset
4770 translation. Translation happens only if iconv() is available; HAVE_ICONV
4771 indicates this at build time. HEADERS_CHARSET gives the charset to
4772 translate to; headers_charset can change it in the configuration, and
4773 "headers charset" can change it in an individual filter file.
4775 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4776 that creates RFC 2047 encoded "words" labels them as that charset instead
4777 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4778 expansion operator; (ii) when Exim creates a From: line for a local
4779 message; (iii) when a header line is rewritten to include a "phrase" part.
4781 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4782 buggy, causing it to skip the first lines of messages whose message ID
4783 ended in 'D'. This would not have bitten before Exim release 4.14, because
4784 message IDs were unlikely to end in 'D' before then. The effect was to have
4785 incorrect size information for certain domains.
4787 45. #include "config.h" was missing at the start of the crypt16.c module. This
4788 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4791 46. If there was a timeout during a "random" callout check, Exim treated it as
4792 a failure of the random address, and carried on sending RSET and the real
4793 address. If the delay was just some slowness somewhere, the response to the
4794 original RCPT would be taken as a response to RSET and so on, causing
4795 mayhem of various kinds.
4797 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4798 when I implemented it. It didn't allow for the fact that some option values
4799 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4800 the right test for positive values.
4802 48. Domain names in DNS records are case-independent. Exim always looks them up
4803 in lower case. Some resolvers return domain names in exactly the case they
4804 appear in the zone file, that is, they may contain uppercase letters. Not
4805 all resolvers do this - some return always lower case. Exim was treating a
4806 change of case by a resolver as a change of domain, similar to a widening
4807 of a domain abbreviation. This triggered its re-routing code and so it was
4808 trying to route what was effectively the same domain again. This normally
4809 caused routing to fail (because the router wouldn't handle the domain
4810 twice). Now Exim checks for this case specially, and just changes the
4811 casing of the domain that it ultimately uses when it transmits the message
4814 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4817 50. If a filter generated a file delivery with a non-absolute name (possible if
4818 no home directory exists for the router), the forbid_file option was not
4821 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4822 the existing equality matching.
4824 52. Exim was using ints instead of ino_t variables in some places where it was
4825 dealing with inode numbers.
4827 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4828 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4829 called TMPDIR, and if it finds it is different, it changes its value.
4831 54. The smtp_printf() function is now made available to local_scan() so
4832 additional output lines can be written before returning. There is also an
4833 smtp_fflush() function to enable the detection of a dropped connection.
4834 The variables smtp_input and smtp_batched_input are exported to
4837 55. Changed the default runtime configuration: the message "Unknown user"
4838 has been removed from the ACL, and instead placed on the localuser router,
4839 using the cannot_route_message feature. This means that any verification
4840 failures that generate their own messages won't get overridden. Similarly,
4841 the "Unrouteable address" message that was in the ACL for unverifiable
4842 relay addresses has also been removed.
4844 56. Added hosts_avoid_esmtp to the smtp transport.
4846 57. The exicyclog script was not checking for the esoteric option
4847 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4848 will work only if exicyclog is run under the appropriate euid.
4850 58. Following a discussion on the list, the rules by which Exim recognises line
4851 endings on incoming messages have been changed. The -dropcr and drop_cr
4852 options are now no-ops, retained only for backwards compatibility. The
4853 following line terminators are recognized: LF CRLF CR. However, special
4854 processing applies to CR:
4856 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4857 nor a local message in the state where . is a terminator.
4859 (ii) If a bare CR is encountered in a header line, an extra space is added
4860 after the line terminator so as not to end the header. The reasoning
4861 behind this is that bare CRs in header lines are most likely either
4862 to be mistakes, or people trying to play silly games.
4864 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4865 was being incorrectly given as 18 bytes larger than it should have been.
4866 This is a VOB (very old bug).
4868 60. This may never have affected anything current, but just in case it has:
4869 When the local host is found other than at the start of a list of hosts,
4870 the local host, those with the same MX, and any that follow, are discarded.
4871 When the list in question was part of a longer list of hosts, the following
4872 hosts (not currently being processed) were also being discarded. This no
4873 longer happens. I'm not sure if this situation could ever has previously
4876 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4879 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4880 header. This is something that is recommended in a new Internet Draft, and
4881 is something that is documented as being done by Sendmail. There are two
4882 possible values. For messages generated by the autoreply transport, Exim
4885 Auto-Submitted: auto-replied
4887 whereas for all other generated messages (e.g. bounces) it adds
4889 Auto-Submitted: auto-generated
4891 63. The "personal" condition in filters now includes a test for the
4892 Auto-Submitted: header. If it contains the string "auto-" the message it
4893 not considered personal.
4895 64. Added rcpt_include_affixes as a generic transport option.
4897 65. Added queue_only_override (default true).
4899 66. Added the syslog_duplication option.
4901 67. If what should have been the first header line of a message consisted of
4902 a space followed by a colon, Exim was mis-interpreting it as a header line.
4903 It isn't of course - it is syntactically invalid and should therefore be
4904 treated as the start of the message body. The misbehaviour could have
4905 caused a number of strange effects, including loss of data in subsequent
4906 header lines, and spool format errors.
4908 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4909 client host had authenticated. This control can now be exercised by an ACL
4910 for more flexibility.
4912 69. By default, callouts do not happen when testing with -bh. There is now a
4913 variant, -bhc, which does actually run the callout code, including
4914 consulting and updating the callout cache.
4916 70. Added support for saslauthd authentication, courtesy of Alexander
4919 71. If statvfs() failed on the spool or log directories while checking their
4920 size for availability, Exim confusingly gave the error "space shortage".
4921 Furthermore, in debugging mode it crashed with a floating point exception.
4922 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4923 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4924 serious problem, Exim now writes to the main and panic logs when this
4925 happens, with details of the failure. It then refuses to accept the
4926 incoming message, giving the message "spool directory problem" or "log
4927 directory problem" with a 421 code for SMTP messages.
4929 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4930 0, 1, and 2 exist, because some OS complain for execs without them (see
4931 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4932 descriptors. However, the code omitted to check that the open succeeded,
4933 causing mysterious errors if for some reason the permissions on /dev/null
4934 got screwed. Now Exim writes a message to the main and panic logs, and
4935 bombs out if it can't open /dev/null.
4937 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4938 interact so that it is all more flexible. It is supposed to remain
4939 backwards compatible. Also added extra_local_interfaces.
4941 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4942 to bomb out with an assertion failure - to the client this appears as a
4943 connection drop. This problem occurs in the part of the code that was taken
4944 from the Samba project. Fortunately, the assertion is in a very simple
4945 function, so I have fixed this by reproducing the function inline in the
4946 one place where it is called, and arranging for authentication to fail
4947 instead of killing the process with assert().
4949 75. The SPA client code was not working when the server requested OEM rather
4950 than Unicode encoding.
4952 76. Added code to make require_files with a specific uid setting more usable in
4953 the case where statting the file as root fails - usually a non-root-mounted
4954 NFS file system. When this happens and the failure is EACCES, Exim now
4955 forks a subprocess and does the per-uid checking as the relevant uid.
4957 77. Added process_log_path.
4959 78. If log_file_path was not explicitly set, a setting of check_log_space or
4960 check_log_inodes was ignored.
4962 79. If a space check for the spool or log partitions fails, the incident is now
4963 logged. Of course, in the latter case the data may get lost...
4965 80. Added the %p formatting code to string_format() so that it can be used to
4966 print addresses in debug_print(). Adjusted all the address printing in the
4967 debugging in store.c to use %p rather than %d.
4969 81. There was a concern that a line of code in smtp_in.c could overflow a
4970 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4971 initially expressed, the concern was not well-founded, because trailing
4972 spaces are removed early. However, if the trailing spaces were followed by
4973 a NULL, they did not get removed, so the overflow was possible. Two fixes
4976 (a) I re-wrote the offending code in a cleaner fashion.
4977 (b) If an incoming SMTP command contains a NULL character, it is rejected
4980 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4981 runs initgroups(), so that if the Exim user is in any additional groups,
4982 they will be used during message reception.
4988 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob