1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.555 2008/10/12 09:58:13 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
81 TK/01 Add preliminary DKIM support. Currently requires a forked version of
82 ALT-N's libdkim that I have put here:
83 http://duncanthrax.net/exim-experimental/
85 Note to Michael Haardt: I had to rename some vars in sieve.c. They
86 were called 'true' and it seems that C99 defines that as a reserved
87 keyword to be used with 'bool' variable types. That means you could
88 not include C99-style headers which use bools without triggering
89 build errors in sieve.c.
91 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
92 as mailq or other aliases. Changed the --help handling significantly
93 to do whats expected. exim_usage() emits usage/help information.
95 SC/01 Added the -bylocaldomain option to eximstats.
97 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
99 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
101 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
107 PH/01 Another patch from the Sieve maintainer.
109 PH/02 When an IPv6 address is converted to a string for single-key lookup
110 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
111 dots are used instead of colons so that keys in lsearch files need not
112 contain colons. This was done some time before quoting was made available
113 in lsearch files. However, iplsearch files do require colons in IPv6 keys
114 (notated using the quote facility) so as to distinguish them from IPv4
115 keys. This meant that lookups for IP addresses in host lists did not work
116 for iplsearch lookups.
118 This has been fixed by arranging for IPv6 addresses to be expressed with
119 colons if the lookup type is iplsearch. This is not incompatible, because
120 previously such lookups could never work.
122 The situation is now rather anomolous, since one *can* have colons in
123 ordinary lsearch keys. However, making the change in all cases is
124 incompatible and would probably break a number of configurations.
126 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
129 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
130 conversion specification without a maximum field width, thereby enabling
131 a rogue spamd server to cause a buffer overflow. While nobody in their
132 right mind would setup Exim to query an untrusted spamd server, an
133 attacker that gains access to a server running spamd could potentially
134 exploit this vulnerability to run arbitrary code as the Exim user.
136 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
137 $primary_hostname instead of what libspf2 thinks the hosts name is.
139 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
140 a directory entry by the name of the lookup key. Previously, if a
141 symlink pointed to a non-existing file or a file in a directory that
142 Exim lacked permissions to read, a lookup for a key matching that
143 symlink would fail. Now it is enough that a matching directory entry
144 exists, symlink or not. (Bugzilla 503.)
146 PH/03 The body_linecount and body_zerocount variables are now exported in the
149 PH/04 Added the $dnslist_matched variable.
151 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
152 This means they are set thereafter only if the connection becomes
155 PH/06 Added the client_condition to authenticators so that some can be skipped
156 by clients under certain conditions.
158 PH/07 The error message for a badly-placed control=no_multiline_responses left
159 "_responses" off the end of the name.
161 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
163 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
164 (without spaces) instead of just copying the configuration text.
166 PH/10 Added the /noupdate option to the ratelimit ACL condition.
168 PH/11 Added $max_received_linelength.
170 PH/12 Added +ignore_defer and +include_defer to host lists.
172 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
173 way in which PCRE > 7.0 is built.
175 PH/14 Implemented queue_only_load_latch.
177 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
178 MAIL command. The effect was to mangle the value on 64-bit systems.
180 PH/16 Another patch from the Sieve maintainer.
182 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
184 PH/18 If a system quota error occurred while trying to create the file for
185 a maildir delivery, the message "Mailbox is full" was not appended to the
186 bounce if the delivery eventually timed out. Change 4.67/27 below applied
187 only to a quota excession during the actual writing of the file.
189 PH/19 It seems that peer DN values may contain newlines (and other non-printing
190 characters?) which causes problems in log lines. The DN values are now
191 passed through string_printing() before being added to log lines.
193 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
194 and InterBase are left for another time.)
196 PH/21 Added message_body_newlines option.
198 PH/22 Guard against possible overflow in moan_check_errorcopy().
200 PH/23 POSIX allows open() to be a macro; guard against that.
202 PH/24 If the recipient of an error message contained an @ in the local part
203 (suitably quoted, of course), incorrect values were put in $domain and
204 $local_part during the evaluation of errors_copy.
210 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
211 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
214 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
215 issue a MAIL command.
217 PH/02 In an ACL statement such as
219 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
221 if a client was not listed at all, or was listed with a value other than
222 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
223 the condition was not true (as it should be), so access was not denied.
224 The bug was that the ! inversion was incorrectly passed on to the second
225 item. This has been fixed.
227 PH/03 Added additional dnslists conditions == and =& which are different from
228 = and & when the dns lookup returns more than one IP address.
230 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
231 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
233 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
234 FSYNC, which compiles an option called disable_fsync that allows for
235 bypassing fsync(). The documentation is heavily laced with warnings.
237 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
239 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
240 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
241 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
242 including adding "make clean"; (3) Added -fPIC when compiling the test
243 dynamically loaded module, to get rid of a warning.
245 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
246 message fails, move_frozen_messages = true and ignore_bounce_errors_after
247 = 0s. The bug is otherwise harmless.
249 PH/07 There was a bug in the dovecot authenticator such that the value of
250 $auth1 could be overwritten, and so not correctly preserved, after a
251 successful authentication. This usually meant that the value preserved by
252 the server_setid option was incorrect.
254 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
256 PH/09 Installed PCRE release 7.0.
258 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
259 run for batched SMTP input. It is now run at the start of every message
260 in the batch. While fixing this I discovered that the process information
261 (output by running exiwhat) was not always getting set for -bs and -bS
262 input. This is fixed, and it now also says "batched" for BSMTP.
264 PH/11 Added control=no_pipelining.
266 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
267 patch, slightly modified), and move the expansion of helo_data till after
268 the connection is made in the smtp transport (so it can use these
271 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
273 PH/14 Added log_selector = +pid.
275 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
277 PH/16 Add ${if forany and ${if forall.
279 PH/17 Added dsn_from option to vary the From: line in DSNs.
281 PH/18 Flush SMTP output before performing a callout, unless control =
282 no_callout_flush is set.
284 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
285 was true (the default) a successful delivery failed to delete the retry
286 item, thus causing premature timeout of the address. The bug is now
289 PH/20 Added hosts_avoid_pipelining to the smtp transport.
291 PH/21 Long custom messages for fakedefer and fakereject are now split up
292 into multiline reponses in the same way that messages for "deny" and
293 other ACL rejections are.
295 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
296 with slight modification.
298 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
299 draft, changing the syntax and factoring some duplicate code".
301 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
302 for deliveries of the second and subsequent messages over the same SMTP
305 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
306 ${reduce, with only minor "tidies".
308 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
310 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
311 expansion side effects.
313 PH/27 When a message times out after an over-quota error from an Exim-imposed
314 quota, the bounce message says "mailbox is full". This message was not
315 being given when it was a system quota that was exceeded. It now should
318 MH/03 Made $recipients available in local_scan(). local_scan() already has
319 better access to the recipient list through recipients_list[], but
320 $recipients can be useful in postmaster-provided expansion strings.
322 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
323 in the case of a MAIL command with additional options following the
324 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
325 were accidentally chopped off.
327 PH/29 SMTP synchronization checks are implemented when a command is read -
328 there is a check that no more input is waiting when there shouldn't be
329 any. However, for some commands, a delay in an ACL can mean that it is
330 some time before the response is written. In this time, more input might
331 arrive, invalidly. So now there are extra checks after an ACL has run for
332 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
333 pipelining has not been advertised.
335 PH/30 MH's patch to allow iscntrl() characters to be list separators.
337 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
338 returned in the SMTP response when smtp_return_error_details was false.
341 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
342 instead of the C I/O that was originally supplied, because problems were
345 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
346 Exim which did not show up earlier: it was assuming that a call to
347 SSL_CTX_set_info_callback() might give an error value. In fact, there is
348 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
349 was a macro that became an assignment, so it seemed to work. This has
350 changed to a proper function call with a void return, hence the compile
351 error. Exim's code has been fixed.
353 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
356 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
358 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
360 PH/37 If a message is not accepted after it has had an id assigned (e.g.
361 because it turns out to be too big or there is a timeout) there is no
362 "Completed" line in the log. When some messages of this type were
363 selected by exigrep, they were listed as "not completed". Others were
364 picked up by some special patterns. I have improved the selection
365 criteria to be more general.
367 PH/38 The host_find_failed option in the manualroute router can now be set
368 to "ignore", to completely ignore a host whose IP address cannot be
369 found. If all hosts are ignored, the behaviour is controlled by the new
370 host_all_ignored option.
372 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
373 homing or because of multiple MX records with /mx) generated more than
374 one IP address, and the following item turned out to be the local host,
375 all the secondary addresses of the first item were incorrectly removed
376 from the list, along with the local host and any following hosts (which
377 is what is supposed to happen).
379 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
380 whoever called Exim into the -H file. In the case of the daemon it was
381 behaving confusingly. When first started, it used values for whoever
382 started the daemon, but after a SIGHUP it used the Exim user (because it
383 calls itself on a restart). I have changed the code so that it now always
386 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
387 message are rejected with the same error (e.g. no authentication or bad
388 sender address), and a DATA command is nevertheless sent (as can happen
389 with PIPELINING or a stupid MUA), the error message that was given to the
390 RCPT commands is included in the rejection of the DATA command. This is
391 intended to be helpful for MUAs that show only the final error to their
394 PH/42 Another patch from the Sieve maintainer.
396 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
397 Eximstats - Fixed some broken HTML links and added missing column headers
399 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
400 columns for Rejects, Temp Rejects, Ham, and Spam rows.
402 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
404 PH/43 Yet another patch from the Sieve maintainer.
406 PH/44 I found a way to check for a TCP/IP connection going away before sending
407 the response to the final '.' that terminates a message, but only in the
408 case where the client has not sent further data following the '.'
409 (unfortunately, this is allowed). However, in many cases there won't be
410 any further data because there won't be any more messages to send. A call
411 to select() can be used: if it shows that the input is "ready", there is
412 either input waiting, or the socket has been closed. An attempt to read
413 the next input character can distinguish the two cases. Previously, Exim
414 would have sent an OK response which the client would never have see.
415 This could lead to message repetition. This fix should cure that, at
416 least in a lot of common cases.
418 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
419 advertised in response to EHLO.
425 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
426 fixed by 4.65/MH/01 (is this a record?) are fixed:
428 (i) An empty string was always treated as zero by the numeric comparison
429 operators. This behaviour has been restored.
431 (ii) It is documented that the numeric comparison operators always treat
432 their arguments as decimal numbers. This was broken in that numbers
433 starting with 0 were being interpreted as octal.
435 While fixing these problems I realized that there was another issue that
436 hadn't been noticed. Values of message_size_limit (both the global option
437 and the transport option) were treated as octal if they started with 0.
438 The documentation was vague. These values are now always treated as
439 decimal, and I will make that clear in the documentation.
445 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
446 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
449 MH/01 Don't check that the operands of numeric comparison operators are
450 integers when their expansion is in "skipping" mode (fixes bug
451 introduced by 4.64-PH/07).
453 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
454 child addresses, Exim now panics and dies. Previously, because the count
455 is held in a short int, deliveries were likely to be lost. As such a
456 large number of recipients for a single message is ridiculous
457 (performance will be very, very poor), I have chosen to impose a limit
458 rather than extend the field.
464 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
465 leftover -K file (the existence of which was triggered by #402).
466 While we were at it, introduced process PID as part of the -K
467 filename. This should rule out race conditions when creating
470 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
471 processing considerably. Previous code took too long for large mails,
472 triggering a timeout which in turn triggers #401.
474 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
475 in the DK code in transports.c. sendfile() is not really portable,
476 hence the _LINUX specificness.
478 TF/01 In the add_headers option to the mail command in an Exim filter,
479 there was a bug that Exim would claim a syntax error in any
480 header after the first one which had an odd number of characters
483 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
484 callout verification, Exim cached a "reject" for the entire domain. This
485 is correct for most verifications, but it is not correct for a recipient
486 verification with use_sender or use_postmaster set, because in that case
487 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
488 case of MAIL FROM:<> rejection from other early rejections (e.g.
489 rejection of HELO). When verifying a recipient using a non-null MAIL
490 address, the cache is ignored if it shows MAIL FROM:<> rejection.
491 Whatever the result of the callout, the value of the domain cache is
492 left unchanged (for any other kind of callout, getting as far as trying
493 RCPT means that the domain itself is ok).
495 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
498 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
499 manifest itself as EPIPE rather than ECONNECT. When tidying away a
500 session, the daemon ignores ECONNECT errors and logs others; it now
501 ignores EPIPE as well.
503 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
504 (quoted-printable decoding).
506 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
507 later the small subsequent patch to fix an introduced bug.
509 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
511 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
513 PH/08 An error is now given if message_size_limit is specified negative.
515 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
516 to be given (somewhat) arbitrary names.
518 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
521 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
522 miscellaneous code fixes
524 PH/10 Added the log_reject_target ACL modifier to specify where to log
527 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
528 hostname. This is wrong, because it relates to the incoming message (and
529 probably the interface on which it is arriving) and not to the outgoing
530 callout (which could be using a different interface). This has been
531 changed to use the value of the helo_data option from the smtp transport
532 instead - this is what is used when a message is actually being sent. If
533 there is no remote transport (possible with a router that sets up host
534 addresses), $smtp_active_hostname is used.
536 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
537 tweaks were necessary in order to get it to work (see also 21 below):
538 (a) The code assumed that strncpy() returns a negative number on buffer
539 overflow, which isn't the case. Replaced with Exim's string_format()
541 (b) There were several signed/unsigned issues. I just did the minimum
542 hacking in of casts. There is scope for a larger refactoring.
543 (c) The code used strcasecmp() which is not a standard C function.
544 Replaced with Exim's strcmpic() function.
545 (d) The code set only $1; it now sets $auth1 as well.
546 (e) A simple test gave the error "authentication client didn't specify
547 service in request". It would seem that Dovecot has changed its
548 interface. Fortunately there's a specification; I followed it and
549 changed what the client sends and it appears to be working now.
551 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
554 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
555 address A is aliased to B and C, where B exists and C does not. Without
556 -v the output is "A verified" because verification stops after a
557 successful redirection if more than one address is generated. However,
558 with -v the child addresses are also verified. Exim was outputting "A
559 failed to verify" and then showing the successful verification for C,
560 with its parentage. It now outputs "B failed to verify", showing B's
561 parentage before showing the successful verification of C.
563 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
564 look up a TXT record in a specific list after matching in a combined
567 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
568 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
569 they consult the DNS. I had assumed they would set it the way they
570 wanted; and indeed my experiments on Linux seem to show that in some
571 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
572 To be on the safe side, however, I have now made the interface to
573 host_find_byname() similar to host_find_bydns(), with an argument
574 containing the DNS resolver options. The host_find_byname() function now
575 sets these options at its start, just as host_find_bydns() does. The smtp
576 transport options dns_qualify_single and dns_search_parents are passed to
577 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
578 of host_find_byname() use the default settings of RES_DEFNAMES
579 (qualify_single) but not RES_DNSRCH (search_parents).
581 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
582 spool_read_header() do less string testing, by means of a preliminary
583 switch on the second character of optional "-foo" lines. (This is
584 overdue, caused by the large number of possibilities that now exist.
585 Originally there were few.) While I was there, I also converted the
586 str(n)cmp tests so they don't re-test the leading "-" and the first
587 character, in the hope this might squeeze out yet more improvement.
589 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
590 flag allowing group syntax was set by the header_syntax check but not
591 turned off, possible causing trouble later; (2) The flag was not being
592 set at all for the header_verify test, causing "group"-style headers to
593 be rejected. I have now set it in this case, and also caused header_
594 verify to ignore an empty address taken from a group. While doing this, I
595 came across some other cases where the code for allowing group syntax
596 while scanning a header line wasn't quite right (mostly, not resetting
597 the flag correctly in the right place). These bugs could have caused
598 trouble for malformed header lines. I hope it is now all correct.
600 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
601 with the "reply" argument non-NULL. The code, however (which originally
602 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
603 but it didn't always do it. This confused somebody who was copying the
604 code for some other use. I have removed all the tests.
606 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
607 feature that was used to support insecure browsers during the U.S. crypto
608 embargo. It requires special client support, and Exim is probably the
609 only MTA that supported it -- and would never use it because real RSA is
610 always available. This code has been removed, because it had the bad
611 effect of slowing Exim down by computing (never used) parameters for the
612 RSA_EXPORT functionality.
614 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
615 authenticator to fail if there's a tab character in the incoming data
616 (there should never be unless someone is messing about, as it's supposed
617 to be base64-encoded). Also added, on Timo's advice, the "secured" option
618 if the connection is using TLS or if the remote IP is the same as the
619 local IP, and the "valid-client-cert option" if a client certificate has
622 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
623 authenticators. This can be used for authorization after authentication
624 succeeds. (In the case of plaintext, it servers for both authentication
627 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
628 if any retry times were supplied.
630 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
631 connection, where there is no client IP address to check. In this
632 situation, the verify now always succeeds.
634 PH/25 Applied John Jetmore's -Mset patch.
636 PH/26 Added -bem to be like -Mset, but loading a message from a file.
638 PH/27 In a string expansion for a processed (not raw) header when multiple
639 headers of the same name were present, leading whitespace was being
640 removed from all of them, but trailing whitespace was being removed only
641 from the last one. Now trailing whitespace is removed from each header
642 before concatenation. Completely empty headers in a concatenation (as
645 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
646 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
648 PH/29 [Removed. This was a change that I later backed out, and forgot to
649 correct the ChangeLog entry (that I had efficiently created) before
650 committing the later change.]
652 PH/30 Exim was sometimes attempting to deliver messages that had suffered
653 address errors (4xx response to RCPT) over the same connection as other
654 messages routed to the same hosts. Such deliveries are always "forced",
655 so retry times are not inspected. This resulted in far too many retries
656 for the affected addresses. The effect occurred only when there were more
657 hosts than the hosts_max_try setting in the smtp transport when it had
658 the 4xx errors. Those hosts that it had tried were not added to the list
659 of hosts for which the message was waiting, so if all were tried, there
660 was no problem. Two fixes have been applied:
662 (i) If there are any address or message errors in an SMTP delivery, none
663 of the hosts (tried or untried) are now added to the list of hosts
664 for which the message is waiting, so the message should not be a
665 candidate for sending over the same connection that was used for a
666 successful delivery of some other message. This seems entirely
667 reasonable: after all the message is NOT "waiting for some host".
668 This is so "obvious" that I'm not sure why it wasn't done
669 previously. Hope I haven't missed anything, but it can't do any
670 harm, as the worst effect is to miss an optimization.
672 (ii) If, despite (i), such a delivery is accidentally attempted, the
673 routing retry time is respected, so at least it doesn't keep
674 hammering the server.
676 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
677 in ${readsocket because some servers need this prod.
679 PH/32 Added some extra debug output when updating a wait-xxx database.
681 PH/33 The hint "could be header name not terminated by colon", which has been
682 given for certain expansion errors for a long time, was not being given
683 for the ${if def:h_colon_omitted{... case.
685 PH/34 The spec says: "With one important exception, whenever a domain list is
686 being scanned, $domain contains the subject domain." There was at least
687 one case where this was not true.
689 PH/35 The error "getsockname() failed: connection reset by peer" was being
690 written to the panic log as well as the main log, but it isn't really
691 panic-worthy as it just means the connection died rather early on. I have
692 removed the panic log writing for the ECONNRESET error when getsockname()
695 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
696 runs only) independently of the message's sender address. This meant
697 that, if the 4xx error was in fact related to the sender, a different
698 message to the same recipient with a different sender could confuse
699 things. In particualar, this can happen when sending to a greylisting
700 server, but other circumstances could also provoke similar problems.
701 I have changed the default so that the retry time for these errors is now
702 based a combination of the sender and recipient addresses. This change
703 can be overridden by setting address_retry_include_sender=false in the
706 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
707 remote server are returned as part of bounce messages. This was not
708 happening for LMTP over a pipe (the lmtp transport), but now it is the
709 same for both kinds of LMTP.
711 PH/38 Despite being documented as not happening, Exim was rewriting addresses
712 in header lines that were in fact CNAMEs. This is no longer the case.
714 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
715 and queue runs started by the daemon processed all messages. This has
716 been fixed so that -R and -S can now usefully be given with -q<time>.
718 PH/40 Import PCRE release 6.7 (fixes some bugs).
720 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
722 PH/42 Give an error if -q is specified more than once.
724 PH/43 Renamed the variables $interface_address and $interface_port as
725 $received_ip_address and $received_port, to make it clear that these
726 values apply to message reception, and not to the outgoing interface when
727 a message is delivered. (The old names remain recognized, of course.)
729 PH/44 There was no timeout on the connect() call when using a Unix domain
730 socket in the ${readsocket expansion. There now is.
732 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
733 be meaningful with "accept".
735 SC/01 Eximstats V1.43
736 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
738 SC/02 Eximstats V1.44
739 Use a glob alias rather than an array ref in the generated
740 parser. This improves both readability and performance.
742 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
743 Collect SpamAssassin and rejection statistics.
744 Don't display local sender or destination tables unless
745 there is data to show.
746 Added average volumes into the top table text output.
748 SC/04 Eximstats V1.46
749 Collect data on the number of addresses (recipients)
750 as well as the number of messages.
752 SC/05 Eximstats V1.47
753 Added 'Message too big' to the list of mail rejection
754 reasons (thanks to Marco Gaiarin).
756 SC/06 Eximstats V1.48
757 Mainlog lines which have GMT offsets and are too short to
758 have a flag are now skipped.
760 SC/07 Eximstats V1.49 (Alain Williams)
761 Added the -emptyok flag.
763 SC/08 Eximstats V1.50
764 Fixes for obtaining the IP address from reject messages.
766 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
767 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
768 whitesspace changes from 4.64-PH/27
770 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
773 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
774 are found, allow negative numbers in numeric criteria)
776 JJ/06 exipick.20061117.2, added new $message_body_missing variable
778 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
779 to match changes made in 4.64-PH/43
781 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
783 PH/47 Put in an explicit test for a DNS lookup of an address record where the
784 "domain" is actually an IP address, and force a failure. This locks out
785 those revolvers/nameservers that support "A-for-A" lookups, in
786 contravention of the specifications.
788 PH/48 When a host name was looked up from an IP address, and the subsequent
789 forward lookup of the name timed out, the host name was left in
790 $sender_host_name, contrary to the specification.
792 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
793 restricted to single-key lookups, Exim was not diagnosing an error if
794 * or *@ was used with a query-style lookup.
796 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
798 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
799 long ago, but noone interested enough thought of it. Let's just say that
800 the "1.1" means that there are some new functions that weren't there at
801 some point in the past.
803 PH/51 Error processing for expansion failure of helo_data from an smtp
804 transport during callout processing was broken.
806 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
807 tested/used via the -bh/-bhc/-bs options.
809 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
810 bug, fixed in subsequent PCRE releases).
812 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
813 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
815 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
821 SC/01 Use a glob alias rather than an array ref in eximstats generated
822 parser. This improves both readability and performance.
824 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
825 Don't display local sender or destination tables in eximstats unless
826 there is data to show.
827 Added average volumes into the eximstats top table text output.
829 SC/03 Collect data on the number of addresses (recipients) as well
830 as the number of messages in eximstats.
832 TF/01 Correct an error in the documentation for the redirect router. Exim
833 does (usually) call initgroups() when daemonizing.
835 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
836 with consistent privilege compared to when running as a daemon.
838 TF/03 Note in the spec that $authenticated_id is not set for local
839 submissions from trusted users.
841 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
842 Thanks to Dean Brooks <dean@iglou.com> for the patch.
844 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
845 by adding some example configuration directives to the default
846 configuration file. A little bit of work is required to uncomment the
847 directives and define how usernames and passwords are checked, but
848 there is now a framework to start from.
850 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
851 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
852 without this. I don't know how relevant this is to other LDAP libraries.
854 PH/02 Add the verb name to the "unknown ACL verb" error.
856 PH/03 Magnus Holmgren's patch for filter_prepend_home.
858 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
860 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
861 directory not expanded when it should be if an expanded home directory
862 was set for the address (which is overridden by the transport).
864 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
867 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
868 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
869 because it is too late at that time, and has no effect.
871 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
872 security issue with \' (bugzilla #107). I could not use the
873 PQescapeStringConn() function, because it needs a PGconn value as one of
876 PH/08 When testing addresses using -bt, indicate those final addresses that
877 are duplicates that would not cause an additional delivery. At least one
878 person was confused, thinking that -bt output corresponded to deliveries.
879 (Suppressing duplicates isn't a good idea as you lose the information
880 about possibly different redirections that led to the duplicates.)
882 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
883 systems where poll() doesn't work, in particular OS X.
885 PH/10 Added more information to debugging output for retry time not reached.
887 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
888 operations in malware.c.
890 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
893 PH/13 If write_rejectlog was set false when logging was sent to syslog with
894 syslog_duplication set false, log lines that would normally be written
895 both the the main log and to the reject log were not written to syslog at
898 PH/14 In the default configuration, change the use of "message" in ACL warn
899 statements to "add_header".
901 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
902 not followed by a command (e.g. "seen endif").
904 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
905 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
908 PH/17 Added extra conditions to the default value of delay_warning_condition
912 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
913 { match{$h_precedence:}{(?i)bulk|list|junk} } \
914 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
917 The Auto-Submitted: and various List- headers are standardised, whereas I
918 don't think Precedence: ever was.
920 PH/18 Refactored debugging code in route_finduser() to show more information,
921 in particular, the error code if getpwnam() issues one.
923 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
924 This is apparently needed in addition to the PH/07 change above to avoid
925 any possible encoding problems.
927 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
928 but not after initializing Perl.
930 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
931 output them only if debugging. By default they are written stderr,
932 apparently, which is not desirable.
934 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
937 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
940 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
942 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
943 authenticated or an ident call has been made. Suppress the default
944 values for $authenticated_id and $authenticated_sender (but permit -oMai
945 and -oMas) when testing with -bh.
947 PH/24 Re-jigged the order of the tests in the default configuration so that the
948 tests for valid domains and recipients precede the DNS black list and CSA
949 tests, on the grounds that those ones are more expensive.
951 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
952 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
953 This bug exists in every version of Exim that I still have, right back to
956 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
957 However, an attempt to turn on case-sensitivity in a regex key by
958 including (?-i) didn't work because the subject string was already
959 lowercased, and the effects were non-intuitive. It turns out that a
960 one-line patch can be used to allow (?-i) to work as expected.
966 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
967 other effects) broke the use of negated acl sub-conditions.
969 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
972 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
973 "Deny" causes Exim to reject the incoming connection with a 554 error.
974 Unfortunately, if there is a major crisis, such as a disk failure,
975 tcp-wrappers gives "deny", whereas what one would like would be some
976 kind of temporary error. A kludge has been added to help with this.
977 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
978 554 error is used if errno is still zero or contains ENOENT (which occurs
979 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
982 PH/03 Add -lutil to the default FreeBSD LIBS setting.
984 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
985 errors. Otherwise a message that provokes a temporary error (when other
986 messages do not) can cause a whole host to time out.
988 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
989 addresses were routed directly to files or pipes from a redirect router.
990 File deliveries just didn't batch; pipe deliveries might have suffered
993 PH/06 A failure to get a lock for a hints database would erroneously always say
994 "Failed to get write lock", even when it was really a read lock.
996 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
997 of 0600. This has been changed to use the value of the lockfile_mode
998 option (which defaults to 0600).
1000 PH/08 Applied small patch from the Sieve maintainer.
1002 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1003 folder from quota calculations, a direct delivery into this folder messed
1004 up the contents of the maildirsize file. This was because the regex was
1005 used only to exclude .Trash (or whatever) when the size of the mailbox
1006 was calculated. There was no check that a delivery was happening into an
1007 excluded directory. This bug has been fixed by ignoring all quota
1008 processing for deliveries into excluded directories.
1010 PH/10 Added the maildirfolder_create_regex option to appendfile.
1016 PH/01 The code for finding all the local interface addresses on a FreeBSD
1017 system running IPv6 was broken. This may well have applied to all BSD
1018 systems, as well as to others that have similar system calls. The broken
1019 code found IPv4 interfaces correctly, but gave incorrect values for the
1020 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1021 that it would not match correctly against @[] and not recognize the IPv6
1024 PH/02 The ipliteral router was not recognizing addresses of the form user@
1025 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1027 PH/03 Added disable_ipv6.
1029 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1030 decoded form, because it is most often used to construct To: headers
1031 lines in autoreplies, and the decoded form may well be syntactically
1032 invalid. However, $reply_address has leading white space removed, and all
1033 newlines turned into spaces so that the autoreply transport does not
1036 PH/05 If group was specified without a user on a router, and no group or user
1037 was specified on a transport, the group from the router was ignored.
1039 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1040 for visible compile-time settings that can be used to change these
1041 numbers, for those that want even more. Backwards compatibility with old
1042 spool files has been maintained. However, going back to a previous Exim
1043 release will lost any variables that are in spool files.
1045 PH/07 Two small changes when running in the test harness: increase delay when
1046 passing a TCP/IP connection to a new process, in case the original
1047 process has to generate a bounce, and remove special handling of
1048 127.0.0.2 (sic), which is no longer necessary.
1050 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1051 be the same on different OS.
1053 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1056 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1057 whether --show-vars was specified or not
1059 JJ/02 exipick: Added support for new ACL variable spool format introduced
1062 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1063 syntactically invalid From: or Reply-to: line, and a filter used this to
1064 generate an autoreply, and therefore failed to obtain an address for the
1065 autoreply, Exim could try to deliver to a non-existent relative file
1066 name, causing unrelated and misleading errors. What now happens is that
1067 it logs this as a hard delivery error, but does not attempt to create a
1070 PH/11 The exinext utility has a -C option for testing purposes, but although
1071 the given file was scanned by exinext itself; it wasn't being passed on
1072 when Exim was called.
1074 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1075 an end-of-file indication when reading a command response.
1077 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1078 compiled. In many other places in Exim, IPv6 addresses are always
1079 recognized, so I have changed this. It also means that IPv4 domain
1080 literals of the form [IPV4:n.n.n.n] are now always recognized.
1082 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1083 used if the router is not running as root, for example, when verifying at
1084 ACL time, or when using -bh. The debugging output from this situation was
1085 non-existent - all you got was a failure to exec. I have made two
1088 (a) Failures to set uid/gid, the current directory, or a process leader
1089 in a subprocess such as that created by queryprogram now generate
1090 suitable debugging ouput when -d is set.
1092 (b) The queryprogram router detects when it is not running as root,
1093 outputs suitable debugging information if -d is set, and then runs
1094 the subprocess without attempting to change uid/gid.
1096 PH/15 Minor change to Makefile for building test_host (undocumented testing
1099 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1100 additional section of a DNS packet that returns MX or SRV records.
1101 Instead, it always explicitly searches for A/AAAA records. This avoids
1102 major problems that occur when a DNS server includes only records of one
1103 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1104 fixed another bug: if SRV records were looked up and the corresponding
1105 address records were *not* found in the additional section, the port
1106 values from the SRV records were lost.
1108 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1109 using the correct key (the original address) when searching the retry
1110 rules in order to find which one to use for generating the retry hint.
1112 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1113 adding the default one. Similarly, if it contains a Reply-To: header, the
1114 errors_reply_to option, if set, is not used.
1116 PH/19 When calculating a retry time, Exim used to measure the "time since
1117 failure" by looking at the "first failed" field in the retry record. Now
1118 it does not use this if it is later than than the arrival time of the
1119 message. Instead it uses the arrival time. This makes for better
1120 behaviour in cases where some deliveries succeed, thus re-setting the
1121 "first failed" field. An example is a quota failure for a huge message
1122 when small messages continue to be delivered. Without this change, the
1123 "time since failure" will always be short, possible causing more frequent
1124 delivery attempts for the huge message than are intended.
1125 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1127 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1128 $1, $2, $3) because the numerical variables can be reset during some
1129 expansion items (e.g. "match"), thereby losing the authentication data.
1131 PH/21 Make -bV show the size of off_t variables so that the test suite can
1132 decide whether to run tests for quotas > 2G.
1134 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1135 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1136 filecount value is greater than 2G or if a quota value is greater than 2G
1137 on a system where the size of off_t is not greater than 4, a panic error
1140 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1141 never match. The debug and -bh output now contains an explicit error
1142 message indicating a malformed IPv4 address or mask.
1144 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1145 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1146 PH/23 above applies.
1148 PH/25 Do not write to syslog when running in the test harness. The only
1149 occasion when this arises is a failure to open the main or panic logs
1150 (for which there is an explicit test).
1152 PH/26 Added the /no_tell option to "control=freeze".
1154 PH/27 If a host name lookup failed very early in a connection, for example, if
1155 the IP address matched host_lookup and the reverse lookup yielded a name
1156 that did not have a forward lookup, an error message of the form "no IP
1157 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1158 could be logged. Now it outputs the IP address instead of "NULL".
1160 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1161 allows the sender and the authenticated sender to be set when
1162 submitting a message from within Exim. Since child_open_exim() is
1163 documented for local_scan(), the new function should be too.
1165 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1166 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1167 results in an empty string is now treated as unset.
1169 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1171 PH/31 Added sender_verify_fail logging option.
1173 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1174 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1175 tidied the source and removed it altogether.
1177 PH/33 When a queue run was abandoned because the load average was too high, a
1178 log line was always written; now it is written only if the queue_run log
1179 selector is set. In addition, the log line for abandonment now contains
1180 information about the queue run such as the pid. This is always present
1181 in "start" and "stop" lines but was omitted from the "abandon" line.
1183 PH/34 Omit spaces between a header name and the colon in the error message that
1184 is given when verify = headers_syntax fails (if there are lots of them,
1185 the message gets confusing).
1187 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1188 names, as there are now some PTR records that contain slashes. This check
1189 is only to protect against broken name servers that fall over on strange
1190 characters, so the fact that it applies to all lookups doesn't matter.
1192 PH/36 Now that the new test suite is complete, we can remove some of the
1193 special code in Exim that was needed for the old test suite. For example,
1194 sorting DNS records because real resolvers return them in an arbitrary
1195 order. The new test suite's fake resolver always returns records in the
1198 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1199 bounces) except when queue_only is set, to avoid logging races between
1200 the different processes.
1202 PH/38 Panic-die if .include specifies a non-absolute path.
1204 PH/39 A tweak to the "H" retry rule from its user.
1206 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1207 a label. They prevented compilation on older perls.
1209 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1210 a warning to be raised on newish perls.
1212 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1213 on queue. Changes to match documented behaviour of showing count of
1214 messages matching specified criteria.
1216 PH/40 Changed the default ident timeout from 30s to 5s.
1218 PH/41 Added support for the use of login_cap features, on those BSD systems
1219 that have them, for controlling the resources used by pipe deliveries.
1221 PH/42 The content-scanning code uses fopen() to create files in which to put
1222 message data. Previously it was not paying any attention to the mode of
1223 the files. Exim runs with umask(0) because the rest of the code creates
1224 files with open(), and sets the required mode explicitly. Thus, these
1225 files were ending up world-writeable. This was not a big issue, because,
1226 being within the spool directory, they were not world-accessible. I have
1227 created a function called modefopen, which takes an additional mode
1228 argument. It sets umask(777), creates the file, chmods it to the required
1229 mode, then resets the umask. All the relevant calls to fopen() in the
1230 content scanning code have been changed to use this function.
1232 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1233 to 24 hours. This avoids potential overflow problems when processing G
1234 and H retry rules. I suspect nobody ever tinkers with this value.
1236 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1238 PH/45 When the plaintext authenticator is running as a client, the server's
1239 challenges are checked to ensure they are valid base64 strings. By
1240 default, the authentication attempt is cancelled if an invalid string is
1241 received. Setting client_ignore_invalid_base64 true ignores these errors.
1242 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1243 they are received. Thus, the responses can be made to depend on the
1244 challenges. If an invalid string is ignored, an empty string is placed in
1247 PH/46 Messages that are created by the autoreply transport now contains a
1248 References: header, in accordance with RFCs 2822 and 3834.
1250 PH/47 Added authenticated_sender_force to the smtp transport.
1252 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1254 PH/49 Installed latest patch from the Sieve maintainer.
1256 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1257 was also set, the appendfile transport was unnecessarily scanning a
1258 directory of message files (e.g. for maildir delivery) to find the count
1259 of files (along with the size), even though it did not need this
1260 information. It now does the scan only if it needs to find either the
1261 size of the count of files.
1263 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1265 PH/52 Two bugs concerned with error handling when the smtp transport is
1268 (i) Exim was not creating retry information for temporary errors given
1269 for individual recipients after the DATA command when the smtp transport
1270 was used in LMTP mode. This meant that they could be retried too
1271 frequently, and not timed out correctly.
1273 (ii) Exim was setting the flag that allows error details to be returned
1274 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1275 recipients that were returned after the DATA command.
1277 PH/53 This is related to PH/52, but is more general: for any failing address,
1278 when detailed error information was permitted to be returned to the
1279 sender, but the error was temporary, then after the final timeout, only
1280 "retry timeout exceeded" was returned. Now it returns the full error as
1281 well as "retry timeout exceeded".
1283 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1284 do this, and (what is worse) MTAs that accept it.
1286 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1287 will now be deprecated.
1289 PH/56 New os.c-cygwin from the Cygwin maintainer.
1291 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1292 formats (previously only available in exim formats via -bpr, -bpru,
1293 and -bpra. Now also available in native and exiqgrep formats)
1295 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1296 with very large, slow to parse queues
1298 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1300 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1302 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1303 responses to authentication challenges, though it was showing the
1304 challenges; (ii) I've removed the CR characters from the debug output for
1307 PH/58 Allow for the insertion of a newline as well as a space when a string
1308 is turned into more than one encoded-word during RFC 2047 encoding. The
1309 Sieve code now uses this.
1311 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1312 data_4xx, lost_connection, tls_required.
1314 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1315 message was being sent as an SMTP response.
1317 PH/61 Add -l and -k options to exicyclog.
1319 PH/62 When verifying, if an address was redirected to one new address, so that
1320 verification continued, and the new address failed or deferred after
1321 having set something in $address_data, the value of $address_data was not
1322 passed back to the ACL. This was different to the case when no
1323 redirection occurred. The value is now passed back in both cases.
1325 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1326 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1327 use login_cap.h, so on its own it isn't the distinguishing feature. The
1328 new name refers directly to the setclassresources() function.
1330 PH/65 Added configuration files for NetBSD3.
1332 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1334 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1335 is preferred over IPv4.
1337 PH/68 The bounce_return_message and bounce_return_body options were not being
1338 honoured for bounces generated during the reception of non-SMTP messages.
1339 In particular, this applied to messages rejected by the ACL. This bug has
1340 been fixed. However, if bounce_return_message is true and bounce_return_
1341 body is false, the headers that are returned for a non-SMTP message
1342 include only those that have been read before the error was detected.
1343 (In the case of an ACL rejection, they have all been read.)
1345 PH/69 The HTML version of the specification is now built in a directory called
1346 spec_html instead of spec.html, because the latter looks like a path with
1347 a MIME-type, and this confuses some software.
1349 PH/70 Catch two compiler warnings in sieve.c.
1351 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1352 function verify_get_ident() calls ip_connect() to connect a socket, but
1353 if the "connect()" function timed out, ip_connect() used to close the
1354 socket. However, verify_get_ident() also closes the socket later, and in
1355 between Exim writes to the log, which may get opened at this point. When
1356 the socket was closed in ip_connect(), the log could get the same file
1357 descriptor number as the socket. This naturally causes chaos. The fix is
1358 not to close the socket in ip_connect(); the socket should be closed by
1359 the function that creates it. There was only one place in the code where
1360 this was missing, in the iplookup router, which I don't think anybody now
1361 uses, but I've fixed it anyway.
1363 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1364 well as to direct DNS lookups. Otherwise the handling of names in host
1365 lists is inconsistent and therefore confusing.
1371 PH/01 Two changes to the default runtime configuration:
1373 (1) Move the checks for relay_from_hosts and authenticated clients from
1374 after to before the (commented out) DNS black list checks.
1376 (2) Add control=submission to the relay_from_hosts and authenticated
1377 clients checks, on the grounds that messages accepted by these
1378 statements are most likely to be submissions.
1380 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1382 (1) Generate an error if the third argument for the ${prvs expansion is
1385 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1388 (3) Reset the variables that are obtained from the first argument of
1389 ${prvscheck and used in the second argument before leaving the code,
1390 because their memory is reclaimed, so using them afterwards may do
1393 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1394 one (it's much easier than Tom thought :-).
1396 (5) Because of (4), we can now allow for the use of $prvscheck_result
1397 inside the third argument.
1399 PH/03 For some reason, the default setting of PATH when running a command from
1400 a pipe transport was just "/usr/bin". I have changed it to
1403 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1404 anything to be listed in the output from -bV.
1406 PH/05 When a filter generated an autoreply, the entire To: header line was
1407 quoted in the delivery log line, like this:
1409 => >A.N.Other <ano@some.domain> <original@ddress> ...
1411 This has been changed so that it extracts the operative address. There
1412 may be more than one such address. If so, they are comma-separated, like
1415 => >ano@some.domain,ona@other.domain <original@ddress> ...
1417 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1418 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1419 not being looked up in the rDNS to get a host name, Exim was showing the
1420 IP address twice in Received: lines, even though the IP addresses were
1421 identical. For example:
1423 Received: from [1.2.3.4] (helo=[1.2.3.4])
1425 However, if the real host name was known, it was omitting the HELO data
1426 if it matched the actual IP address. This has been tidied up so that it
1427 doesn't show the same IP address twice.
1429 PH/07 When both +timestamp and +memory debugging was on, the value given by
1430 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1431 called by the debug printing, thereby overwriting the timestamp buffer.
1432 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1434 PH/08 When the original message was included in an autoreply transport, it
1435 always said "this is a copy of the message, including all the headers",
1436 even if body_only or headers_only was set. It now gives an appropriate
1439 PH/09 Applied a patch from the Sieve maintainer which:
1441 o fixes some comments
1442 o adds the (disabled) notify extension core
1443 o adds some debug output for the result of if/elsif tests
1444 o points to the current vacation draft in the documentation
1445 and documents the missing references header update
1449 o fixes a bug in processing the envelope test (when testing
1450 multiple envelope elements, the last element determinted the
1453 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1454 Electronic Mail") by including:
1456 Auto-submitted: auto-generated
1458 in the messages that it generates (bounce messages and others, such as
1459 warnings). In the case of bounce messages for non-SMTP mesages, there was
1460 also a typo: it was using "Auto_submitted" (underscore instead of
1461 hyphen). Since every message generated by Exim is necessarily in response
1462 to another message, thes have all been changed to:
1464 Auto-Submitted: auto-replied
1466 in accordance with these statements in the RFC:
1468 The auto-replied keyword:
1470 - SHOULD be used on messages sent in direct response to another
1471 message by an automatic process,
1473 - MUST NOT be used on manually-generated messages,
1475 - MAY be used on Delivery Status Notifications (DSNs) and Message
1476 Disposition Notifications (MDNs),
1478 - MUST NOT be used on messages generated by automatic or periodic
1479 processes, except for messages which are automatic responses to
1482 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1483 to the default Received: header definition.
1485 PH/12 Added log selector acl_warn_skipped (default on).
1487 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1488 variables because (a) they are in the wrong storage pool and (b) even if
1489 they were copied, it wouldn't work properly because of the caching.
1491 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1492 checking when decoding. Apparently there are clients that generate
1493 overlong encoded strings. Why am I not surprised?
1495 PH/15 If the first argument of "${if match_address" was not empty, but did not
1496 contain an "@" character, Exim crashed. Now it writes a panic log message
1497 and treats the condition as false.
1499 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1501 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1502 extension "envelope-auth". The code is finished and in agreement with
1503 other implementations, but there is no documentation so far and in fact,
1504 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1505 not changing the active code.
1507 Print executed "if" and "elsif" statements when debugging is used. This
1508 helps a great deal to understand what a filter does.
1510 Document more things not specified clearly in RFC3028. I had all this
1511 sorted out, when out of a sudden new issues came to my mind. Oops."
1513 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1516 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1517 canonical form (as documented). However, after a host name lookup from
1518 the IP address, check_host() was doing a simple string comparison with
1519 addresses acquired from the DNS when checking that the found name did
1520 have the original IP as one of its addresses. Since any found IPv6
1521 addresses are likely to be in abbreviated form, the comparison could
1522 fail. Luckily, there already exists a function for doing the comparison
1523 by converting both addresses to binary, so now that is used instead of
1524 the text comparison.
1526 PH/20 There was another similar case to PH/19, when a complete host name was
1527 given in a host list; looking up its IP address could give an abbreviated
1528 form, whereas the current host's name might or might not be abbreviated.
1529 The same fix has been applied.
1535 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1536 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1539 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1540 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1542 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1544 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1545 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1546 Also, for "Auto-Submitted", treat anything other than "no" as
1547 non-personal, in accordance with RFC 3834. (Previously it treated
1548 anything starting "auto-" as non-personal.)
1550 TF/01 The control=submission/name=... option had a problem with syntax
1551 errors if the name included a slash character. The /name= option
1552 now slurps the rest of the string, so it can include any characters
1553 but it must come last in the list of options (after /sender_retain
1556 PH/05 Some modifications to the interface to the fake nameserver for the new
1564 TK/01 Added the "success_on_redirect" address verification option. See
1565 NewStuff for rationale and an example.
1567 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1569 PH/02 Patch to exigrep to allow it to work on syslog lines.
1571 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1572 fread() to skip over the body file's header line, because in Cygwin the
1573 header line is locked and is inaccessible.
1575 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1576 co-exist for some time) to make it clear that it is the Exim ID that is
1577 referenced, not the Message-ID: header line.
1579 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1580 string_format() function, because snprintf() does not exist on all
1583 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1584 ${stat: expansion item.
1586 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1587 protocol synchronization error", to keep the pedants happy.
1589 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1590 well as for IRIX systems, when gcc is being used. See the host.c source
1593 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1595 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1598 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1599 transport and to the smtp transport in LMTP mode.
1601 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1603 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1604 run a filter in a subprocess. This could lead to confusion in subsequent
1605 lookups in the parent process. There should also be a search_tidyup() at
1606 the end of the subprocess.
1608 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1609 only if the host matched helo_try_verify_hosts, which caused the
1610 verification to occur when the EHLO/HELO command was issued. The ACL just
1611 tested the remembered result. Now, if a previous verification attempt has
1612 not happened, "verify = helo" does it there and then.
1614 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1616 TK/03 Fix log output including CR from clamd.
1618 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1619 exist provoked a memory error which could cause a segfault.
1621 PH/15 Installed PCRE 6.2
1623 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1625 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1626 of the problem. Specifically, suggested +O2 rather than +O1 for the
1629 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1631 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1632 record, the port was not correctly logged when the outgoing_port log
1633 selector was set (it logged the transort's default port).
1635 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1636 fallback_hosts, and "hosts" in the smtp transport.
1638 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1639 host errors such as "Connection refused".
1641 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1642 authentication with radiusclient 0.4.9:
1644 - Error returned from rc_read_config was caught wrongly
1645 - Username/password not passed on to radius server due to wrong length.
1647 The presumption is that some radiusclient API changes for 4.51/PH/17
1648 were not taken care of correctly. The code is still untested by me (my
1649 Linux distribution still has 0.3.2 of radiusclient), but it was
1650 contributed by a Radius user.
1652 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1653 expanding the "port" option of the smtp transport.
1655 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1656 while reading a MIME header. Thanks to Tom Hughes for a patch.
1658 PH/24 Include config.h inside local_scan.h so that configuration settings are
1661 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1662 that in an ACL for RCPT (for example), you can examine exactly what was
1665 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1666 commands, but it was not correctly comparing the address with the actual
1667 client host address. Thus, it would show the EHLO address in Received:
1668 header lines when this was not necessary.
1670 PH/27 Added the % operator to ${eval:}.
1672 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1673 it should be ignoring failures (because with -C, for example, it has lost
1674 privilege). It wasn't ignoring creation failures other than "already
1677 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1680 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1681 because an input line was too long, either on its own, or by virtue of
1682 too many continuations, the temporary file was not being removed, and the
1683 return code was incorrect.
1685 PH/31 Missing "BOOL" in function definition in filtertest.c.
1687 PH/32 Applied Sieve patches from the maintainer.
1689 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1691 PH/33 Added "verify = not_blind".
1693 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1694 Local/Makefile (with some defaults set). These are used in built scripts
1695 such as exicyclog, but they have never been used in the exim_install
1696 script (though there are many overriding facilities there). I have
1697 arranged that the exim_install script now takes note of these two
1700 PH/35 Installed configuration files for Dragonfly.
1702 PH/36 When a locally submitted message by a trusted user did not contain a
1703 From: header, and the sender address was obtained from -f or from an SMTP
1704 MAIL command, and the trusted user did not use -F to supply a sender
1705 name, $originator_name was incorrectly used when constructing a From:
1706 header. Furthermore, $originator_name was used for submission mode
1707 messages from external hosts without From: headers in a similar way,
1708 which is clearly wrong.
1710 PH/37 Added control=suppress_local_fixups.
1712 PH/38 When log_selector = +received_sender was set, and the addition of the
1713 sender made the log line's construction buffer exactly full, or one byte
1714 less than full, an overflow happened when the terminating "\n" was
1717 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1718 when the result of a list match is failure because a DNS lookup failed.
1720 PH/40 RM_COMMAND is now used in the building process.
1722 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1723 the "build-* directories that it finds.
1725 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1726 address in a domain literal was a prefix of an interface address.
1728 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1729 when verifying a sender address, unless rewrite_headers is false.
1731 PH/44 Wrote a long comment about why errors_to addresses are verified as
1732 recipients, not senders.
1734 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1735 the ratelimit ACL was added.
1737 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1739 PH/46 Added extra information about PostgreSQL errors to the error string.
1741 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1742 avoiding the need to install special zones in a real server. This is
1743 backwards compatible; if it can't find the fake resolver, it drops back.
1744 Thus, both old and new test suites can be run.
1746 TF/02 Added util/ratelimit.pl
1748 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1749 clock is set back in time.
1751 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1752 Candler <B.Candler@pobox.com>.
1754 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1755 OK for addresses that are the result of redirections.
1757 PH/48 A number of further additions for the benefit of the new test suite,
1758 including a fake gethostbyname() that interfaces to the fake DNS resolver
1761 TF/06 The fix for widen_domains has also been applied to qualify_single and
1762 search_parents which are the other dnslookup options that can cause
1765 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1768 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1770 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1771 service" if it encountered an MX record with an empty target hostname.
1772 The message is now "an MX or SRV record indicated no SMTP service".
1774 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1775 if the DNS of the sending site is misconfigured. This is quite a
1776 common situation. This change restores the behaviour of treating a
1777 helo verification defer as a failure.
1779 PH/51 If self=fail was set on a router, the bounce message did not include the
1780 actual error message.
1786 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1788 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1789 command itself ended in error, the underlying message about the transport
1790 filter timeout was being overwritten with the pipe command error. Now the
1791 underlying error message should be appended to the second error message.
1793 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1795 PH/02 Reduce the amount of output that "make" produces by default. Full output
1796 can still be requested.
1798 PH/03 The warning log line about a condition test deferring for a "warn" verb
1799 was being output only once per connection, rather than after each
1800 occurrence (because it was using the same function as for successful
1801 "warn" verbs). This seems wrong, so I have changed it.
1803 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1804 it should not have, which might have caused a crash in the right
1805 circumstances, but probably never did.
1807 PH/04 Installed a modified version of Tony Finch's patch to make submission
1808 mode fix the return path as well as the Sender: header line, and to
1809 add a /name= option so that you can make the user's friendly name appear
1812 TF/03 Added the control = fakedefer ACL modifier.
1814 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1815 Mark Lowes for thorough testing.
1817 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1819 TK/03 Merged latest SRS patch from Miles Wilton.
1821 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1822 with the definition in sysexits.h (which is #included earlier).
1823 Fortunately, Exim does not actually use EX_OK. The code used to try to
1824 preserve the sysexits.h value, by assumimg that macro definitions were
1825 scanned for macro replacements. I have been disabused of this notion,
1826 so now the code just undefines EX_OK before #including unistd.h.
1828 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1829 in the smtp transport. When a block could not be written in a single
1830 write() function, the timeout was being re-applied to each part-write.
1831 This seems wrong - if the receiver was accepting one byte at a time it
1832 would take for ever. The timeout is now adjusted when this happens. It
1833 doesn't have to be particularly precise.
1835 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1836 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1838 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1839 without a domain if the check to <postmaster@domain> fails.
1841 SC/01 Eximstats: added -xls and the ability to specify output files
1842 (patch written by Frank Heydlauf).
1844 SC/02 Eximstats: use FileHandles for outputing results.
1846 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1848 SC/04 Eximstats: fixed display of large numbers with -nvr option
1850 SC/05 Eximstats: fixed merging of reports with empty tables.
1852 SC/06 Eximstats: added the -include_original_destination flag
1854 SC/07 Eximstats: removed tabs and trailing whitespace.
1856 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1858 TK/06 MBOX spool code: Add real "From " MBOX separator line
1859 so the .eml file is really in mbox format (even though
1860 most programs do not really care). Patch from Alex Miller.
1862 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1863 The latter is generated from $received_to and is only set if the
1864 message has one envelope recipient. SA can use these headers,
1865 obviously out-of-the-box. Patch from Alex Miller.
1867 PH/08 The ${def test on a variable was returning false if the variable's
1868 value was "0", contrary to what the specification has always said!
1869 The result should be true unless the variable is empty.
1871 PH/09 The syntax error of a character other than { following "${if
1872 def:variable_name" (after optional whitespace) was not being diagnosed.
1873 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1874 accidental colon was present, for example, could give incorrect results.
1876 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1877 result is used (not including appendfile, where other changes are about
1880 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1881 This involved changing a lot of size variables from int to off_t. It
1882 should work with maildirs and everything.
1884 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1885 spamd dying while we are connected to it.
1887 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1890 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1891 function for BDB 4.3.
1893 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1895 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1896 the sources for PCRE have been re-arranged and more files are now
1899 PH/15 The code I had for printing potentially long long variables in PH/11
1900 above was not the best (it lost precision). The length of off_t variables
1901 is now inspected at build time, and an appropriate printing format (%ld
1902 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1903 to be "long long int" or "long int". This is needed for the internal
1904 formatting function string_vformat().
1906 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1907 the configuration file to be ":syslog", then the script "guesses" where
1908 the logs files are, rather than using the compiled in default. In our
1909 case the guess is not the same as the compiled default, so the script
1910 suddenly stopped working when I started to use syslog. The patch checks
1911 to see if log_file_path is "". If so, it attempts to read it from exim
1912 with no configuration file to get the compiled in version, before it
1913 falls back to the previous guessing code."
1915 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1916 implementing BATV in an Exim configuration. See NewStuff for the gory
1919 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1920 Makefile that are specific to HP-UX.
1922 PH/18 If the "use_postmaster" option was set for a recipient callout together
1923 with the "random" option, the postmaster address was used as the MAIL
1924 FROM address for the random test, but not for the subsequent recipient
1925 test. It is now used for both.
1927 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1928 patch removes a few documentation additions to RFC 3028, because the
1929 latest draft now contains them. It adds the new en;ascii-case comparator
1930 and a new error check for 8bit text in MIME parts. Comparator and
1931 require names are now matched exactly. I enabled the subaddress
1932 extension, but it is not well tested yet (read: it works for me)."
1934 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1935 rework some of the code of TK/09 above to avoid the hardwired use of
1936 "%lld" and "long long". Replaced the call to snprintf() with a call to
1939 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1940 records point to non-existent hosts", "retry timeout exceeded", and
1941 "retry time not reached for any host after a long failure period".
1943 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1944 experimental DomainKeys support:
1946 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1947 (2) On an error such as an illegally used "control", the wrong name for
1948 the control was given.
1950 These problems did NOT occur unless DomainKeys support was compiled.
1952 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1954 PH/24 Added ${if match_ip condition.
1956 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1957 databases so that it will be absolutely obvious if a crash occurs in the
1958 DB library. This is a regular occurrence (often caused by mis-matched
1961 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1962 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1963 picked up on a user's system that detects such things. There doesn't seem
1964 to be a gcc warning option for this - only an attribute that has to be
1965 put on the function's prototype. It seems that in Fedora Core 4 they have
1966 set this on a number of new functions. No doubt there will be more in due
1969 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1970 specify a transport. However, if an address that was verified by such a
1971 router was the subject of a callout, Exim crashed because it tried to
1972 read the rcpt_include_affixes from the non-existent transport. Now it
1973 just assumes that the setting of that option is false. This bug was
1974 introduced by 4.51/PH/31.
1976 PH/28 Changed -d+all to exclude +memory, because that information is very
1977 rarely of interest, but it makes the output a lot bigger. People tend to
1978 do -d+all out of habit.
1980 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1981 code in os-type was giving problems when libc.so lives in lib64, like on
1984 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1985 aren't the modern standard, and it seems that some systems' include files
1986 don't always have them. Exim was already checking for some of the newer
1987 ones like T_AAAA, and defining it itself. I've added checks for all the
1988 record types that Exim uses.
1990 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1991 not automatically generating a new one, as it is supposed to. This
1992 prevented TLS from working. If the file did exist, but contained invalid
1993 data, a new version was generated, as expected. It was only the case of a
1994 non-existent file that was broken.
1996 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1997 with a change in libdomainkeys > 0.64.
1999 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2000 from DNS. If the selector record carries the flag, it now has
2001 precedence over the domain-wide flag.
2003 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2005 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2006 the use of a port name that isn't defined in /etc/services) occurred, the
2007 message was deferred as in a normal delivery, and thus remained on the
2008 spool, instead of being failed because of the mua_wrapper setting. This
2009 is now fixed, and I tidied up some of the mua_wrapper messages at the
2012 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2013 the messages in a hash of arrays rather than using individual hashes.
2014 This is a bit cleaner and results in dramatic memory savings, albeit
2015 at a slight CPU cost.
2017 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2018 as requested by Marc Sherman.
2020 SC/10 Eximstats: added histograms for user specified patterns as requested
2023 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2025 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2026 fopen() in the content-scanning modules that did not already have it.
2032 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2033 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2035 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2037 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2039 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2040 to test Sieve filters that use "vacation".
2042 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2043 that changes the way the GnuTLS parameters are stored in the cache file.
2044 The new format can be generated externally. For backward compatibility,
2045 if the data in the cache doesn't make sense, Exim assumes it has read an
2046 old-format file, and it generates new data and writes a new file. This
2047 means that you can't go back to an older release without removing the
2050 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2051 work if there are any delivery delays because "one_time" forces the
2052 parent to be marked "delivered", so its unseen clone is never tried
2053 again. For this reason, Exim now forbids the simultaneous setting of
2056 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2057 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2058 entry if you want to know the details. The fix, however, neglected to
2059 consider the case when local delivery batching is involved. The test for
2060 "previously delivered" was not happening when checking to see if an
2061 address could be batched with a previous (undelivered) one; under
2062 certain circumstances this could lead to multiple deliveries to the same
2065 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2066 in its include files, and this causes problems building Exim.
2068 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2069 header_syntax) but Exim was just ignoring anything given after a slash.
2070 In particular, this caused confusion with an attempt to use "verify =
2071 reverse_host_lookup/defer_ok". An error is now given when options are
2072 supplied for verify items that do not have them. (Maybe reverse_host_
2073 lookup should have a defer_ok option, but that's a different point.)
2075 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2076 defined by RFC 821) to 2048, because there were problems with some AUTH
2077 commands, and RFC 1869 says the size should be increased for extended
2078 SMTP commands that take arguments.
2080 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2083 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2084 "unknown" error; now it says that the functionality isn't in the binary.
2086 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2087 an address' error message when a string expansion fails (syntax or
2088 whatever). Otherwise the password may appear in the log. Following change
2089 PH/42 below, there is no longer a chance of it appearing in a bounce
2092 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2094 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2095 of its addresses was ever tried. (Bugzilla bug #2.)
2097 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2098 the result incorrectly in the debug output. (It correctly added a newline
2099 to what was transported.)
2101 TF/01 Added $received_time.
2103 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2104 commented out examples of how to interface to a virus scanner and to
2105 SpamAssassin. Also added commented examples of av_scanner and
2106 spamd_address settings.
2108 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2109 and controls are allowed in which ACLs. There were a couple of minor
2110 errors. Some of the entries in the conditions table (which is a table of
2111 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2112 negation of where the condition IS allowed.
2114 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2116 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2117 header file does not have a version number, so I've had to invent a new
2118 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2119 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2120 radiusclient), but it was contributed by a Radius user.
2122 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2123 files or directories, for OpenSSL.
2125 PH/19 When an Exim process that is running as root has to create an Exim log
2126 file, it does so in a subprocess that runs as exim:exim so as to get the
2127 ownership right at creation (otherwise, other Exim processes might see
2128 the file with the wrong ownership). There was no test for failure of this
2129 fork() call, which would lead to the process getting stuck as it waited
2130 for a non-existent subprocess. Forks do occasionally fail when resources
2131 run out. I reviewed all the other calls to fork(); they all seem to check
2134 PH/20 When checking for unexpected SMTP input at connect time (before writing
2135 the banner), Exim was not dealing correctly with a non-positive return
2136 from the read() function. If the client had disconnected by this time,
2137 the result was a log entry for a synchronization error with an empty
2138 string after "input=" when read() returned zero. If read() returned -1
2139 (an event I could not check), uninitialized data bytes were printed.
2140 There were reports of junk text (parts of files, etc) appearing after
2143 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2145 PH/22 Added support for macro redefinition, and (re)definition in between
2146 driver and ACL definitions.
2148 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2149 forgetting to use the resulting value; it was using the unexpanded value.
2151 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2152 hadn't been configured. The fix is from Juergen Kreileder, who
2153 understands it better than I do:
2155 "Here's what I see happening with three configured cyrus_sasl
2156 authenticators configured (plain, login, cram-md5):
2158 On startup auth_cyrus_sasl_init() gets called for each of these.
2159 This means three calls to sasl_listmech() without a specified mech_list.
2160 => SASL tests which mechs of all available mechs actually work
2161 => three warnings about OTP not working
2162 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2164 With the patch, sasl_listmech() also gets called three times. But now
2165 SASL's mech_list option is set to the server_mech specified in the the
2166 authenticator. Or in other words, the answer from sasl_listmech()
2167 gets limited to just the mech you're testing for (which is different
2169 => the return list contains just 'plain' or 'login', 'cram-md5' or
2170 nothing depending on the value of ob->server_mech.
2172 I've just tested the patch: Authentication still works fine,
2173 unavailable mechs specified in the exim configuration are still
2174 caught, and the auth.log warnings about OTP are gone."
2176 PH/25 When debugging is enabled, the contents of the command line are added
2177 to the debugging output, even when log_selector=+arguments is not
2180 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2181 answer is "GNU", and only if the return is "GNU/something" is the answer
2184 PH/27 $acl_verify_message is now set immediately after the failure of a
2185 verification in an ACL, and so is available in subsequent modifiers. In
2186 particular, the message can be preserved by coding like this:
2188 warn !verify = sender
2189 set acl_m0 = $acl_verify_message
2191 Previously, $acl_verify_message was set only while expanding "message"
2192 and "log_message" when a very denied access.
2194 PH/28 Modified OS/os.c-Linux with
2196 -#ifndef OS_LOAD_AVERAGE
2197 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2199 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2200 nomenclature these days.)
2202 PH/29 Installed patch from the Sieve maintainer that adds the options
2203 sieve_useraddress and sieve_subaddress to the redirect router.
2205 PH/30 In these circumstances:
2206 . Two addresses routed to the same list of hosts;
2207 . First host does not offer TLS;
2208 . First host accepts first address;
2209 . First host gives temporary error to second address;
2210 . Second host offers TLS and a TLS session is established;
2211 . Second host accepts second address.
2212 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2213 and peerdn, if requested) that were in fact used only for the second
2216 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2217 attention to any local part prefix or suffix that was matched by the
2218 router that accepted the address. It now behaves in the same way as it
2219 does for delivery: the affixes are removed from the local part unless
2220 rcpt_include_affixes is set on the transport.
2222 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2223 timeout during the DATA phase of an incoming message.
2225 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2226 applied a patch sanctioned by the Sieve maintainer.
2228 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2229 the uid or gid is negative. A case of a negative gid caused this to be
2230 noticed. The fix allows for either to be negative.
2232 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2233 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2236 PH/36 The change PH/12 above was broken. Fixed it.
2238 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2239 the grounds that routing the same address twice would always produce the
2240 same answer. This might have been true once, but it is certainly no
2241 longer true now. Routing a child address may depend on the previous
2242 routing that produced that child. Some complicated redirection strategies
2243 went wrong when messages had multiple recipients, and made Exim's
2244 behaviour dependent on the order in which the addresses were given.
2246 I have moved the duplicate checking until after the routing is complete.
2247 Exim scans the addresses that are assigned to local and remote
2248 transports, and removes any duplicates. This means that more work will be
2249 done, as duplicates will always all be routed, but duplicates are
2250 presumably rare, so I don't expect this is of any significance.
2252 For deliveries to pipes, files, and autoreplies, the duplicate checking
2253 still happens during the routing process, since they are not going to be
2256 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2257 It corrects a timeout issue with spamd. This is Ian's comment: "The
2258 background is that sometimes spamd either never reads data from a
2259 connection it has accepted, or it never writes response data. The exiscan
2260 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2261 blindly assumes that writes won't block so it may never time out."
2263 PH/39 Allow G after quota size as well as K and M.
2265 PH/40 The value set for $authenticated_id in an authenticator may not contain
2266 binary zeroes or newlines because the value is written to log lines and
2267 to spool files. There was no check on this. Now the value is run through
2268 the string_printing() function so that such characters are converted to
2269 printable escape sequences.
2271 PH/41 $message_linecount is a new variable that contains the total number of
2272 lines in the message. Compare $body_linecount, which is the count for the
2275 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2276 bounce and delay warning messages, except in certain special cases, which
2279 (a) An SMTP error message from a remote host;
2280 (b) A message specified in a :fail: redirection;
2281 (c) A message specified in a "fail" command in a system filter;
2282 (d) A message specified in a FAIL return from the queryprogram router;
2283 (e) A message specified by the cannot_route_message router option.
2285 In these cases only, Exim does include the error details in bounce and
2286 warning messages. There are also a few cases where bland messages such
2287 as "unrouteable address" or "local delivery error" are given.
2289 PH/43 $value is now also set for the "else" part of a ${run expansion.
2291 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2292 being worked on, but at least Exim now implements the latest version to
2295 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2296 process to complete was treated as a delivery failure, a timeout while
2297 writing the message to the pipe was logged, but erroneously treated as a
2298 successful delivery. Such timeouts include transport filter timeouts. For
2299 consistency with the overall process timeout, these timeouts are now
2300 treated as errors, giving rise to delivery failures by default. However,
2301 there is now a new Boolean option for the pipe transport called
2302 timeout_defer, which, if set TRUE, converts the failures into defers for
2303 both kinds of timeout. A transport filter timeout is now identified in
2306 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2307 systems where "make" and "gmake" are different, calling "gmake" at top
2308 level broke things. I've arranged for the value of $(MAKE) to be passed
2309 from the Makefile to this script so that it can call the same version of
2313 A note about Exim versions 4.44 and 4.50
2314 ----------------------------------------
2316 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2317 changes of various kinds. As a consequence, a big documentation update was
2318 needed. This delayed the release for rather longer than seemed good, especially
2319 in the light of a couple of (minor) security issues. Therefore, the changes
2320 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2321 release. So 4.44 and 4.50 are in effect two different branches that both start
2324 I have left the 4.50 change log unchanged; it contains all the changes since
2325 4.43. The change log for 4.44 is below; many of its items are identical to
2326 those for 4.50. This seems to be the most sensible way to preserve the
2327 historical information.
2333 1. Minor wording change to the doc/README.SIEVE file.
2335 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2336 computation of the current number of files was incorrect.
2338 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2339 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2342 4. Give more explanation in the error message when the command for a transport
2343 filter fails to execute.
2345 5. There are several places where Exim runs a non-Exim command in a
2346 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2347 was being done only for the command run by the queryprogram router. It is
2348 now done for all such subprocesses. The other cases are: ${run, transport
2349 filters, and the commands run by the lmtp and pipe transports.
2351 6. Added CONFIGURE_GROUP build-time option.
2353 7. Some older OS have a limit of 256 on the maximum number of file
2354 descriptors. Exim was using setrlimit() to set 1000 as a large value
2355 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2356 systems. I've change it so that if it can't get 1000, it tries for 256.
2358 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2359 was an oversight, and furthermore, ever since the addition of extra
2360 controls (e.g. 4.43/32), the checks on when to allow different forms of
2361 "control" were broken. There should now be diagnostics for all cases when a
2362 control that does not make sense is encountered.
2364 9. Added the /retain_sender option to "control=submission".
2366 10. $recipients is now available in the predata ACL (oversight).
2368 11. Tidy the search cache before the fork to do a delivery from a message
2369 received from the command line. Otherwise the child will trigger a lookup
2370 failure and thereby defer the delivery if it tries to use (for example) a
2371 cached ldap connection that the parent has called unbind on.
2373 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2374 of $address_data from the recipient verification was clobbered by the
2375 sender verification.
2377 13. The value of address_data from a sender verification is now available in
2378 $sender_address_data in subsequent conditions in the ACL statement.
2380 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2382 15. Added a new option "connect=<time>" to callout options, to set a different
2385 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2386 was its contents. (It was OK if the option was not defined at all.)
2388 17. A "Completed" log line is now written for messages that are removed from
2389 the spool by the -Mrm option.
2391 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2392 information about exactly what failed.
2394 19. Added -dd to debug only the daemon process.
2396 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2397 handles timeouts, both on the server side and network timeouts. Renamed the
2398 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2400 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2401 It is now set to "smtps".
2403 22. $host_address is now set to the target address during the checking of
2404 ignore_target_hosts.
2406 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2407 being passed; this would have caused $sender_host_name to have been used if
2408 matching the list had actually called for a host name (not very likely,
2409 since this list is usually IP addresses). A host name is now passed as
2412 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2413 code that specifies a non-restarting handler (typically sigaction() in
2414 modern systems) in an attempt to fix a rare and obscure crash bug.
2416 25. Narrowed the window for a race in the daemon that could cause it to ignore
2417 SIGCHLD signals. This is not a major problem, because they are used only to
2418 wake it up if nothing else does.
2420 26. A malformed maildirsize file could cause Exim to calculate negative values
2421 for the mailbox size or file count. Odd effects could occur as a result.
2422 The maildirsize information is now recalculated if the size or filecount
2425 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2426 support for a long time. Removed HAVE_SYS_VFS_H.
2428 28. Installed the latest version of exipick from John Jetmore.
2430 29. In an address list, if the pattern was not a regular expression, an empty
2431 subject address (from a bounce message) matched only if the pattern was an
2432 empty string. Non-empty patterns were not even tested. This was the wrong
2433 because it is perfectly reasonable to use an empty address as part of a
2434 database query. An empty address is now tested by patterns that are
2435 lookups. However, all the other forms of pattern expect the subject to
2436 contain a local part and a domain, and therefore, for them, an empty
2437 address still always fails if the pattern is not itself empty.
2439 30. Exim went into a mad DNS loop when attempting to do a callout where the
2440 host was specified on an smtp transport, and looking it up yielded more
2441 than one IP address.
2443 31. Re-factored the code for checking spool and log partition space into a
2444 function that finds that data and another that does the check. The former
2445 is then used to implement four new variables: $spool_space, $log_space,
2446 $spool_inodes, and $log_inodes.
2448 32. The RFC2047 encoding function was originally intended for short strings
2449 such as real names; it was not keeping to the 75-character limit for
2450 encoded words that the RFC imposes. It now respects the limit, and
2451 generates multiple encoded words if necessary. To be on the safe side, I
2452 have increased the buffer size for the ${rfc2047: expansion operator from
2455 33. It is now permitted to omit both strings after an "if" condition; if the
2456 condition is true, the result is "true". As before, when the second string
2457 is omitted, a false condition yields an empty string. This makes it less
2458 cumbersome to write custom ACL and router conditions.
2460 34. Failure to deliver a bounce message always caused it to be frozen, even if
2461 there was an errors_to setting on the router. The errors_to setting is now
2464 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2465 canonical form (fully expanded) before being placed in
2466 $sender_host_address.
2468 36. The table in the code that translates DNS record types into text (T_A to
2469 "A" for instance) was missing entries for NS and CNAME. It is just possible
2470 that this could have caused confusion if both these types were looked up
2471 for the same domain, because the text type is used as part of Exim's
2472 per-process caching. But the chance of anyone hitting this buglet seems
2475 37. The dnsdb lookup has been extended in a number of ways.
2477 (1) There is a new type, "zns", which walks up the domain tree until it
2478 finds some nameserver records. It should be used with care.
2480 (2) There is a new type, "mxh", which is like "mx" except that it returns
2481 just the host names, not the priorities.
2483 (3) It is now possible to give a list of domains (or IP addresses) to be
2484 looked up. The behaviour when one of the lookups defers can be
2485 controlled by a keyword.
2487 (4) It is now possible to specify the separator character for use when
2488 multiple records are returned.
2490 38. The dnslists ACL condition has been extended: it is now possible to supply
2491 a list of IP addresses and/or domains to be looked up in a particular DNS
2494 39. Added log_selector=+queue_time_overall.
2496 40. When running the queue in the test harness, wait just a tad after forking a
2497 delivery process, to get repeatability of debugging output.
2499 41. Include certificate and key file names in error message when GnuTLS fails
2500 to set them up, because the GnuTLS error message doesn't include the name
2501 of the failing file when there is a problem reading it.
2503 42. Allow both -bf and -bF in the same test run.
2505 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2507 44. The "Exiscan patch" is now merged into the mainline Exim source.
2509 45. Sometimes the final signoff response after QUIT could fail to get
2510 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2511 < 0 before doing a fflush(). This bug looks as though it goes back to the
2512 introduction of TLS in release 3.20, but "sometimes" must have been rare
2513 because the tests only now provoked it.
2515 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2516 (this can affect the format of dates).
2518 47. exim_tidydb, when checking for the continued existence of a message for
2519 which it has found a message-specific retry record, was not finding
2520 messages that were in split spool directories. Consequently, it was
2521 deleting retry records that should have stayed in existence.
2523 48. Steve fixed some bugs in eximstats.
2525 49. The SPA authentication driver was not abandoning authentication and moving
2526 on to the next authenticator when an expansion was forced to fail,
2527 contradicting the general specification for all authenticators. Instead it
2528 was generating a temporary error. It now behaves as specified.
2530 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2531 (the order specifies the preference for clients). The order is now AES256,
2532 AES128, 3DES, ARCFOUR128.
2534 51. Small patch to Sieve code - explicitly set From: when generating an
2537 52. Exim crashed if a remote delivery caused a very long error message to be
2538 recorded - for instance if somebody sent an entire SpamAssassin report back
2539 as a large number of 550 error lines. This bug was coincidentally fixed by
2540 increasing the size of one of Exim's internal buffers (big_buffer) that
2541 happened as part of the Exiscan merge. However, to be on the safe side, I
2542 have made the code more robust (and fixed the comments that describe what
2545 53. Now that there can be additional text after "Completed" in log lines (if
2546 the queue_time_overall log selector is set), a one-byte patch to exigrep
2547 was needed to allow it to recognize "Completed" as not the last thing in
2550 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2551 patch that reportedly fixes this has been added. I am not expert enough to
2552 create a test for it. This is what the patch creator wrote:
2554 "I found a little strange behaviour of ldap code when working with
2555 Windows 2003 AD Domain, where users was placed in more than one
2556 Organization Units. When I tried to give exim partial DN, the exit code
2557 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2558 But simultaneously result of request was absolutely normal ldap result,
2559 so I produce this patch..."
2561 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2562 REFERENCE, so I have modified the code to exclude the patch when that macro
2565 55. Some experimental protocols are using DNS PTR records for new purposes. The
2566 keys for these records are domain names, not reversed IP addresses. The
2567 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2568 leaves it alone. Component reversal etc. now happens only for IP addresses.
2571 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2573 57. Double the size of the debug message buffer (to 2048) so that more of very
2574 long debug lines gets shown.
2576 58. The exicyclog utility now does better if the number of log files to keep
2577 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2579 59. Two changes related to the smtp_active_hostname option:
2581 (1) $smtp_active_hostname is now available as a variable.
2582 (2) The default for smtp_banner uses $smtp_active_hostname instead
2583 of $primary_hostname.
2585 60. The host_aton() function is supposed to be passed a string that is known
2586 to be a valid IP address. However, in the case of IPv6 addresses, it was
2587 not checking this. This is a hostage to fortune. Exim now panics and dies
2588 if the condition is not met. A case was found where this could be provoked
2589 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2590 components; fortuitously, this particular loophole had already been fixed
2591 by change 4.50/55 above.
2593 If there are any other similar loopholes, the new check in host_aton()
2594 itself should stop them being exploited. The report I received stated that
2595 data on the command line could provoke the exploit when Exim was running as
2596 exim, but did not say which command line option was involved. All I could
2597 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2598 running as the user.
2601 61. There was a buffer overflow vulnerability in the SPA authentication code
2602 (which came originally from the Samba project). I have added a test to the
2603 spa_base64_to_bits() function which I hope fixes it.
2606 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2607 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2609 63. The daemon start-up calls getloadavg() while still root for those OS that
2610 need the first call to be done as root, but it missed one case: when
2611 deliver_queue_load_max is set with deliver_drop_privilege. This is
2612 necessary for the benefit of the queue runner, because there is no re-exec
2613 when deliver_drop_privilege is set.
2615 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2616 This has been fixed.
2618 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2619 was in use, was not putting the data itself into the right store pool;
2620 consequently, it could be overwritten for a subsequent message in the same
2621 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2624 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2626 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2627 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2628 the function were treating the return as a boolean value, which happened to
2629 work because 0=false and not-0=true, but is not correct code.
2631 68. The host_aton() function was not handling scoped IPv6 addresses (those
2632 with, for example, "%eth0" on the end) correctly.
2634 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2635 negated items (that is, ~something) in unsigned ints. Some compilers
2636 apparently mutter when there is no cast.
2638 70. If an address verification called from an ACL failed, and did not produce a
2639 user-specific message (i.e. there was only a "system" message), nothing was
2640 put in $acl_verify_message. In this situation, it now puts the system
2643 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2644 session; change 4.31/43 added the unwanted input to the log line - except
2645 that it did not do this in the start of session case. It now does.
2647 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2648 This is wrong and can cause the other end to generate a synchronization
2649 error if it is another Exim or anything else that does the synchronization
2650 check. A QUIT command is no longer sent after a timeout.
2652 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2653 during host lookups.
2655 74. The defer_ok option of callout verification was not working if it was used
2656 when verifying addresses in header lines, that is, for this case:
2658 verify = header_sender/callout=defer_ok
2660 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2661 those file descriptors could be used for SMTP connections. If anything
2662 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2663 could be sent to the SMTP client, causing chaos. The daemon now opens
2664 stdin, stdout, and stderr to /dev/null when it puts itself into the
2667 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2668 log by default. The user can override this with suitable Perl magic.
2670 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2671 the log message when discard triggers, was not working for the DATA ACL or
2672 for the non-SMTP ACL.
2674 78. Error message wording change in sieve.c.
2676 79. If smtp_accept_max_per_host was set, the number of connections could be
2677 restricted to fewer than expected, because the daemon was trying to set up
2678 a new connection before checking whether the processes handling previous
2679 connections had finished. The check for completed processes is now done
2680 earlier. On busy systems, this bug wouldn't be noticed because something
2681 else would have woken the daemon, and it would have reaped the completed
2684 80. If a message was submitted locally by a user whose login name contained one
2685 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2686 It caused a spool format error. I have fixed the spool reading code. A
2687 related problem was that the "from" clause in the Received: line became
2688 illegal because of the space(s). It is now covered by ${quote_local_part.
2690 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2693 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2695 83. Patch from Sieve maintainer to fix unterminated string problem in
2696 "vacation" handling.
2698 84. Some minor changes to the Linux configuration files to help with other
2699 OS variants using glibc.
2701 85. One more patch for Sieve to update vacation handling to latest spec.
2704 ----------------------------------------------------
2705 See the note above about the 4.44 and 4.50 releases.
2706 ----------------------------------------------------
2712 1. Change 4.43/35 introduced a bug that caused file counts to be
2713 incorrectly computed when quota_filecount was set in an appendfile
2716 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2717 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2720 3. Give more explanation in the error message when the command for a transport
2721 filter fails to execute.
2723 4. There are several places where Exim runs a non-Exim command in a
2724 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2725 was being done only for the command run by the queryprogram router. It is
2726 now done for all such subprocesses. The other cases are: ${run, transport
2727 filters, and the commands run by the lmtp and pipe transports.
2729 5. Some older OS have a limit of 256 on the maximum number of file
2730 descriptors. Exim was using setrlimit() to set 1000 as a large value
2731 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2732 systems. I've change it so that if it can't get 1000, it tries for 256.
2734 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2735 was an oversight, and furthermore, ever since the addition of extra
2736 controls (e.g. 4.43/32), the checks on when to allow different forms of
2737 "control" were broken. There should now be diagnostics for all cases when a
2738 control that does not make sense is encountered.
2740 7. $recipients is now available in the predata ACL (oversight).
2742 8. Tidy the search cache before the fork to do a delivery from a message
2743 received from the command line. Otherwise the child will trigger a lookup
2744 failure and thereby defer the delivery if it tries to use (for example) a
2745 cached ldap connection that the parent has called unbind on.
2747 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2748 of $address_data from the recipient verification was clobbered by the
2749 sender verification.
2751 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2752 was its contents. (It was OK if the option was not defined at all.)
2754 11. A "Completed" log line is now written for messages that are removed from
2755 the spool by the -Mrm option.
2757 12. $host_address is now set to the target address during the checking of
2758 ignore_target_hosts.
2760 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2761 being passed; this would have caused $sender_host_name to have been used if
2762 matching the list had actually called for a host name (not very likely,
2763 since this list is usually IP addresses). A host name is now passed as
2766 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2767 code that specifies a non-restarting handler (typically sigaction() in
2768 modern systems) in an attempt to fix a rare and obscure crash bug.
2770 15. Narrowed the window for a race in the daemon that could cause it to ignore
2771 SIGCHLD signals. This is not a major problem, because they are used only to
2772 wake it up if nothing else does.
2774 16. A malformed maildirsize file could cause Exim to calculate negative values
2775 for the mailbox size or file count. Odd effects could occur as a result.
2776 The maildirsize information is now recalculated if the size or filecount
2779 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2780 support for a long time. Removed HAVE_SYS_VFS_H.
2782 18. Updated exipick to current release from John Jetmore.
2784 19. Allow an empty sender to be matched against a lookup in an address list.
2785 Previously the only cases considered were a regular expression, or an
2788 20. Exim went into a mad DNS lookup loop when doing a callout where the
2789 host was specified on the transport, if the DNS lookup yielded more than
2792 21. The RFC2047 encoding function was originally intended for short strings
2793 such as real names; it was not keeping to the 75-character limit for
2794 encoded words that the RFC imposes. It now respects the limit, and
2795 generates multiple encoded words if necessary. To be on the safe side, I
2796 have increased the buffer size for the ${rfc2047: expansion operator from
2799 22. Failure to deliver a bounce message always caused it to be frozen, even if
2800 there was an errors_to setting on the router. The errors_to setting is now
2803 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2804 canonical form (fully expanded) before being placed in
2805 $sender_host_address.
2807 24. Updated eximstats to version 1.33
2809 25. Include certificate and key file names in error message when GnuTLS fails
2810 to set them up, because the GnuTLS error message doesn't include the name
2811 of the failing file when there is a problem reading it.
2813 26. Expand error message when OpenSSL has problems setting up cert/key files.
2816 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2817 (this can affect the format of dates).
2819 28. exim_tidydb, when checking for the continued existence of a message for
2820 which it has found a message-specific retry record, was not finding
2821 messages that were in split spool directories. Consequently, it was
2822 deleting retry records that should have stayed in existence.
2824 29. eximstats updated to version 1.35
2825 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2826 1.35 - bugfix such that pie charts by volume are generated correctly
2828 30. The SPA authentication driver was not abandoning authentication and moving
2829 on to the next authenticator when an expansion was forced to fail,
2830 contradicting the general specification for all authenticators. Instead it
2831 was generating a temporary error. It now behaves as specified.
2833 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2834 (the order specifies the preference for clients). The order is now AES256,
2835 AES128, 3DES, ARCFOUR128.
2837 31. Small patch to Sieve code - explicitly set From: when generating an
2840 32. Exim crashed if a remote delivery caused a very long error message to be
2841 recorded - for instance if somebody sent an entire SpamAssassin report back
2842 as a large number of 550 error lines. This bug was coincidentally fixed by
2843 increasing the size of one of Exim's internal buffers (big_buffer) that
2844 happened as part of the Exiscan merge. However, to be on the safe side, I
2845 have made the code more robust (and fixed the comments that describe what
2848 33. Some experimental protocols are using DNS PTR records for new purposes. The
2849 keys for these records are domain names, not reversed IP addresses. The
2850 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2851 leaves it alone. Component reversal etc. now happens only for IP addresses.
2854 34. The host_aton() function is supposed to be passed a string that is known
2855 to be a valid IP address. However, in the case of IPv6 addresses, it was
2856 not checking this. This is a hostage to fortune. Exim now panics and dies
2857 if the condition is not met. A case was found where this could be provoked
2858 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2859 components; fortuitously, this particular loophole had already been fixed
2860 by change 4.50/55 or 4.44/33 above.
2862 If there are any other similar loopholes, the new check in host_aton()
2863 itself should stop them being exploited. The report I received stated that
2864 data on the command line could provoke the exploit when Exim was running as
2865 exim, but did not say which command line option was involved. All I could
2866 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2867 running as the user.
2870 35. There was a buffer overflow vulnerability in the SPA authentication code
2871 (which came originally from the Samba project). I have added a test to the
2872 spa_base64_to_bits() function which I hope fixes it.
2875 36. The daemon start-up calls getloadavg() while still root for those OS that
2876 need the first call to be done as root, but it missed one case: when
2877 deliver_queue_load_max is set with deliver_drop_privilege. This is
2878 necessary for the benefit of the queue runner, because there is no re-exec
2879 when deliver_drop_privilege is set.
2881 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2882 was in use, was not putting the data itself into the right store pool;
2883 consequently, it could be overwritten for a subsequent message in the same
2884 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2887 38. Sometimes the final signoff response after QUIT could fail to get
2888 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2889 < 0 before doing a fflush(). This bug looks as though it goes back to the
2890 introduction of TLS in release 3.20, but "sometimes" must have been rare
2891 because the tests only now provoked it.
2897 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2898 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2899 Now it returns an integer. A number of calls were still expecting a T/F
2900 return. Fortuitously, in all cases, the tests worked in OK situations,
2901 which is the norm. However, things would have gone wrong on any write
2902 failures on the smtp file descriptor. This function is used when sending
2903 messages over SMTP and also when doing verify callouts.
2905 2. When Exim is called to do synchronous delivery of a locally submitted
2906 message (the -odf or -odi options), it no longer closes stderr before doing
2909 3. Implemented the mua_wrapper option.
2911 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2913 5. Implemented the functions header_remove(), header_testname(),
2914 header_add_at_position(), and receive_remove_recipient(), and exported them
2917 6. If an ACL "warn" statement specified the addition of headers, Exim already
2918 inserted X-ACL-Warn: at the start if there was no header name. However, it
2919 was not making this test for the second and subsequent header lines if
2920 there were newlines in the string. This meant that an invalid header could
2921 be inserted if Exim was badly configured.
2923 7. Allow an ACL "warn" statement to add header lines at the start or after all
2924 the Received: headers, as well as at the end.
2926 8. Added the rcpt_4xx retry error code.
2928 9. Added postmaster_mailfrom=xxx to callout verification option.
2930 10. Added mailfrom=xxxx to the callout verification option, for verify=
2933 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2934 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2936 12. Inserted some casts to stop certain compilers warning when using pointer
2937 differences as field lengths or precisions in printf-type calls (mostly
2938 affecting debugging statements).
2940 13. Added optional readline() support for -be (dynamically loaded).
2942 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2943 same clock tick as a message's arrival, so that its received time was the
2944 same as the "first fail" time on the retry record, and that message
2945 remained on the queue past the ultimate address timeout, every queue runner
2946 would try a delivery (because it was past the ultimate address timeout) but
2947 after another failure, the ultimate address timeout, which should have then
2948 bounced the address, did not kick in. This was a "< instead of <=" error;
2949 in most cases the first failure would have been in the next clock tick
2950 after the received time, and all would be well.
2952 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2953 being recognized when the domain list was tested by the match_domain
2954 condition in an expansion string.
2956 16. Added the ${str2b64: operator.
2958 17. Exim was always calling setrlimit() to set a large limit for the number of
2959 processes, without checking whether the existing limit was already
2960 adequate. (It did check for the limit on file descriptors.) Furthermore,
2961 errors from getrlimit() and setrlimit() were being ignored. Now they are
2962 logged to the main and panic logs, but Exim does carry on, to try to do its
2963 job under whatever limits there are.
2965 18. Imported PCRE 5.0.
2967 19. Trivial typo in log message " temporarily refused connection" (the leading
2970 20. If the log selector return_path_on_delivery was set and an address was
2971 redirected to /dev/null, the delivery process crashed because it assumed
2972 that a return path would always be set for a "successful" delivery. In this
2973 case, the whole delivery is bypassed as an optimization, and therefore no
2976 21. Internal re-arrangement: the function for sending a challenge and reading
2977 a response while authentication was assuming a zero-terminated challenge
2978 string. It's now changed to take a pointer and a length, to allow for
2979 binary data in such strings.
2981 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2983 23. Exim was not respecting finduser_retries when seeking the login of the
2984 uid under which it was called; it was always trying 10 times. (The default
2985 setting of finduser_retries is zero.) Also, it was sleeping after the final
2986 failure, which is pointless.
2988 24. Implemented tls_on_connect_ports.
2990 25. Implemented acl_smtp_predata.
2992 26. If the domain in control=submission is set empty, Exim assumes that the
2993 authenticated id is a complete email address when it generates From: or
2994 Sender: header lines.
2996 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2997 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2998 chown and chgrp in /bin and hostname in /usr/bin.
3000 28. Exim was keeping the "process log" file open after each use, just as it
3001 does for the main log. This opens the possibility of it remaining open for
3002 long periods when the USR1 signal hits a daemon. Occasional processlog
3003 errors were reported, that could have been caused by this. Anyway, it seems
3004 much more sensible not to leave this file open at all, so that is what now
3007 29. The long-running daemon process does not normally write to the log once it
3008 has entered its main loop, and it closes the log before doing so. This is
3009 so that log files can straightforwardly be renamed and moved. However,
3010 there are a couple of unusual error situations where the daemon does write
3011 log entries, and I had neglected to close the log afterwards.
3013 30. The text of an SMTP error response that was received during a remote
3014 delivery was being truncated at 512 bytes. This is too short for some of
3015 the long messages that one sometimes sees. I've increased the limit to
3018 31. It is now possible to make retry rules that apply only when a message has a
3019 specific sender, in particular, an empty sender.
3021 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3022 it possible to be selective about when SMTP synchronization is enforced.
3024 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3026 32. Implemented hosts_connection_nolog.
3028 33. Added an ACL for QUIT.
3030 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3033 35. Added mailbox_size and mailbox_filecount to appendfile.
3035 36. Added control = no_multiline_responses to ACLs.
3037 37. There was a bug in the logic of the code that waits for the clock to tick
3038 in the case where the clock went backwards by a substantial amount such
3039 that the microsecond fraction of "now" was more than the microsecond
3040 fraction of "then" (but the whole seconds number was less).
3042 38. Added support for the libradius Radius client library this is found on
3043 FreeBSD (previously only the radiusclient library was supported).
3049 1. When certain lookups returned multiple values in the form name=value, the
3050 quoting of the values was not always being done properly. Specifically:
3051 (a) If the value started with a double quote, but contained no whitespace,
3053 (b) If the value contained whitespace other than a space character (i.e.
3054 tabs or newlines or carriage returns) it was not quoted.
3055 This fix has been applied to the mysql and pgsql lookups by writing a
3056 separate quoting function and calling it from the lookup code. The fix
3057 should probably also be applied to nisplus, ibase and oracle lookups, but
3058 since I cannot test any of those, I have not disturbed their existing code.
3060 2. A hit in the callout cache for a specific address caused a log line with no
3061 reason for rejecting RCPT. Now it says "Previous (cached) callout
3062 verification failure".
3064 3. There was an off-by-one bug in the queryprogram router. An over-long
3065 return line was truncated at 256 instead of 255 characters, thereby
3066 overflowing its buffer with the terminating zero. As well as fixing this, I
3067 have increased the buffer size to 1024 (and made a note to document this).
3069 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3070 when Exim is waiting for an SMTP response from a remote server, Exim
3071 restarts its select() call on the socket, thereby resetting its timeout.
3072 This is not a problem when such interrupts are rare. Somebody set up a cron
3073 job to run exiwhat every 2 minutes, which is less than the normal select()
3074 timeout (5 or 10 minutes). This meant that the select() timeout never
3075 kicked in because it was always reset. I have fixed this by comparing the
3076 time when an interrupt arrives with the time at the start of the first call
3077 to select(). If more time than the timeout has elapsed, the interrupt is
3078 treated as a timeout.
3080 5. Some internal re-factoring in preparation for the addition of Sieve
3081 extensions (by MH). In particular, the "personal" test is moved to a
3082 separate function, and given an option for scanning Cc: and Bcc: (which is
3083 not set for Exim filters).
3085 6. When Exim created an email address using the login of the caller as the
3086 local part (e.g. when creating a From: or Sender: header line), it was not
3087 quoting the local part when it contained special characters such as @.
3089 7. Installed new OpenBSD configuration files.
3091 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3092 try to make them clearer.
3094 9. Callout options, other than the timeout value, were being ignored when
3095 verifying sender addresses in header lines. For example, when using
3097 verify = header_sender/callout=no_cache
3099 the cache was (incorrectly) being used.
3101 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3102 only the Cygwin environment.
3104 11. When return_path_on_delivery was set as a log selector, if different remote
3105 addresses in the same message used different return paths and parallel
3106 remote delivery occurred, the wrong values would sometimes be logged.
3107 (Whenever a remote delivery process finished, the return path value from
3108 the most recently started remote delivery process was logged.)
3110 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3111 lines when AUTH and/or TLS are in use. This is the "received protocol"
3112 field. Exim used to use "asmtp" for authenticated SMTP, without any
3113 indication (in the protocol name) for TLS use. Now it follows the RFC and
3114 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3115 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3116 names appear in log lines as well as in Received: header lines.
3118 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3119 extensions, and comparison tests, and to fix some bugs.
3121 14. Changes to the "personal" filter test:
3123 (1) The test was buggy in that it was just doing the equivalent of
3124 "contains" tests on header lines. For example, if a user's address was
3125 anne@some.where, the "personal" test would incorrectly be true for
3127 To: susanne@some.where
3129 This test is now done by extracting each address from the header in turn,
3130 and checking the entire address. Other tests that are part of "personal"
3131 are now done using regular expressions (for example, to check local parts
3132 of addresses in From: header lines).
3134 (2) The list of non-personal local parts in From: addresses has been
3135 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3136 taken from the Sieve specification recommendations.
3138 (3) If the message contains any header line starting with "List-" it is
3139 treated as non-personal.
3141 (4) The test for "circular" in the Subject: header line has been removed
3142 because it now seems ill-conceived.
3144 15. Minor typos in src/EDITME comments corrected.
3146 16. Installed latest exipick from John Jetmore.
3148 17. If headers_add on a router specified a text string that was too long for
3149 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3150 of string_sprintf() is now avoided.
3152 18. $message_body_size was not set (it was always zero) when running the DATA
3153 ACL and the local_scan() function.
3155 19. For the "mail" command in an Exim filter, no default was being set for
3156 the once_repeat time, causing a random time value to be used if "once" was
3157 specified. (If the value happened to be <= 0, no repeat happened.) The
3158 default is now 0s, meaning "never repeat". The "vacation" command was OK
3159 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3160 (I found it when inspecting the code).
3162 20. There is now an overall timeout for performing a callout verification. It
3163 defaults to 4 times the callout timeout, which applies to individual SMTP
3164 commands during the callout. The overall timeout applies when there is more
3165 than one host that can be tried. The timeout is checked before trying the
3166 next host. This prevents very long delays if there are a large number of
3167 hosts and all are timing out (e.g. when the network connections are timing
3168 out). The value of the overall timeout can be changed by specifying an
3169 additional sub-option for "callout", called "maxwait". For example:
3171 verify = sender/callout=5s,maxwait=20s
3173 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3174 to the end before writing, but this should make it even safer).
3176 22. Exim was forgetting that it had advertised PIPELINING for the second and
3177 subsequent messages on an SMTP connection. It was also not resetting its
3178 memory on STARTTLS and an internal HELO.
3180 23. When Exim logs an SMTP synchronization error within a session, it now
3181 records whether PIPELINING has been advertised or not.
3183 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3184 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3185 rather than long int.
3187 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3189 26. Added the never_mail option to autoreply.
3195 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3196 crash if the getsockname() call failed; this can happen if a connection is
3197 closed very soon after it is established. The problem was simply in the
3198 order in which certain operations were done, causing Exim to try to write
3199 to the SMTP stream before it had set up the file descriptor. The bug has
3200 been fixed by making things happen in the correct order.
3206 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3207 before the connection was closed, thus losing the rejection response.
3209 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3210 some early Solaris releases, but causes trouble in current releases where
3211 socklen_t is defined.
3213 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3216 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3219 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3220 that releasing the top of it at the end releases what was used for sub-
3221 expansions (unless the block got too big). However, discard this block if
3222 the first thing is a variable or header, so that we can use its block when
3223 it is dynamic (useful for very large $message_headers, for example).
3225 6. Lookups now cache *every* query, not just the most recent. A new, separate
3226 store pool is used for this. It can be recovered when all lookup caches are
3227 flushed. Lookups now release memory at the end of their result strings.
3228 This has involved some general refactoring of the lookup sources.
3230 7. Some code has been added to the store_xxx() functions to reduce the amount
3231 of flapping under certain conditions.
3233 8. log_incoming_interface used to affect only the <= reception log lines. Now
3234 it causes the local interface and port to be added to several more SMTP log
3235 lines, for example "SMTP connection from", and rejection lines.
3237 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3239 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3241 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3242 could be overwritten at the end of the current message (or the start of a
3243 new message if it was set in a HELO ACL). The value is now preserved for
3244 the duration of the SMTP connection.
3246 12. If a transport had a headers_rewrite setting, and a matching header line
3247 contained an unqualified address, that address was qualified, even if it
3248 did not match any rewriting rules. The underlying bug was that the values
3249 of the flags that permit the existence of unqualified sender and recipient
3250 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3251 non-local messages, and by -bnq for local messages) were not being
3252 preserved with the message after it was received.
3254 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3255 "next input=" as part of the text comprising the host identity instead of
3256 the correct text. The code was using the same buffer for two different
3257 strings. However, depending on which order the printing function evaluated
3258 its arguments, the bug did not always show up. Under Linux, for example, my
3259 test suite worked just fine.
3261 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3262 doesn't work with some older versions of Perl. It has been changed to "my",
3263 which in any case is probably the better facility to use.
3265 15. A really picky compiler found some instances of statements for creating
3266 error messages that either had too many or two few arguments for the format
3269 16. The size of the buffer for calls to the DNS resolver has been increased
3270 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3271 for addresses that have a lot of PTR records. This alleviates a problem; it
3272 does not fully solve it.
3274 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3275 buffer now truncates the list and logs the incident, which is the same
3276 action as happens when Exim is looking up a host name and its aliases.
3277 Previously in this situation something unpredictable would happen;
3278 sometimes it was "internal error: store_reset failed".
3280 18. If a server dropped the connection unexpectedly when an Exim client was
3281 using GnuTLS and trying to read a response, the client delivery process
3282 crashed while trying to generate an error log message.
3284 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3285 string, for example:
3287 warn message = H1: something\nH2: something
3289 the text was added as a single header line from Exim's point of view
3290 though it ended up OK in the delivered message. However, searching for the
3291 second and subsequent header lines using $h_h2: did not work. This has been
3292 fixed. Similarly, if a system filter added multiple headers in this way,
3293 the routers could not see them.
3295 20. Expanded the error message when iplsearch is called with an invalid key to
3296 suggest using net-iplsearch in a host list.
3298 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3299 ACLs are no longer actually imposed (and a message to that effect is
3302 22. If a "gecos" field in a passwd entry contained escaped characters, in
3303 particular, if it contained a \" sequence, Exim got it wrong when building
3304 a From: or a Sender: header from that name. A second bug also caused
3305 incorrect handling when an unquoted " was present following a character
3306 that needed quoting.
3308 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3309 was not being matched caselessly.
3311 24. Arranged for all hyphens in the exim.8 source to be escaped with
3314 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3315 properly. Recipient callout cache records were still being keyed to include
3316 the sender, even when use_sender was set false. This led to far more
3317 callouts that were necessary. The sender is no longer included in the key
3318 when use_sender is false.
3320 26. Added "control = submission" modifier to ACLs.
3322 27. Added the ${base62d: operator to decode base 62 numbers.
3324 28. dnsdb lookups can now access SRV records.
3326 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3327 the configuration file.
3329 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3330 (-v) mode. This makes the output for a verbose queue run more intelligible.
3332 31. Added a use_postmaster feature to recipient callouts.
3334 32. Added the $body_zerocount variable, containing the number of binary zero
3335 bytes in the message body.
3337 33. The time of last modification of the "new" subdirectory is now used as the
3338 "mailbox time last read" when there is a quota error for a maildir
3341 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3343 35. Added +ignore_unknown as a special item in host lists.
3345 36. Code for decoding IPv6 addresses in host lists is now included, even if
3346 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3347 address was recognized as an IP address, but was then not correctly decoded
3348 into binary, causing unexpected and incorrect effects when compared with
3355 1. Very minor rewording of debugging text in manualroute to say "list of
3356 hosts" instead of "hostlist".
3358 2. If verify=header_syntax was set, and a header line with an unqualified
3359 address (no domain) and a large number of spaces between the end of the
3360 name and the colon was received, the reception process suffered a buffer
3361 overflow, and (when I tested it) crashed. This was caused by some obsolete
3362 code that should have been removed. The fix is to remove it!
3364 3. When running in the test harness, delay a bit after writing a bounce
3365 message to get a bit more predictability in the log output.
3367 4. Added a call to search_tidyup() just before forking a reception process. In
3368 theory, someone could use a lookup in the expansion of smtp_accept_max_
3369 per_host which, without the tidyup, could leave open a database connection.
3371 5. Added the variables $recipient_data and $sender_data which get set from a
3372 lookup success in an ACL "recipients" or "senders" condition, or a router
3373 "senders" option, similar to $domain_data and $local_part_data.
3375 6. Moved the writing of debug_print from before to after the "senders" test
3378 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3379 problems for message scanning, either using a data ACL, or using
3380 local_scan() because the Received: header was not generated till after they
3381 were called (in order to set the time as the time of reception completion).
3382 I have revised the way this works. The header is now generated after the
3383 body is received, but before the ACL or local_scan() are called. After they
3384 are run, the timestamp in the header is updated.
3390 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3391 before starting a queue runner without re-exec. This happened only when
3392 deliver_drop_privilege was set or when the Exim user was set to root. The
3393 effect of the bug was that timeouts during subsequent deliveries caused
3394 crashes instead of being properly handled. The handler is now left at its
3395 default (and expected) setting.
3397 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3398 message, again when deliver_drop_privilege is set or Exim is run as root.
3399 The bug described in (1) was not present in this case, but the tidying up
3400 of the other signals was missing. I have made the two cases consistent.
3402 3. The ignore_target_hosts setting on a manualroute router was being ignored
3403 for hosts that were looked up using the /MX notation.
3405 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3408 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3409 operated on the sender address. After changing the $sender_address to <>
3410 for the sender address verify, Exim was re-instated it as the original
3411 (before rewriting) address, but remembering that it had rewritten it, so it
3412 wasn't rewriting it again. This bug also had the effect of breaking the
3413 sender address verification caching when the sender address was rewritten.
3415 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3416 This has been changed so that if the ip literal address matches
3417 ignore_target_hosts, the router declines.
3419 7. Added expansion conditions match_domain, match_address, and match_local_
3420 part (NOT match_host).
3422 8. The placeholder for the Received: header didn't have a length field set.
3424 9. Added code to Exim itself and to exim_lock to test for a specific race
3425 condition that could lead to file corruption when using MBX delivery. The
3426 issue is with the lockfile that is created in /tmp. If this file is removed
3427 after a process has opened it but before that process has acquired a lock,
3428 there is the potential for a second process to recreate the file and also
3429 acquire a lock. This could lead to two Exim processes writing to the file
3430 at the same time. The added code performs the same test as UW imapd; it
3431 checks after acquiring the lock that its file descriptor still refers to
3432 the same named file.
3434 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3435 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3438 11. Added the smtp_active_hostname option. If used, this will typically be made
3439 to depend on the incoming interface address. Because $interface_address is
3440 not set up until the daemon has forked a reception process, error responses
3441 that can happen earlier (such as "too many connections") no longer contain
3444 12. If an expansion in a condition on a "warn" statement fails because a lookup
3445 defers, the "warn" statement is abandoned, and the next ACL statement is
3446 processed. Previously this caused the whole ACL to be aborted.
3448 13. Added the iplsearch lookup type.
3450 14. Added ident_timeout as a log selector.
3452 15. Added tls_certificate_verified as a log selector.
3454 16. Added a global option tls_require_ciphers (compare the smtp transport
3455 option of the same name). This controls incoming TLS connections.
3457 17. I finally figured out how to make tls_require_ciphers do a similar thing
3458 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3459 before starting the TLS session.
3461 18. Tabs are now shown as \t in -bP output.
3463 19. If the log selector return_path_on_delivery was set, Exim crashed when
3464 bouncing a message because it had too many Received: header lines.
3466 20. If two routers both had headers_remove settings, and the first one included
3467 a superfluous trailing colon, the final name in the first list and the
3468 first name in the second list were incorrectly joined into one item (with a
3469 colon in the middle).
3475 1. Added -C and -D options to the exinext utility, mainly to make it easier
3476 to include in the automated testing, but these could be helpful when
3477 multiple configurations are in use.
3479 2. The exinext utility was not formatting the output nicely when there was
3480 an alternate port involved in the retry record key, nor when there was a
3481 message id as well (for retries that were specific to a specific message
3482 and a specific host). It was also confused by IPv6 addresses, because of
3483 the additional colons they contain. I have fixed the IPv4 problem, and
3484 patched it up to do a reasonable job for IPv6.
3486 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3487 delivery, the log line now contains "pipelined" if PIPELINING was used.
3489 4. An SMTP transport process used to panic and die if the bind() call to set
3490 an explicit outgoing interface failed. This has been changed; it is now
3491 treated in the same way as a connect() failure.
3493 5. A reference to $sender_host_name in the part of a conditional expansion
3494 that was being skipped was still causing a DNS lookup. This no longer
3497 6. The def: expansion condition was not recognizing references to header lines
3498 that used bh_ and bheader_.
3500 7. Added the _cache feature to named lists.
3502 8. The code for checking quota_filecount in the appendfile transport was
3503 allowing one more file than it should have been.
3505 9. For compatibility with Sendmail, the command line option
3513 and sets the incoming protocol and host name (for trusted callers). The
3514 host name and its colon can be omitted when only the protocol is to be set.
3515 Note the Exim already has two private options, -pd and -ps, that refer to
3516 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3517 "s", but I don't think that's a major issue.
3519 10. A number of refactoring changes to the code, none of which should affect
3522 (a) The number of logging options was getting close to filling up the
3523 32-bit word that was used as a bit map. I have split them into two classes:
3524 those that are passed in the argument to log_write(), and those that are
3525 only ever tested independently outside of that function. These are now in
3526 separate 32-bit words, so there is plenty of room for expansion again.
3527 There is no change in the user interface or the logging behaviour.
3529 (b) When building, for example, log lines, the code previously used a
3530 macro that called string_cat() twice, in order to add two strings. This is
3531 not really sufficiently general. Furthermore, there was one instance where
3532 it was actually wrong because one of the argument was used twice, and in
3533 one call a function was used. (As it happened, calling the function twice
3534 did not affect the overall behaviour.) The macro has been replaced by a
3535 function that can join an arbitrary number of extra strings onto a growing
3538 (c) The code for expansion conditions now uses a table and a binary chop
3539 instead of a serial search (which was left over from when there were very
3540 few conditions). Also, it now recognizes conditions like "pam" even when
3541 the relevant support is not compiled in: a suitably worded error message is
3542 given if an attempt is made to use such a condition.
3544 11. Added ${time_interval:xxxxx}.
3546 12. A bug was causing one of the ddress fields not to be passed back correctly
3547 from remote delivery subprocesses. The field in question was not being
3548 subsequently used, so this caused to problems in practice.
3550 13. Added new log selectors queue_time and deliver_time.
3552 14. Might have fixed a bug in maildirsizefile handling that threw up
3553 "unexpected character" debug warnings, and recalculated the data
3554 unnecessarily. In any case, I expanded the warning message to give more
3557 15. Added the message "Restricted characters in address" to the statements in
3558 the default ACL that block characters like @ and % in local parts.
3560 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3561 Three changes have been made:
3563 (a) There was a serious bug; a negative response to MAIL caused the whole
3564 recipient domain to be cached as invalid, thereby blocking all messages
3565 to all local parts at the same domain, from all senders. This bug has
3566 been fixed. The domain is no longer cached after a negative response to
3567 MAIL if the sender used is not empty.
3569 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3572 (c) A new callout option, "use_sender" has been added for people who want
3573 the modified behaviour.
3579 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3582 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3583 indeed breaks things for older releases.
3585 3. Added additional logging to the case where there is a problem reading data
3586 from a filter that is running in a subprocess using a pipe, in order to
3587 try to track down a specific problem.
3589 4. Testing facility fudge: when running in the test harness and attempting
3590 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3591 getting "No route to host". Convert this to a timeout.
3593 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3596 6. Some OS don't have socklen_t but use size_t instead. This affects the
3597 fifth argument of getsockopt() amongst other things. This is now
3598 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3599 can be set for individual OS. I have set it for SunOS5, OSF1, and
3600 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3601 some earlier ones do not.
3603 7. Change 4.30/15 was not doing the test caselessly.
3605 8. The standard form for an IPv6 address literal was being rejected by address
3606 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3607 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3608 this, as well as the form without the "IPv6" on the front (but only when
3609 address literals are enabled, of course).
3611 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3613 10. Exim crashed if a message with an empty sender address specified by -f
3614 encountered a router with an errors_to setting. This could be provoked only
3615 by a command such as
3619 where an empty string was supplied; "<>" did not hit this bug.
3621 11. Installed PCRE release 4.5.
3623 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3624 remained set. It is now erased.
3626 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3627 times from message ids (which are base 36 rather than the normal 62).
3629 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3630 were being counted as actual protocol errors, and logged if the log
3631 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3632 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3633 and DATA following a set of rejected RCPTs do not count as protocol errors.
3634 In other words, Exim assumes they were pipelined, though this may not
3635 actually be the case. Of course, in all cases the client gets an
3636 appropriate error code.
3638 15. If a lookup fails in an ACL condition, a message about the failure may
3639 be available; it is used if testing the ACL cannot continue, because most
3640 such messages specify what the cause of the deferral is. However, some
3641 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3642 that caused an old message to be retained and used if a later statement
3643 caused a defer, replacing the real cause of the deferral.
3645 16. If an IP address had so many PTR records that the DNS lookup buffer
3646 was not large enough to hold them, Exim could crash while trying to process
3647 the truncated data. It now detects and logs this case.
3649 17. Further to 4.21/58, another change has been made: if (and only if) the
3650 first line of a message (the first header line) ends with CRLF, a bare LF
3651 in a subsequent header line has a space inserted after it, so as not to
3652 terminate the header.
3654 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3655 unnecessarily, used atoi() instead of strtol(), and didn't check the
3656 termination when getting file sizes from file names by regex.
3658 19. Completely re-implemented the support for maildirsize files, in the light
3659 of a number of problems with the previous contributed implementation
3660 (4.30/29). In particular:
3662 . If the quota is zero, the maildirsize file is maintained, but no quota is
3665 . If the maildir directory does not exist, it is created before any attempt
3666 to write a maildirsize file.
3668 . The quota value in the file is just a cache; if the quota is changed in
3669 the transport, the new value overrides.
3671 . A regular expression is available for excluding directories from the
3674 20. The autoreply transport checks the characters in options that define the
3675 message's headers; it allows continued headers, but it was checking with
3676 isspace() after an embedded newline instead of explicitly looking for a
3679 21. If all the "regular" hosts to which an address was routed had passed their
3680 expiry times, and had not reached their retry times, the address was
3681 bounced, even if fallback hosts were defined. Now Exim should go on to try
3684 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3685 equivalent code in the SMTP transport. Some hosts send humungous responses
3686 to HELO/EHLO, more than 1024 it seems.
3688 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3689 gives compiler warnings in some environments. I've now done it "properly",
3692 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3693 (because of problems with the built-in one) was declared to return uschar *
3694 instead of char *, causing compiler failure.
3696 25. Fixed a file descriptor leak when processing alias/forward files.
3698 26. Fixed a minor format string issue in dbfn.c.
3700 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3702 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3703 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3706 29. When a sender address is verified, it is cached, to save repeating the test
3707 when there is more than one recipient in a message. However, when the
3708 verification involves a callout, it is possible for different callout
3709 options to be set for different recipients. It is too complicated to keep
3710 track of this in the cache, so now Exim always runs a verification when a
3711 callout is required, relying on the callout cache for the optimization.
3712 The overhead is duplication of the address routing, but this should not be
3715 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3716 to be verified with callout=postmaster, and the main callout worked but the
3717 postmaster check failed, the verification correctly failed. However, if a
3718 subsequent RCPT command asked for sender verification *without* the
3719 postmaster check, incorrect caching caused this verification also to fail,
3722 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3723 it was not caching the DNS options (qualify_single, search_parents) that
3724 were used when the lookup failed. A subsequent lookup with different
3725 options therefore always gave the same answer, though there were cases
3726 where it should not have. (Example: a "domains = !$mx_any" option on a
3727 dnslookup router: the "domains" option is always processed without any
3728 widening, but the router might have qualify_single set.) Now Exim uses the
3729 cached value only when the same options are set.
3731 32. Added John Jetmore's "exipick" utility to the distribution.
3733 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3734 than a timeout (e.g. a certificate is required, and is not provided), an
3735 Exim server now closes the connection immediately. Previously it waited for
3736 the client to close - but if the client is SSL, it seems that they each
3737 wait for each other, leading to a delay before one of them times out.
3739 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3740 maintained 0.8.x compatibility because I don't think many are using it, and
3741 it is clearly obsolete.
3743 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3746 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3747 client certificate was expired. A simple patch fixes this, though I don't
3748 understand the full logic of why the verify callback is called multiple
3751 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3752 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3753 which causes problems with some clients (such as the Certicom SSL Plus
3754 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3755 disables the coutermeasure allowing Eudora to connect."
3757 38. Exim was not checking that a write() to a log file succeeded. This could
3758 lead to Bad Things if a log got too big, in particular if it hit a file
3759 size limit. Exim now panics and dies if it cannot write to a log file, just
3760 as it does if it cannot open a log file.
3762 39. Modified OS/Makefile-Linux so that it now contains
3764 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3766 The two -D definitions ensure that Exim is compiled with large file
3767 support, which makes it possible to handle log files that are bigger than
3770 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3771 instance) a domain was checked against a named list that involved a lookup,
3772 causing $domain_data to be set, then another domain was checked against the
3773 same list, then the first domain was re-checked, the value of $domain_data
3774 after the final check could be wrong. In particular, if the second check
3775 failed, it could be set empty. This bug probably also applied to
3778 41. The strip_trailing_dot option was not being applied to the address given
3779 with the -f command-line option.
3781 42. The code for reading a message's header from the spool was incrementing
3782 $received_count, but never initializing it. This meant that the value was
3783 incorrect (doubled) while delivering a message in the same process in which
3784 it was received. In the most common configuration of Exim, this never
3785 happens - a fresh exec is done - but it can happen when
3786 deliver_drop_privilege is set.
3788 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3789 it now includes up to 150 characters of the unexpected data in the log
3792 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3793 and building data strings. The size of both of these buffers was 10 000
3794 bytes - far larger than anybody would *ever* want, thought I. Needless to
3795 say, somebody hit the limit. I have increased the maximum line length to
3796 20 000 and the maximum data length of concatenated lines to 100 000. I have
3797 also fixed two bugs, because there was no checking on these buffers. Tsk,
3798 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3799 buffer is too small.
3801 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3802 lsearch lookups. Now it does.
3804 46. When parsing a route_list item in a manualroute router, a fixed-length
3805 buffer was used for the list of hosts. I made this 1024 bytes long,
3806 thinking that nobody would ever have a list of hosts that long. Wrong.
3807 Somebody had a whole pile of complicated expansion conditions, and the
3808 string was silently truncated, leading to an expansion error. It turns out
3809 that it is easier to change to an unlimited length (owing to other changes
3810 that have happened since this code was originally written) than to build
3811 structure for giving a limitation error. The length of the item that
3812 expands into the list of hosts is now unlimited.
3814 47. The lsearch lookup could not handle data where the length of text line was
3815 more than 4095 characters. Such lines were truncated, leading to shortened
3816 data being returned. It should now handle lines of any length.
3818 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3819 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3822 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3823 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3824 no longer happens in comment lines. A list of replacements is now placed
3825 at the head of all of the source files, except those whose only change is
3826 to replace PERL_COMMAND in the very first #! line.
3828 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3829 messages on the queue, with a bottom-up merge sort, using code contributed
3830 by Michael Haardt. This should make operations like -bp somewhat faster on
3831 large queues. It won't affect queue runners, except when queue_run_in_order
3834 51. Installed eximstats 1.31 in the distribution.
3836 52. Added support for SRV lookups to the dnslookup router.
3838 53. If an ACL referred to $message_body or $message_body_end, the value was not
3839 reset for any messages that followed in the same SMTP session.
3841 54. The store-handling optimization for building very long strings was not
3842 differentiating between the different store pools. I don't think this
3843 actually made any difference in practice, but I've tidied it.
3845 55. While running the routers to verify a sender address, $sender_address
3846 was still set to the sender address. This is wrong, because when routing to
3847 send a bounce to the sender, it would be empty. Therefore, I have changed
3848 it so that, while verifying a sender address, $sender_address is set to <>.
3849 (There is no change to what happens when verifying a recipient address.)
3851 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3852 target A or AAAA records (if not already returned) without resetting the
3853 qualify_single or search_parents options of the DNS resolver. These are
3854 inappropriate in this case because the targets of MX and SRV records must
3855 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3856 target that, when qualified, matched something in the local domain. These
3857 two options are now turned off when doing these lookups.
3859 57. It seems that at least some releases of Reiserfs (which does not have the
3860 concept of a fixed number of inodes) returns zero and not -1 for the
3861 number of available inodes. This interacted badly with check_spool_inodes,
3862 which assumed that -1 was the "no such thing" setting. What I have done is
3863 to check that the total number of inodes is greater than zero before doing
3864 the test of how many are available.
3866 58. When a "warn" ACL statement has a log_message modifier, the message is
3867 remembered, and not repeated. This is to avoid a lot of repetition when a
3868 message has many recipients that cause the same warning to be written.
3869 Howewer, Exim was preserving the list of already written lines for an
3870 entire SMTP session, which doesn't seem right. The memory is now reset if a
3871 new message is started.
3873 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3874 debugging output unless log_rewrite was also set.
3876 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3877 of (int)(handle) when we know that handle contains (void *)(-1).
3879 61. The Exim daemon panic-logs an error return when it closes the incoming
3880 connection. However "connection reset by peer" seems to be common, and
3881 isn't really an error worthy of noting specially, so that particular error
3884 62. When Exim is trying to find all the local interfaces, it used to panic and
3885 die if the ioctl to get the interface flags failed. However, it seems that
3886 on at least one OS (Solaris 9) it is possible to have an interface that is
3887 included in the list of interfaces, but for which you get a failure error
3888 for this call. This happens when the interface is not "plumbed" into a
3889 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3890 failure of the "get flags" call assumes that the interface is down.
3892 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3893 makes life easier for people who are doing arithmetic on fields extracted
3894 from dates, where you often get leading zeros that should not be
3895 interpreted as octal.
3897 64. Added qualify_domain to the redirect router, to override the global
3900 65. If a pathologically long header line contained very many addresses (the
3901 report of this problem mentioned 10 000) and each of them was rewritten,
3902 Exim could use up a very large amount of memory. (It kept on making new
3903 copies of the header line as it rewrote, and never released the old ones.)
3904 At the expense of a bit more processing, the header rewriting function has
3905 been changed so that it no longer eats memory in this way.
3907 66. The generation of the Received: header has been moved from the time that a
3908 message starts to be received, to the time that it finishes. The timestamp
3909 in the Received: header should now be very close to that of the <= log
3910 line. There are two side-effects of this change:
3912 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3913 logged header lines no longer include the local Received: line, because
3914 it has not yet been created. The same applies to a copy of the message
3915 that is returned to a non-SMTP sender when a message is rejected.
3917 (b) When a filter file is tested using -bf, no additional Received: header
3918 is added to the test message. After some thought, I decided that this
3921 This change does not affect the value of $received_for. It is still set
3922 after address rewriting, but before local_scan() is called.
3924 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3926 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3927 gave an unhelpful panic error message, and a defer error. I have managed to
3928 change this behaviour so that it now rejects any supplied certificate,
3929 which seems right, as the list of acceptable certificates is empty.
3931 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3932 gave an unhelpful defer error. I have not managed to make this reject any
3933 supplied certificates, but the error message it gives is "no certificate
3934 supplied", which is not helpful.
3936 70. exigrep's output now also includes lines that are not associated with any
3937 message, but which match the given pattern. Implemented by a patch from
3938 Martin Sluka, which also tidied up the Perl a bit.
3940 71. Recipient callout verification, like sender verification, was using <> in
3941 the MAIL FROM command. This isn't really the right thing, since the actual
3942 sender may affect whether the remote host accepts the recipient or not. I
3943 have changed it to use the actual sender in the callout; this means that
3944 the cache record is now keyed on a recipient/sender pair, not just the
3945 recipient address. There doesn't seem to be a real danger of callout loops,
3946 since a callout by the remote host to check the sender would use <>.
3947 [SEE ABOVE: changed after hitting problems.]
3949 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3950 temporary errors. However, in the case of such a code being given after
3951 the end of a data transmission (i.e. after ".") Exim was failing to write
3952 a retry record for the message. (Yes, there was some broken host that was
3953 actually sending 8xx at this point.)
3955 73. An unknown lookup type in a host list could cause Exim to panic-die when
3956 the list was checked. (An example that provoked this was putting <; in the
3957 middle of a list instead of at the start.) If this happened during a DATA
3958 ACL check, a -D file could be left lying around. This kind of configuration
3959 error no longer causes Exim to die; instead it causes a defer errror. The
3960 incident is still logged to the main and panic logs.
3962 74. Buglet left over from Exim 3 conversion. The message "too many messages
3963 in one connection" was written to the rejectlog but not the mainlog, except
3964 when address rewriting (yes!) was being logged.
3966 75. Added write_rejectlog option.
3968 76. When a system filter was run not as root (that is, when system_filter_user
3969 was set), the values of the $n variables were not being returned to the
3970 main process; thus, they were not subsequently available in the $sn
3973 77. Added +return_path_on_delivery log selector.
3975 78. A connection timeout was being treated differently from recipients deferred
3976 when testing hosts_max_try with a message that was older than the host's
3977 retry timeout. (The host should not be counted, thus allowing all hosts to
3978 be tried at least once before bouncing.) This may have been the cause of an
3979 occasionally reported bug whereby a message would remain on the queue
3980 longer than the retry timeout, but would be bounced if a delivery was
3981 forced. I say "may" because I never totally pinned down the problem;
3982 setting up timeout/retry tests is difficult. See also the next item.
3984 79. The ultimate address timeout was not being applied to errors that involved
3985 a combination of host plus message (for example, a timeout on a MAIL
3986 command). When an address resolved to a number of possible hosts, and they
3987 were not all tried for each delivery (e.g. because of hosts_max_try), a
3988 message could remain on the queue longer than the retry timeout.
3990 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3993 81. Fixed an obscure SMTP outgoing bug which required at least the following
3994 conditions: (a) there was another message waiting for the same server;
3995 (b) the server returned 5xx to all RCPT commands in the first message so
3996 that the message was not completed; (c) the server dropped the connection
3997 or gave a negative response to the RSET that Exim sends to abort the
3998 transaction. The observed case was a dropped connection after DATA that had
3999 been sent in pipelining mode. That is, the server had advertised PIPELINING
4000 but was not implementing it correctly. The effect of the bug was incorrect
4001 behaviour, such as trying another host, and this could lead to a crash.
4007 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4008 and daemon.c were passed as pointers to ints; they should have been
4009 pointers to socklen_t variables (which are typically unsigned ints).
4011 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4014 3. Fixed a really odd bug that affected only the testing scheme; patching a
4015 certain fixed string in the binary changed the value of another string that
4016 happened to be identical to the end of the original first string.
4018 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4019 name", it returns that address as the IP address. On some operating
4020 systems (e.g. Solaris), it also passes back the IP address string as the
4021 "host name". However, on others (e.g. Linux), it passes back an empty
4022 string. Exim wasn't checking for this, and was changing the host name to an
4023 empty string, assuming it had been canonicized.
4025 5. Although rare, it is permitted to have more than one PTR record for a given
4026 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4027 all the names associated with an address, because they do in Solaris.
4028 However, it seems that they do not in Linux for data that comes from the
4029 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4030 I found this out when I moved to a new Linux workstation and tried to run
4031 the Exim test suite.
4033 To get round this problem I have changed the code so that it now does its
4034 own call to the DNS to look up PTR records when searching for a host name.
4035 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4036 addresses that are only in /etc/hosts are still found.
4038 This behaviour is, however, controlled by an option called host_lookup_
4039 order, which defaults to "bydns:byaddr". If people want to use the other
4040 order, or indeed, just use one or the other means of lookup, they can
4041 specify it in this variable.
4043 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4044 some operating systems, this comes back from gethostbyaddr() as an empty
4045 string, and this is what Exim used to test for. However, it seems that in
4046 other systems, "." is yielded. Exim now tests for this case too.
4048 7. The values of check_spool_space and check_log_space are now held internally
4049 as a number of kilobytes instead of an absolute number of bytes. If a
4050 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4051 kilobyte. This means that much larger values can be stored.
4053 8. Exim monitor: an attempt to get the action menu when not actually pointing
4054 at a message produces an empty menu entitled "No message selected". This
4055 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4056 no entries in it ("Shell widget menu has zero width and/or height"). So I
4057 have added a single, blank menu entry in this case.
4059 9. Added ${quote_local_part.
4061 10. MIME decoding is now applied to the contents of Subject: header lines when
4064 11. Now that a reference to $sender_host_address automatically causes a reverse
4065 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4066 host lookup before query-style lookups in lists that might use this
4067 variable. This has therefore been abolished, and the "net-" prefix is no
4068 longer necessary for query-style lookups.
4070 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4071 have been a typo for LFLAGS, so it has been changed.
4073 13. The install script calls Exim with "-C /dev/null" in order to find the
4074 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4075 to be output. Howeve, since Exim outputs its version number before the
4076 error, it didn't break the script. It just looked ugly. I fixed this by
4077 always allowing "-C /dev/null" if the caller is root.
4079 14. Ignore overlarge ACL variable number when reading spool file - insurance
4080 against a later release with more variables having written the file.
4082 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4083 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4084 this, as well as the form without the "IPv6" on the front.
4086 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4087 OS/Makefile-Darwin file.
4089 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4090 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4092 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4093 variable. However, some OS use ints for inodes. I've added cast to long int
4094 to get rid of the compiler warning.
4096 19. I had forgotten to lock out "/../" in configuration file names when
4097 ALT_CONFIG_PREFIX was set.
4099 20. Routers used for verification do not need to specify transports. However,
4100 if such a router generated a host list, and callout was configured, Exim
4101 crashed, because it could not find a port number from the (non-existent)
4102 transport. It now assumes port 25 in this circumstance.
4104 21. Added the -t option to exigrep.
4106 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4107 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4108 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4109 them is set, LOOKUP_LSEARCH is forced.
4111 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4113 24. Added sender and host information to the "rejected by local_scan()" log
4114 line; previously there was no indication of these.
4116 25. Added .include_if_exists.
4118 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4119 Linux. It turns out that not all file systems support this. Apparently some
4120 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4121 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4122 not supported on the file descriptor, is now ignored when Exim is trying to
4123 sync a directory. This applies only to Linux.
4125 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4127 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4128 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4131 29. Added support for maildirsize files from supplied patch (modified a bit).
4133 30. The use of :fail: followed by an empty string could lead Exim to respond to
4134 sender verification failures with (e.g.):
4136 550 Verification failed for <xxx>
4137 550 Sender verify failed
4139 where the first response line was missing the '-' that indicates it is not
4140 the final line of the response.
4142 31. The loop for finding the name of the user that called Exim had a hardwired
4143 limit of 10; it now uses the value of finduser_retries, which is used for
4144 all other user lookups.
4146 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4149 33. Exim was neglecting to zero errno before one call of strtol() when
4150 expanding a string and expecting an integer value. On some systems this
4151 resulted in spurious "integer overflow" errors. Also, it was casting the
4152 result into an int without checking.
4154 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4155 did not work. The code looks as if it has *never* worked, though it appears
4156 to have been documented since at least releast 1.62. I have made it work.
4158 35. The "timeout_DNS" error in retry rules, also documented since at least
4159 1.62, also never worked. As it isn't clear exactly what this means, and
4160 clearly it isn't a major issue, I have abolished the feature by treating it
4161 as "timeout", and writing a warning to the main and panic logs.
4163 36. The display of retry rules for -brt wasn't always showing the error code
4166 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4167 timeout_connect_A, timeout_connect_MX.
4169 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4170 to the empty sender.
4172 39. The daemon was not analysing the content of -oX till after it had closed
4173 stderr and disconnected from the controlling terminal. This meant that any
4174 syntax errors were only noted on the panic log, and the return code from
4175 the command was 0. By re-arranging the code a little, I've made the
4176 decoding happen first, so such errors now appear on stderr, and the return
4177 code is 1. However, the actual setting up of the sockets still happens in
4178 the disconnected process, so errors there are still only recorded on the
4181 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4182 connections (as happens on some IP stacks) was logged at start up time as
4183 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4184 it from "IPv6 and IPv4", which means that two separate sockets are being
4187 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4188 says whether AF_INET or AF_INET6 was passed as an argument.
4190 42. Exiwhat output was messed up when time zones were included in log
4193 43. Exiwhat now gives more information about the daemon's listening ports,
4194 and whether -tls-on-connect was used.
4196 44. The "port" option of the smtp transport is now expanded.
4198 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4199 silently ignored. Now an error message is written to the main and panic
4202 46. There's a new ACL modifier called "logwrite" which writes to a log file
4203 as soon as it is encountered.
4205 47. Added $local_user_uid and $local_user_gid at routing time.
4207 48. Exim crashed when trying to verify a sender address that was being
4210 49. Exim was recognizing only a space character after ".include". It now also
4211 recognizes a tab character.
4213 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4214 extracting the relevant information from the specification. The man page no
4215 longer contains scrambled data for the -d option, and I've added a section
4216 at the front about calling Exim under different names.
4218 51. Added "extra_headers" argument to the "mail" command in filter files.
4220 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4223 53. Installed eximstats 1.29.
4225 54. Added transport_filter_timeout as a generic transport option.
4227 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4228 Cc: header lines. This was required by RFC 822, but it not required by RFC
4231 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4232 incoming messages that did not have them. Now it does so only if the
4233 message originates locally, that is, if there is no associated remote host
4234 address. When Resent- header lines are present, this applies to the Resent-
4235 lines rather than the non-Resent- lines.
4237 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4238 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4240 58. Messages for configuration errors now include the name of the main
4241 configuration file - useful now that there may be more than one file in a
4242 list (.included file names were always shown).
4244 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4245 for those rare installations that do not start the daemon as root or run it
4246 setuid root. I've cut out the call to initgroups() if the daemon is not
4249 60. The Exim user and group can now be bound into the binary as text strings
4250 that are looked up at the start of Exim's processing.
4252 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4254 62. Added $mailstore_basename variable.
4256 63. Installed patch to sieve.c from Michael Haardt.
4258 64. When Exim failed to open the panic log after failing to open the main log,
4259 the original message it was trying to log was written to stderr and debug
4260 output, but if they were not available (the usual case in production), it
4261 was lost. Now it is written to syslog before the two lines that record the
4262 failures to open the logs.
4264 65. Users' Exim filters run in subprocesses under the user's uid. It is
4265 possible for a "deliver" command or an alias in a "personal" command to
4266 provoke an address rewrite. If logging of address rewriting is configured,
4267 this fails because the process is not running as root or exim. There may be
4268 a better way of dealing with this, but for the moment (because 4.30 needs
4269 to be released), I have disabled address rewrite logging when running a
4270 filter in a non-root, non-exim process.
4276 1. The buildconfig auxiliary program wasn't quoting the value set for
4277 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4278 not defined. This bug was masked in 4.22 by the effect that was fixed in
4281 2. Some messages that were rejected after a message id was allocated were
4282 shown as "incomplete" by exigrep. It no longer does this for messages that
4283 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4285 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4286 have allow_domain_literals set, the ID did not get logged in the <= line.
4287 Domain literals are now always recognized in Message-ID: header lines.
4289 4. The first argument for a ${extract expansion item is the key name or field
4290 number. Leading and trailing spaces in this item were not being ignored,
4291 causing some misleading effects.
4293 5. When deliver_drop_privilege was set, single queue runner processes started
4294 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4295 same command in the process it spins off) were not dropping privilege.
4297 6. When the daemon running as "exim" started a queue runner, it always
4298 re-executed Exim in the spun-off process. This is a waste of effort when
4299 deliver_drop_privilege is set. The new process now just calls the
4300 queue-runner function directly.
4306 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4309 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4310 ignored. Though the use of -oP was forcing the writing of a pid file, it
4311 was always written to the default place.
4313 3. If the message "no IP address found for host xxxx" is generated during
4314 incoming verification, it is now followed by identification of the incoming
4315 connection (so you can more easily find what provoked it).
4317 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4319 5. Added some features to "harden" Exim a bit more against certain attacks:
4321 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4322 be put in Local/Makefile. This is like the never_users runtime option,
4323 but it cannot be overridden. The default setting is "root".
4325 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4326 prefix string with which any file named in a -C command line option
4329 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4330 is retained for -C and -D only if the caller of Exim is root. Without
4331 it, the exim user may also use -C and -D and retain privilege.
4333 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4334 command line option is disabled.
4336 6. Macro names set by the -D option must start with an upper case letter, just
4337 like macro names defined in the configuration file.
4339 7. Added "dereference=" facility to LDAP.
4341 8. Two instances of the typo "uknown" in the source files are fixed.
4343 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4344 the Configure-Makefile script screwed up while processing it.
4346 10. Incorporated PCRE 4.4.
4348 11. The SMTP synchronization check was not operating right at the start of an
4349 SMTP session. For example, it could not catch a HELO sent before the client
4350 waited for the greeting. There is now a check for outstanding input at the
4351 point when the greeting is written. Because of the duplex, asynchronous
4352 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4353 way, but not yet received, when the check is performed.
4355 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4356 on TCP/IP sockets, because this apparently causes some broken clients to
4359 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4360 unchanged) from the Cygwin maintainer.
4362 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4363 was supported instead of testing for mbx. Effectively a typo.
4365 15. The spa authenticator server code was not checking that the input it
4366 received was valid base64.
4368 16. The debug output line for the "set" modifier in ACLs was not showing the
4369 name of the variable that was being set.
4371 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4373 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4374 lookup on its own. Something else was needed to trigger the lookup. For
4375 example, a match in host_lookup or the need for a host name in a host list.
4376 Now, if $sender_host_name is referenced and the host name has not yet been
4377 looked up, a lookup is performed. If the lookup fails, the variable remains
4378 empty, and $host_lookup_failed is set to "1".
4380 19. Added "eqi" as a case-independent comparison operator.
4382 20. The saslauthd authentication condition could segfault if neither service
4383 nor realm was specified.
4385 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4386 error message that was logged was misleading, and incoming SMTP
4387 connections were dropped. The message is now more accurate, and temporary
4388 errors are given to SMTP connections.
4390 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4391 (except RSET) with a 421 error, until QUIT is received. However, it was
4392 failing to send a response to QUIT.
4394 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4395 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4396 ACL is now run after the helo_try_verify_hosts code.
4398 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4399 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4400 case-independent, so other case variants are also recognized). Apparently
4401 some systems use these upper case variants.
4403 25. If more than two messages were waiting for the same host, and a transport
4404 filter was specified for the transport, Exim sent two messages over the
4405 same TCP/IP connection, and then failed with "socket operation on non-
4406 socket" when it tried to send the third.
4408 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4410 27. The extern definition of crypt16() in expand.c was not being excluded when
4411 the OS had its own crypt16() function.
4413 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4414 as a preferred synonym for return_size_limit, both as an option and as an
4417 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4419 30. Changed the default configuration ACL to relax the local part checking rule
4420 for addresses that are not in any local domains. For these addresses,
4421 slashes and pipe symbols are allowed within local parts, but the sequence
4422 /../ is explicitly forbidden.
4424 31. SPA server authentication was not clearing the challenge buffer before
4427 32. log_message in a "warn" ACL statement was writing to the reject log as
4428 well as to the main log, which contradicts the documentation and doesn't
4429 seem right (because no rejection is happening). So I have stopped it.
4431 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4432 However, I am unable to do any testing of this.
4434 34. Fixed an infelicity in the appendfile transport. When checking directories
4435 for a mailbox, to see if any needed to be created, it was accidentally
4436 using path names with one or more superfluous leading slashes; tracing
4437 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4439 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4440 contents are added to the log line that is written for every discarded
4441 recipient. (Previously a log_message setting was ignored.)
4443 36. The ${quote: operator now quotes the string if it is empty.
4445 37. The install script runs exim in order to find its version number. If for
4446 some reason other than non-existence or emptiness, which it checks, it
4447 could not run './exim', it was installing it with an empty version number,
4448 i.e. as "exim-". This error state is now caught, and the installation is
4451 38. An argument was missing from the function that creates an error message
4452 when Exim fails to connect to the socket for saslauthd authentication.
4453 This could cause Exim to crash, or give a corrupted message.
4455 39. Added isip, isip4, and isip6 to ${if conditions.
4457 40. The ACL variables $acl_xx are now saved with the message, and can be
4458 accessed later in routers, transports, and filters.
4460 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4461 strings in the file are not string-expanded.
4463 42. If a MAIL command specified a SIZE value that was too large to fit into an
4464 int variable, the check against message_size_limit failed. Such values are
4465 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4466 day this will have to be increased, but I don't think I want to be around
4467 when emails are that large.
4474 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4475 iconv() is not standard in FreeBSD.
4477 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4478 IPv6 enabled. The observed symptom was a segmentation fault on return from
4479 the function os_common_find_running_interfaces() in src/os.c.
4481 3. In the check_special_case() function in daemon.c I had used "errno" as an
4482 argument name, which causes warnings on some systems. This was basically a
4483 typo, since it was named "eno" in the comments!
4485 4. The code that waits for the clock to tick (at a resolution of some fraction
4486 of a second) so as to ensure message-id uniqueness was always waiting for
4487 at least one whole tick, when it could have waited for less. [This is
4488 almost certainly not relevant at current processor speeds, where it is
4489 unlikely to ever wait at all. But we try to future-proof.]
4491 5. The function that sleeps for a time interval that includes fractions of a
4492 second contained a race. It did not block SIGALRM between setting the
4493 timer, and suspending (a couple of lines later). If the interval was short
4494 and the sigsuspend() was delayed until after it had expired, the suspension
4495 never ended. On busy systems this could lead to processes getting stuck for
4498 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4499 process, before it forks any delivery processes. The open lookup caching
4500 mechanism meant that the open file or database connection was passed into
4501 the delivery process. The problem was that delivery processes always tidy
4502 up cached lookup data. This could cause a problem for the next delivery
4503 process started by the queue runner, because the external queue runner
4504 process does not know about the closure. So the next delivery process
4505 still has data in the lookup cache. In the case of a file lookup, there was
4506 no problem because closing a file descriptor in a subprocess doesn't affect
4507 the parent. However, if the lookup was caching a connection to a database,
4508 the connection was closed, and the second delivery process was likely to
4509 see errors such as "PGSQL: query failed: server closed the connection
4510 unexpectedly". The problem has been fixed by closing all cached lookups
4511 in a queue runner before running a delivery process.
4513 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4514 seem to have the "const" qualifier which it has on other OS. I've
4517 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4518 *of the same type* (client or server) with the same public name that an
4519 error should be diagnosed.
4521 9. When Exim looked up a host name for an IP address, but failed to find the
4522 original IP address when looking up the host name (a safety check), it
4523 output the message "<ip address> does not match any IP for NULL", which was
4524 confusing, to say the least. The bug was that the host name should have
4525 appeared instead of "NULL".
4527 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4528 user that is built into the binary, and the -C or -D options is used, root
4529 privilege is dropped before the configuration file is read. In addition,
4530 logging is switched to stderr instead of the normal log files. If the
4531 configuration then re-defines the Exim user, the unprivileged environment
4532 is probably not what is expected, so Exim logs a panic warning message (but
4535 However, if deliver_drop_privilege is set, the unprivileged state may well
4536 be exactly what is intended, so the warning has been cut out in that case,
4537 and Exim is allowed to try to write to its normal log files.
4543 1. smtp_return_error_details was not giving details for temporary sender
4544 or receiver verification errors.
4546 2. Diagnose a configuration error if two authenticators have the same public
4549 3. Exim used not to create the message log file for a message until the first
4550 delivery attempt. This could be confusing when incoming messages were held
4551 for policy or load reasons. The message log file is now created at the time
4552 the message is received, and an initial "Received" line is written to it.
4554 4. The automatically generated man page for command line options had a minor
4555 bug that caused no ill effects; however, a more serious problem was that
4556 the procedure for building the man page automatically didn't always
4557 operate. Consequently, release 4.20 contains an out-of-date version. This
4558 shouldn't happen again.
4560 5. When building Exim with embedded Perl support, the script that builds the
4561 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4562 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4564 6. The freeze_tell option was not being used for messages that were frozen on
4565 arrival, either by an ACL or by local_scan().
4567 7. Added the smtp_incomplete_transaction log selector.
4569 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4570 was accepting AUTH without a new EHLO.
4572 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4573 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4576 10. Logging of TCP/IP connections (when configured) now happens in the main
4577 daemon process instead of the child process, so that the TCP/IP connection
4578 count is more accurate (but it can never be perfect).
4580 11. The use of "drop" in a nested ACL was not being handled correctly in the
4581 outer ACL. Now, if condition failure induced by the nested "drop" causes
4582 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4583 or "require"), the connection is dropped.
4585 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4586 that yield "discard" can now be used with an "accept" or a "discard" verb,
4587 but an error is generated for any others (because I can't see a useful way
4588 to define what should happen).
4590 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4591 are now processed in the same way as lines in the Exim configuration file.
4592 In particular, continuation lines are supported.
4594 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4596 15. Added -ti meaning -t -i.
4598 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4599 domains, and warn by logging if others are found.
4601 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4602 returned from GIFCONF when Exim is trying to find the list of interfaces on
4603 a host. The code in os.c has been modified to copy each ifreq to an aligned
4604 structure in all cases.
4606 Also, in some cases, the returned ifreq's were being copied to a 'struct
4607 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4608 means the last couple of bytes of an IPv6 address could be chopped if the
4609 ifreq contained only a normal sockaddr (14 bytes storage).
4611 18. Named domain lists were not supported in the hosts_treat_as_local option.
4612 An entry such as +xxxx was not recognized, and was treated as a literal
4615 19. Ensure that header lines added by a DATA ACL are included in the reject log
4616 if the ACL subsequently rejects the message.
4618 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4619 MD5 (which is deprecated).
4621 21. When testing a filter file using -bf, Exim was writing a message when it
4622 took the sender from a "From " line in the message, but it was not doing so
4623 when it took $return_path from a Return-Path: header line. It now does.
4625 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4626 with a valid header line field name (a series of printing characters
4627 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4629 23. Changed "disc" in the source to "disk" to conform to the documentation and
4630 the book and for uniformity.
4632 24. Ignore Sendmail's -Ooption=value command line item.
4634 25. When execve() failed while trying to run a command in a pipe transport,
4635 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4636 could be confused with a return value of 69 from the command itself. This
4637 has been changed to 127, the value the shell returns if it is asked to run
4638 a non-existent command. The wording for the related log line suggests a
4639 non-existent command as the problem.
4641 26. If received_header_text expands to an empty string, do not add a Received:
4642 header line to the message. (Well, it adds a token one on the spool, but
4643 marks it "old" so that it doesn't get used or transmitted.)
4645 27. Installed eximstats 1.28 (addition of -nt option).
4647 28. There was no check for failure on the call to getsockname() in the daemon
4648 code. This can fail if there is a shortage of resources on the system, with
4649 ENOMEM, for example. A temporary error is now given on failure.
4651 29. Contrary to the C standard, it seems that in some environments, the
4652 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4653 program. Exim now does this explicitly; it affects the formatting of
4654 timestamps using strftime().
4656 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4657 complaints. I've now initialized all the variables, to avoid this.
4659 32. Header lines added by a system filter were not being "seen" during
4660 transport-time rewrites.
4662 33. The info_callback() function passed to OpenSSL is set up with type void
4663 (*)(SSL *, int, int), as described somewhere. However, when calling the
4664 function (actually a macro) that sets it up, the type void(*)() is
4665 expected. I've put in a cast to prevent warnings from picky compilers.
4667 34. If a DNS black list lookup found a CNAME record, but there were no A
4668 records associated with the domain it pointed at, Exim crashed.
4670 35. If a DNS black list lookup returned more than one A record, Exim ignored
4671 all but the first. It now scans all returned addresses if a particular IP
4672 value is being sought. In this situation, the contents of the
4673 $dnslist_value variable are a list of all the addresses, separated by a
4676 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4677 to accept a host name and all its aliases if the forward lookup for any of
4678 them yielded the IP address of the incoming connection. Now it accepts only
4679 those names whose forward lookup yields the correct IP address. Any other
4680 names are discarded. This closes a loophole whereby a rogue DNS
4681 administrator could create reverse DNS records to break through a
4682 wildcarded host restriction in an ACL.
4684 37. If a user filter or a system filter that ran in a subprocess used any of
4685 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4686 the wrong values were passed to the pipe command ($thisaddress had the
4687 value of $0, $0 had the value of $1, etc). This bug was introduced by
4688 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4690 38. Improved the line breaking for long SMTP error messages from ACLs.
4691 Previously, if there was no break point between 40 and 75 characters, Exim
4692 left the rest of the message alone. Two changes have been made: (a) I've
4693 reduced the minimum length to 35 characters; (b) if it can't find a break
4694 point between 35 and 75 characters, it looks ahead and uses the first one
4695 that it finds. This may give the occasional overlong line, but at least the
4696 remaining text gets split now.
4698 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4699 file descriptors might be low, and that setting 1000 would always raise it.
4700 It turns out that in some environments, the limit is already over 1000 and
4701 that lowering it causes trouble. So now Exim takes care not to decrease it.
4703 40. When delivering a message, the value of $return_path is set to $sender_
4704 address at the start of routing (routers may change the value). By an
4705 oversight, this default was not being set up when an address was tested by
4706 -bt or -bv, which affected the outcome if any router or filter referred to
4709 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4710 log only when "message" or "log_message" are set. However, if one of the
4711 conditions was an address verification, or a call to a nested ACL, the
4712 messages generated by the underlying test were being passed through. This
4713 no longer happens. The underlying message is available in $acl_verify_
4714 message for both "message" and "log_message" expansions, so it can be
4715 passed through if needed.
4717 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4718 new expansion $bh_ to give the encoded byte string without charset
4719 translation. Translation happens only if iconv() is available; HAVE_ICONV
4720 indicates this at build time. HEADERS_CHARSET gives the charset to
4721 translate to; headers_charset can change it in the configuration, and
4722 "headers charset" can change it in an individual filter file.
4724 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4725 that creates RFC 2047 encoded "words" labels them as that charset instead
4726 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4727 expansion operator; (ii) when Exim creates a From: line for a local
4728 message; (iii) when a header line is rewritten to include a "phrase" part.
4730 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4731 buggy, causing it to skip the first lines of messages whose message ID
4732 ended in 'D'. This would not have bitten before Exim release 4.14, because
4733 message IDs were unlikely to end in 'D' before then. The effect was to have
4734 incorrect size information for certain domains.
4736 45. #include "config.h" was missing at the start of the crypt16.c module. This
4737 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4740 46. If there was a timeout during a "random" callout check, Exim treated it as
4741 a failure of the random address, and carried on sending RSET and the real
4742 address. If the delay was just some slowness somewhere, the response to the
4743 original RCPT would be taken as a response to RSET and so on, causing
4744 mayhem of various kinds.
4746 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4747 when I implemented it. It didn't allow for the fact that some option values
4748 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4749 the right test for positive values.
4751 48. Domain names in DNS records are case-independent. Exim always looks them up
4752 in lower case. Some resolvers return domain names in exactly the case they
4753 appear in the zone file, that is, they may contain uppercase letters. Not
4754 all resolvers do this - some return always lower case. Exim was treating a
4755 change of case by a resolver as a change of domain, similar to a widening
4756 of a domain abbreviation. This triggered its re-routing code and so it was
4757 trying to route what was effectively the same domain again. This normally
4758 caused routing to fail (because the router wouldn't handle the domain
4759 twice). Now Exim checks for this case specially, and just changes the
4760 casing of the domain that it ultimately uses when it transmits the message
4763 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4766 50. If a filter generated a file delivery with a non-absolute name (possible if
4767 no home directory exists for the router), the forbid_file option was not
4770 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4771 the existing equality matching.
4773 52. Exim was using ints instead of ino_t variables in some places where it was
4774 dealing with inode numbers.
4776 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4777 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4778 called TMPDIR, and if it finds it is different, it changes its value.
4780 54. The smtp_printf() function is now made available to local_scan() so
4781 additional output lines can be written before returning. There is also an
4782 smtp_fflush() function to enable the detection of a dropped connection.
4783 The variables smtp_input and smtp_batched_input are exported to
4786 55. Changed the default runtime configuration: the message "Unknown user"
4787 has been removed from the ACL, and instead placed on the localuser router,
4788 using the cannot_route_message feature. This means that any verification
4789 failures that generate their own messages won't get overridden. Similarly,
4790 the "Unrouteable address" message that was in the ACL for unverifiable
4791 relay addresses has also been removed.
4793 56. Added hosts_avoid_esmtp to the smtp transport.
4795 57. The exicyclog script was not checking for the esoteric option
4796 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4797 will work only if exicyclog is run under the appropriate euid.
4799 58. Following a discussion on the list, the rules by which Exim recognises line
4800 endings on incoming messages have been changed. The -dropcr and drop_cr
4801 options are now no-ops, retained only for backwards compatibility. The
4802 following line terminators are recognized: LF CRLF CR. However, special
4803 processing applies to CR:
4805 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4806 nor a local message in the state where . is a terminator.
4808 (ii) If a bare CR is encountered in a header line, an extra space is added
4809 after the line terminator so as not to end the header. The reasoning
4810 behind this is that bare CRs in header lines are most likely either
4811 to be mistakes, or people trying to play silly games.
4813 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4814 was being incorrectly given as 18 bytes larger than it should have been.
4815 This is a VOB (very old bug).
4817 60. This may never have affected anything current, but just in case it has:
4818 When the local host is found other than at the start of a list of hosts,
4819 the local host, those with the same MX, and any that follow, are discarded.
4820 When the list in question was part of a longer list of hosts, the following
4821 hosts (not currently being processed) were also being discarded. This no
4822 longer happens. I'm not sure if this situation could ever has previously
4825 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4828 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4829 header. This is something that is recommended in a new Internet Draft, and
4830 is something that is documented as being done by Sendmail. There are two
4831 possible values. For messages generated by the autoreply transport, Exim
4834 Auto-Submitted: auto-replied
4836 whereas for all other generated messages (e.g. bounces) it adds
4838 Auto-Submitted: auto-generated
4840 63. The "personal" condition in filters now includes a test for the
4841 Auto-Submitted: header. If it contains the string "auto-" the message it
4842 not considered personal.
4844 64. Added rcpt_include_affixes as a generic transport option.
4846 65. Added queue_only_override (default true).
4848 66. Added the syslog_duplication option.
4850 67. If what should have been the first header line of a message consisted of
4851 a space followed by a colon, Exim was mis-interpreting it as a header line.
4852 It isn't of course - it is syntactically invalid and should therefore be
4853 treated as the start of the message body. The misbehaviour could have
4854 caused a number of strange effects, including loss of data in subsequent
4855 header lines, and spool format errors.
4857 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4858 client host had authenticated. This control can now be exercised by an ACL
4859 for more flexibility.
4861 69. By default, callouts do not happen when testing with -bh. There is now a
4862 variant, -bhc, which does actually run the callout code, including
4863 consulting and updating the callout cache.
4865 70. Added support for saslauthd authentication, courtesy of Alexander
4868 71. If statvfs() failed on the spool or log directories while checking their
4869 size for availability, Exim confusingly gave the error "space shortage".
4870 Furthermore, in debugging mode it crashed with a floating point exception.
4871 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4872 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4873 serious problem, Exim now writes to the main and panic logs when this
4874 happens, with details of the failure. It then refuses to accept the
4875 incoming message, giving the message "spool directory problem" or "log
4876 directory problem" with a 421 code for SMTP messages.
4878 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4879 0, 1, and 2 exist, because some OS complain for execs without them (see
4880 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4881 descriptors. However, the code omitted to check that the open succeeded,
4882 causing mysterious errors if for some reason the permissions on /dev/null
4883 got screwed. Now Exim writes a message to the main and panic logs, and
4884 bombs out if it can't open /dev/null.
4886 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4887 interact so that it is all more flexible. It is supposed to remain
4888 backwards compatible. Also added extra_local_interfaces.
4890 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4891 to bomb out with an assertion failure - to the client this appears as a
4892 connection drop. This problem occurs in the part of the code that was taken
4893 from the Samba project. Fortunately, the assertion is in a very simple
4894 function, so I have fixed this by reproducing the function inline in the
4895 one place where it is called, and arranging for authentication to fail
4896 instead of killing the process with assert().
4898 75. The SPA client code was not working when the server requested OEM rather
4899 than Unicode encoding.
4901 76. Added code to make require_files with a specific uid setting more usable in
4902 the case where statting the file as root fails - usually a non-root-mounted
4903 NFS file system. When this happens and the failure is EACCES, Exim now
4904 forks a subprocess and does the per-uid checking as the relevant uid.
4906 77. Added process_log_path.
4908 78. If log_file_path was not explicitly set, a setting of check_log_space or
4909 check_log_inodes was ignored.
4911 79. If a space check for the spool or log partitions fails, the incident is now
4912 logged. Of course, in the latter case the data may get lost...
4914 80. Added the %p formatting code to string_format() so that it can be used to
4915 print addresses in debug_print(). Adjusted all the address printing in the
4916 debugging in store.c to use %p rather than %d.
4918 81. There was a concern that a line of code in smtp_in.c could overflow a
4919 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4920 initially expressed, the concern was not well-founded, because trailing
4921 spaces are removed early. However, if the trailing spaces were followed by
4922 a NULL, they did not get removed, so the overflow was possible. Two fixes
4925 (a) I re-wrote the offending code in a cleaner fashion.
4926 (b) If an incoming SMTP command contains a NULL character, it is rejected
4929 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4930 runs initgroups(), so that if the Exim user is in any additional groups,
4931 they will be used during message reception.
4937 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob