1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.556 2008/10/16 07:57:01 nm4 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TK/01 Added patch by Johannes Berg that expands the main option
10 "spamd_servers" if it starts with a dollar sign.
12 TK/02 Write list of recipients to X-Envelope-Sender header when building
13 the mbox-format spool file for content scanning (suggested by Jakob
16 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
17 (http://www.dcc-servers.net/) support via dccifd. Activated by
18 setting EXPERIMENTAL_DCC=yes in Local/Makefile. Check out
19 experimental_spec.txt for more documentation.
21 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
22 by Mark Daniel Reidel <mr@df.eu>.
24 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
25 When building exim an external PCRE library is now needed -
26 PCRE is a system library on the majority of modern systems.
27 See entry on PCRE_LIBS in EDITME file.
29 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
30 conversation. Added nologin parameter to request.
31 Patch contributed by Kirill Miazine
33 TF/01 Do not log submission mode rewrites if they do not change the address.
35 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
37 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
38 log files in place. Contributed by Roberto Lima
40 NM/04 Bugzilla 667: close socket used by dovecot authenticator
42 TF/03 Bugzilla 615: When checking the local_parts router precondition
43 after a local_part_suffix or local_part_prefix option, Exim now
44 does not use the address's named list lookup cache, since this
45 contains cached lookups for the whole local part.
47 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
48 Robert Millan. Documentation is in experimental-spec.txt
50 TF/04 Bugzilla 668: Fix parallel build (make -j).
52 NM/05 Bugzilla 437: Prevent Maildix aux files being created with mode 000
54 NM/05 Bugzilla 598: Improvement to Dovecot authenticator handling.
55 Patch provided by Jan Srzednicki
57 TF/05 Leading white space used to be stripped from $spam_report which
58 wrecked the formatting. Now it is preserved.
60 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
61 that they are available at delivery time.
63 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
65 TF/08 TLS error reporting now respects the incoming_interface and
66 incoming_port log selectors.
68 TF/09 Produce a more useful error message if an SMTP transport's hosts
69 setting expands to an empty string.
71 NM/06 Bugzilla 744: EXPN did not work under TLS.
72 Patch provided by Phil Pennock
74 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
75 Patch provided by Richard Godbee
77 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
84 TK/01 Add preliminary DKIM support. Currently requires a forked version of
85 ALT-N's libdkim that I have put here:
86 http://duncanthrax.net/exim-experimental/
88 Note to Michael Haardt: I had to rename some vars in sieve.c. They
89 were called 'true' and it seems that C99 defines that as a reserved
90 keyword to be used with 'bool' variable types. That means you could
91 not include C99-style headers which use bools without triggering
92 build errors in sieve.c.
94 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
95 as mailq or other aliases. Changed the --help handling significantly
96 to do whats expected. exim_usage() emits usage/help information.
98 SC/01 Added the -bylocaldomain option to eximstats.
100 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
102 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
104 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
110 PH/01 Another patch from the Sieve maintainer.
112 PH/02 When an IPv6 address is converted to a string for single-key lookup
113 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
114 dots are used instead of colons so that keys in lsearch files need not
115 contain colons. This was done some time before quoting was made available
116 in lsearch files. However, iplsearch files do require colons in IPv6 keys
117 (notated using the quote facility) so as to distinguish them from IPv4
118 keys. This meant that lookups for IP addresses in host lists did not work
119 for iplsearch lookups.
121 This has been fixed by arranging for IPv6 addresses to be expressed with
122 colons if the lookup type is iplsearch. This is not incompatible, because
123 previously such lookups could never work.
125 The situation is now rather anomolous, since one *can* have colons in
126 ordinary lsearch keys. However, making the change in all cases is
127 incompatible and would probably break a number of configurations.
129 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
132 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
133 conversion specification without a maximum field width, thereby enabling
134 a rogue spamd server to cause a buffer overflow. While nobody in their
135 right mind would setup Exim to query an untrusted spamd server, an
136 attacker that gains access to a server running spamd could potentially
137 exploit this vulnerability to run arbitrary code as the Exim user.
139 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
140 $primary_hostname instead of what libspf2 thinks the hosts name is.
142 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
143 a directory entry by the name of the lookup key. Previously, if a
144 symlink pointed to a non-existing file or a file in a directory that
145 Exim lacked permissions to read, a lookup for a key matching that
146 symlink would fail. Now it is enough that a matching directory entry
147 exists, symlink or not. (Bugzilla 503.)
149 PH/03 The body_linecount and body_zerocount variables are now exported in the
152 PH/04 Added the $dnslist_matched variable.
154 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
155 This means they are set thereafter only if the connection becomes
158 PH/06 Added the client_condition to authenticators so that some can be skipped
159 by clients under certain conditions.
161 PH/07 The error message for a badly-placed control=no_multiline_responses left
162 "_responses" off the end of the name.
164 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
166 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
167 (without spaces) instead of just copying the configuration text.
169 PH/10 Added the /noupdate option to the ratelimit ACL condition.
171 PH/11 Added $max_received_linelength.
173 PH/12 Added +ignore_defer and +include_defer to host lists.
175 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
176 way in which PCRE > 7.0 is built.
178 PH/14 Implemented queue_only_load_latch.
180 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
181 MAIL command. The effect was to mangle the value on 64-bit systems.
183 PH/16 Another patch from the Sieve maintainer.
185 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
187 PH/18 If a system quota error occurred while trying to create the file for
188 a maildir delivery, the message "Mailbox is full" was not appended to the
189 bounce if the delivery eventually timed out. Change 4.67/27 below applied
190 only to a quota excession during the actual writing of the file.
192 PH/19 It seems that peer DN values may contain newlines (and other non-printing
193 characters?) which causes problems in log lines. The DN values are now
194 passed through string_printing() before being added to log lines.
196 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
197 and InterBase are left for another time.)
199 PH/21 Added message_body_newlines option.
201 PH/22 Guard against possible overflow in moan_check_errorcopy().
203 PH/23 POSIX allows open() to be a macro; guard against that.
205 PH/24 If the recipient of an error message contained an @ in the local part
206 (suitably quoted, of course), incorrect values were put in $domain and
207 $local_part during the evaluation of errors_copy.
213 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
214 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
217 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
218 issue a MAIL command.
220 PH/02 In an ACL statement such as
222 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
224 if a client was not listed at all, or was listed with a value other than
225 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
226 the condition was not true (as it should be), so access was not denied.
227 The bug was that the ! inversion was incorrectly passed on to the second
228 item. This has been fixed.
230 PH/03 Added additional dnslists conditions == and =& which are different from
231 = and & when the dns lookup returns more than one IP address.
233 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
234 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
236 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
237 FSYNC, which compiles an option called disable_fsync that allows for
238 bypassing fsync(). The documentation is heavily laced with warnings.
240 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
242 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
243 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
244 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
245 including adding "make clean"; (3) Added -fPIC when compiling the test
246 dynamically loaded module, to get rid of a warning.
248 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
249 message fails, move_frozen_messages = true and ignore_bounce_errors_after
250 = 0s. The bug is otherwise harmless.
252 PH/07 There was a bug in the dovecot authenticator such that the value of
253 $auth1 could be overwritten, and so not correctly preserved, after a
254 successful authentication. This usually meant that the value preserved by
255 the server_setid option was incorrect.
257 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
259 PH/09 Installed PCRE release 7.0.
261 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
262 run for batched SMTP input. It is now run at the start of every message
263 in the batch. While fixing this I discovered that the process information
264 (output by running exiwhat) was not always getting set for -bs and -bS
265 input. This is fixed, and it now also says "batched" for BSMTP.
267 PH/11 Added control=no_pipelining.
269 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
270 patch, slightly modified), and move the expansion of helo_data till after
271 the connection is made in the smtp transport (so it can use these
274 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
276 PH/14 Added log_selector = +pid.
278 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
280 PH/16 Add ${if forany and ${if forall.
282 PH/17 Added dsn_from option to vary the From: line in DSNs.
284 PH/18 Flush SMTP output before performing a callout, unless control =
285 no_callout_flush is set.
287 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
288 was true (the default) a successful delivery failed to delete the retry
289 item, thus causing premature timeout of the address. The bug is now
292 PH/20 Added hosts_avoid_pipelining to the smtp transport.
294 PH/21 Long custom messages for fakedefer and fakereject are now split up
295 into multiline reponses in the same way that messages for "deny" and
296 other ACL rejections are.
298 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
299 with slight modification.
301 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
302 draft, changing the syntax and factoring some duplicate code".
304 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
305 for deliveries of the second and subsequent messages over the same SMTP
308 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
309 ${reduce, with only minor "tidies".
311 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
313 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
314 expansion side effects.
316 PH/27 When a message times out after an over-quota error from an Exim-imposed
317 quota, the bounce message says "mailbox is full". This message was not
318 being given when it was a system quota that was exceeded. It now should
321 MH/03 Made $recipients available in local_scan(). local_scan() already has
322 better access to the recipient list through recipients_list[], but
323 $recipients can be useful in postmaster-provided expansion strings.
325 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
326 in the case of a MAIL command with additional options following the
327 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
328 were accidentally chopped off.
330 PH/29 SMTP synchronization checks are implemented when a command is read -
331 there is a check that no more input is waiting when there shouldn't be
332 any. However, for some commands, a delay in an ACL can mean that it is
333 some time before the response is written. In this time, more input might
334 arrive, invalidly. So now there are extra checks after an ACL has run for
335 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
336 pipelining has not been advertised.
338 PH/30 MH's patch to allow iscntrl() characters to be list separators.
340 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
341 returned in the SMTP response when smtp_return_error_details was false.
344 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
345 instead of the C I/O that was originally supplied, because problems were
348 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
349 Exim which did not show up earlier: it was assuming that a call to
350 SSL_CTX_set_info_callback() might give an error value. In fact, there is
351 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
352 was a macro that became an assignment, so it seemed to work. This has
353 changed to a proper function call with a void return, hence the compile
354 error. Exim's code has been fixed.
356 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
359 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
361 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
363 PH/37 If a message is not accepted after it has had an id assigned (e.g.
364 because it turns out to be too big or there is a timeout) there is no
365 "Completed" line in the log. When some messages of this type were
366 selected by exigrep, they were listed as "not completed". Others were
367 picked up by some special patterns. I have improved the selection
368 criteria to be more general.
370 PH/38 The host_find_failed option in the manualroute router can now be set
371 to "ignore", to completely ignore a host whose IP address cannot be
372 found. If all hosts are ignored, the behaviour is controlled by the new
373 host_all_ignored option.
375 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
376 homing or because of multiple MX records with /mx) generated more than
377 one IP address, and the following item turned out to be the local host,
378 all the secondary addresses of the first item were incorrectly removed
379 from the list, along with the local host and any following hosts (which
380 is what is supposed to happen).
382 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
383 whoever called Exim into the -H file. In the case of the daemon it was
384 behaving confusingly. When first started, it used values for whoever
385 started the daemon, but after a SIGHUP it used the Exim user (because it
386 calls itself on a restart). I have changed the code so that it now always
389 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
390 message are rejected with the same error (e.g. no authentication or bad
391 sender address), and a DATA command is nevertheless sent (as can happen
392 with PIPELINING or a stupid MUA), the error message that was given to the
393 RCPT commands is included in the rejection of the DATA command. This is
394 intended to be helpful for MUAs that show only the final error to their
397 PH/42 Another patch from the Sieve maintainer.
399 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
400 Eximstats - Fixed some broken HTML links and added missing column headers
402 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
403 columns for Rejects, Temp Rejects, Ham, and Spam rows.
405 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
407 PH/43 Yet another patch from the Sieve maintainer.
409 PH/44 I found a way to check for a TCP/IP connection going away before sending
410 the response to the final '.' that terminates a message, but only in the
411 case where the client has not sent further data following the '.'
412 (unfortunately, this is allowed). However, in many cases there won't be
413 any further data because there won't be any more messages to send. A call
414 to select() can be used: if it shows that the input is "ready", there is
415 either input waiting, or the socket has been closed. An attempt to read
416 the next input character can distinguish the two cases. Previously, Exim
417 would have sent an OK response which the client would never have see.
418 This could lead to message repetition. This fix should cure that, at
419 least in a lot of common cases.
421 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
422 advertised in response to EHLO.
428 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
429 fixed by 4.65/MH/01 (is this a record?) are fixed:
431 (i) An empty string was always treated as zero by the numeric comparison
432 operators. This behaviour has been restored.
434 (ii) It is documented that the numeric comparison operators always treat
435 their arguments as decimal numbers. This was broken in that numbers
436 starting with 0 were being interpreted as octal.
438 While fixing these problems I realized that there was another issue that
439 hadn't been noticed. Values of message_size_limit (both the global option
440 and the transport option) were treated as octal if they started with 0.
441 The documentation was vague. These values are now always treated as
442 decimal, and I will make that clear in the documentation.
448 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
449 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
452 MH/01 Don't check that the operands of numeric comparison operators are
453 integers when their expansion is in "skipping" mode (fixes bug
454 introduced by 4.64-PH/07).
456 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
457 child addresses, Exim now panics and dies. Previously, because the count
458 is held in a short int, deliveries were likely to be lost. As such a
459 large number of recipients for a single message is ridiculous
460 (performance will be very, very poor), I have chosen to impose a limit
461 rather than extend the field.
467 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
468 leftover -K file (the existence of which was triggered by #402).
469 While we were at it, introduced process PID as part of the -K
470 filename. This should rule out race conditions when creating
473 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
474 processing considerably. Previous code took too long for large mails,
475 triggering a timeout which in turn triggers #401.
477 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
478 in the DK code in transports.c. sendfile() is not really portable,
479 hence the _LINUX specificness.
481 TF/01 In the add_headers option to the mail command in an Exim filter,
482 there was a bug that Exim would claim a syntax error in any
483 header after the first one which had an odd number of characters
486 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
487 callout verification, Exim cached a "reject" for the entire domain. This
488 is correct for most verifications, but it is not correct for a recipient
489 verification with use_sender or use_postmaster set, because in that case
490 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
491 case of MAIL FROM:<> rejection from other early rejections (e.g.
492 rejection of HELO). When verifying a recipient using a non-null MAIL
493 address, the cache is ignored if it shows MAIL FROM:<> rejection.
494 Whatever the result of the callout, the value of the domain cache is
495 left unchanged (for any other kind of callout, getting as far as trying
496 RCPT means that the domain itself is ok).
498 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
501 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
502 manifest itself as EPIPE rather than ECONNECT. When tidying away a
503 session, the daemon ignores ECONNECT errors and logs others; it now
504 ignores EPIPE as well.
506 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
507 (quoted-printable decoding).
509 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
510 later the small subsequent patch to fix an introduced bug.
512 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
514 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
516 PH/08 An error is now given if message_size_limit is specified negative.
518 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
519 to be given (somewhat) arbitrary names.
521 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
524 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
525 miscellaneous code fixes
527 PH/10 Added the log_reject_target ACL modifier to specify where to log
530 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
531 hostname. This is wrong, because it relates to the incoming message (and
532 probably the interface on which it is arriving) and not to the outgoing
533 callout (which could be using a different interface). This has been
534 changed to use the value of the helo_data option from the smtp transport
535 instead - this is what is used when a message is actually being sent. If
536 there is no remote transport (possible with a router that sets up host
537 addresses), $smtp_active_hostname is used.
539 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
540 tweaks were necessary in order to get it to work (see also 21 below):
541 (a) The code assumed that strncpy() returns a negative number on buffer
542 overflow, which isn't the case. Replaced with Exim's string_format()
544 (b) There were several signed/unsigned issues. I just did the minimum
545 hacking in of casts. There is scope for a larger refactoring.
546 (c) The code used strcasecmp() which is not a standard C function.
547 Replaced with Exim's strcmpic() function.
548 (d) The code set only $1; it now sets $auth1 as well.
549 (e) A simple test gave the error "authentication client didn't specify
550 service in request". It would seem that Dovecot has changed its
551 interface. Fortunately there's a specification; I followed it and
552 changed what the client sends and it appears to be working now.
554 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
557 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
558 address A is aliased to B and C, where B exists and C does not. Without
559 -v the output is "A verified" because verification stops after a
560 successful redirection if more than one address is generated. However,
561 with -v the child addresses are also verified. Exim was outputting "A
562 failed to verify" and then showing the successful verification for C,
563 with its parentage. It now outputs "B failed to verify", showing B's
564 parentage before showing the successful verification of C.
566 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
567 look up a TXT record in a specific list after matching in a combined
570 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
571 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
572 they consult the DNS. I had assumed they would set it the way they
573 wanted; and indeed my experiments on Linux seem to show that in some
574 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
575 To be on the safe side, however, I have now made the interface to
576 host_find_byname() similar to host_find_bydns(), with an argument
577 containing the DNS resolver options. The host_find_byname() function now
578 sets these options at its start, just as host_find_bydns() does. The smtp
579 transport options dns_qualify_single and dns_search_parents are passed to
580 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
581 of host_find_byname() use the default settings of RES_DEFNAMES
582 (qualify_single) but not RES_DNSRCH (search_parents).
584 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
585 spool_read_header() do less string testing, by means of a preliminary
586 switch on the second character of optional "-foo" lines. (This is
587 overdue, caused by the large number of possibilities that now exist.
588 Originally there were few.) While I was there, I also converted the
589 str(n)cmp tests so they don't re-test the leading "-" and the first
590 character, in the hope this might squeeze out yet more improvement.
592 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
593 flag allowing group syntax was set by the header_syntax check but not
594 turned off, possible causing trouble later; (2) The flag was not being
595 set at all for the header_verify test, causing "group"-style headers to
596 be rejected. I have now set it in this case, and also caused header_
597 verify to ignore an empty address taken from a group. While doing this, I
598 came across some other cases where the code for allowing group syntax
599 while scanning a header line wasn't quite right (mostly, not resetting
600 the flag correctly in the right place). These bugs could have caused
601 trouble for malformed header lines. I hope it is now all correct.
603 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
604 with the "reply" argument non-NULL. The code, however (which originally
605 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
606 but it didn't always do it. This confused somebody who was copying the
607 code for some other use. I have removed all the tests.
609 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
610 feature that was used to support insecure browsers during the U.S. crypto
611 embargo. It requires special client support, and Exim is probably the
612 only MTA that supported it -- and would never use it because real RSA is
613 always available. This code has been removed, because it had the bad
614 effect of slowing Exim down by computing (never used) parameters for the
615 RSA_EXPORT functionality.
617 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
618 authenticator to fail if there's a tab character in the incoming data
619 (there should never be unless someone is messing about, as it's supposed
620 to be base64-encoded). Also added, on Timo's advice, the "secured" option
621 if the connection is using TLS or if the remote IP is the same as the
622 local IP, and the "valid-client-cert option" if a client certificate has
625 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
626 authenticators. This can be used for authorization after authentication
627 succeeds. (In the case of plaintext, it servers for both authentication
630 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
631 if any retry times were supplied.
633 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
634 connection, where there is no client IP address to check. In this
635 situation, the verify now always succeeds.
637 PH/25 Applied John Jetmore's -Mset patch.
639 PH/26 Added -bem to be like -Mset, but loading a message from a file.
641 PH/27 In a string expansion for a processed (not raw) header when multiple
642 headers of the same name were present, leading whitespace was being
643 removed from all of them, but trailing whitespace was being removed only
644 from the last one. Now trailing whitespace is removed from each header
645 before concatenation. Completely empty headers in a concatenation (as
648 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
649 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
651 PH/29 [Removed. This was a change that I later backed out, and forgot to
652 correct the ChangeLog entry (that I had efficiently created) before
653 committing the later change.]
655 PH/30 Exim was sometimes attempting to deliver messages that had suffered
656 address errors (4xx response to RCPT) over the same connection as other
657 messages routed to the same hosts. Such deliveries are always "forced",
658 so retry times are not inspected. This resulted in far too many retries
659 for the affected addresses. The effect occurred only when there were more
660 hosts than the hosts_max_try setting in the smtp transport when it had
661 the 4xx errors. Those hosts that it had tried were not added to the list
662 of hosts for which the message was waiting, so if all were tried, there
663 was no problem. Two fixes have been applied:
665 (i) If there are any address or message errors in an SMTP delivery, none
666 of the hosts (tried or untried) are now added to the list of hosts
667 for which the message is waiting, so the message should not be a
668 candidate for sending over the same connection that was used for a
669 successful delivery of some other message. This seems entirely
670 reasonable: after all the message is NOT "waiting for some host".
671 This is so "obvious" that I'm not sure why it wasn't done
672 previously. Hope I haven't missed anything, but it can't do any
673 harm, as the worst effect is to miss an optimization.
675 (ii) If, despite (i), such a delivery is accidentally attempted, the
676 routing retry time is respected, so at least it doesn't keep
677 hammering the server.
679 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
680 in ${readsocket because some servers need this prod.
682 PH/32 Added some extra debug output when updating a wait-xxx database.
684 PH/33 The hint "could be header name not terminated by colon", which has been
685 given for certain expansion errors for a long time, was not being given
686 for the ${if def:h_colon_omitted{... case.
688 PH/34 The spec says: "With one important exception, whenever a domain list is
689 being scanned, $domain contains the subject domain." There was at least
690 one case where this was not true.
692 PH/35 The error "getsockname() failed: connection reset by peer" was being
693 written to the panic log as well as the main log, but it isn't really
694 panic-worthy as it just means the connection died rather early on. I have
695 removed the panic log writing for the ECONNRESET error when getsockname()
698 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
699 runs only) independently of the message's sender address. This meant
700 that, if the 4xx error was in fact related to the sender, a different
701 message to the same recipient with a different sender could confuse
702 things. In particualar, this can happen when sending to a greylisting
703 server, but other circumstances could also provoke similar problems.
704 I have changed the default so that the retry time for these errors is now
705 based a combination of the sender and recipient addresses. This change
706 can be overridden by setting address_retry_include_sender=false in the
709 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
710 remote server are returned as part of bounce messages. This was not
711 happening for LMTP over a pipe (the lmtp transport), but now it is the
712 same for both kinds of LMTP.
714 PH/38 Despite being documented as not happening, Exim was rewriting addresses
715 in header lines that were in fact CNAMEs. This is no longer the case.
717 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
718 and queue runs started by the daemon processed all messages. This has
719 been fixed so that -R and -S can now usefully be given with -q<time>.
721 PH/40 Import PCRE release 6.7 (fixes some bugs).
723 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
725 PH/42 Give an error if -q is specified more than once.
727 PH/43 Renamed the variables $interface_address and $interface_port as
728 $received_ip_address and $received_port, to make it clear that these
729 values apply to message reception, and not to the outgoing interface when
730 a message is delivered. (The old names remain recognized, of course.)
732 PH/44 There was no timeout on the connect() call when using a Unix domain
733 socket in the ${readsocket expansion. There now is.
735 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
736 be meaningful with "accept".
738 SC/01 Eximstats V1.43
739 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
741 SC/02 Eximstats V1.44
742 Use a glob alias rather than an array ref in the generated
743 parser. This improves both readability and performance.
745 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
746 Collect SpamAssassin and rejection statistics.
747 Don't display local sender or destination tables unless
748 there is data to show.
749 Added average volumes into the top table text output.
751 SC/04 Eximstats V1.46
752 Collect data on the number of addresses (recipients)
753 as well as the number of messages.
755 SC/05 Eximstats V1.47
756 Added 'Message too big' to the list of mail rejection
757 reasons (thanks to Marco Gaiarin).
759 SC/06 Eximstats V1.48
760 Mainlog lines which have GMT offsets and are too short to
761 have a flag are now skipped.
763 SC/07 Eximstats V1.49 (Alain Williams)
764 Added the -emptyok flag.
766 SC/08 Eximstats V1.50
767 Fixes for obtaining the IP address from reject messages.
769 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
770 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
771 whitesspace changes from 4.64-PH/27
773 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
776 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
777 are found, allow negative numbers in numeric criteria)
779 JJ/06 exipick.20061117.2, added new $message_body_missing variable
781 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
782 to match changes made in 4.64-PH/43
784 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
786 PH/47 Put in an explicit test for a DNS lookup of an address record where the
787 "domain" is actually an IP address, and force a failure. This locks out
788 those revolvers/nameservers that support "A-for-A" lookups, in
789 contravention of the specifications.
791 PH/48 When a host name was looked up from an IP address, and the subsequent
792 forward lookup of the name timed out, the host name was left in
793 $sender_host_name, contrary to the specification.
795 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
796 restricted to single-key lookups, Exim was not diagnosing an error if
797 * or *@ was used with a query-style lookup.
799 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
801 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
802 long ago, but noone interested enough thought of it. Let's just say that
803 the "1.1" means that there are some new functions that weren't there at
804 some point in the past.
806 PH/51 Error processing for expansion failure of helo_data from an smtp
807 transport during callout processing was broken.
809 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
810 tested/used via the -bh/-bhc/-bs options.
812 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
813 bug, fixed in subsequent PCRE releases).
815 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
816 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
818 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
824 SC/01 Use a glob alias rather than an array ref in eximstats generated
825 parser. This improves both readability and performance.
827 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
828 Don't display local sender or destination tables in eximstats unless
829 there is data to show.
830 Added average volumes into the eximstats top table text output.
832 SC/03 Collect data on the number of addresses (recipients) as well
833 as the number of messages in eximstats.
835 TF/01 Correct an error in the documentation for the redirect router. Exim
836 does (usually) call initgroups() when daemonizing.
838 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
839 with consistent privilege compared to when running as a daemon.
841 TF/03 Note in the spec that $authenticated_id is not set for local
842 submissions from trusted users.
844 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
845 Thanks to Dean Brooks <dean@iglou.com> for the patch.
847 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
848 by adding some example configuration directives to the default
849 configuration file. A little bit of work is required to uncomment the
850 directives and define how usernames and passwords are checked, but
851 there is now a framework to start from.
853 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
854 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
855 without this. I don't know how relevant this is to other LDAP libraries.
857 PH/02 Add the verb name to the "unknown ACL verb" error.
859 PH/03 Magnus Holmgren's patch for filter_prepend_home.
861 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
863 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
864 directory not expanded when it should be if an expanded home directory
865 was set for the address (which is overridden by the transport).
867 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
870 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
871 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
872 because it is too late at that time, and has no effect.
874 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
875 security issue with \' (bugzilla #107). I could not use the
876 PQescapeStringConn() function, because it needs a PGconn value as one of
879 PH/08 When testing addresses using -bt, indicate those final addresses that
880 are duplicates that would not cause an additional delivery. At least one
881 person was confused, thinking that -bt output corresponded to deliveries.
882 (Suppressing duplicates isn't a good idea as you lose the information
883 about possibly different redirections that led to the duplicates.)
885 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
886 systems where poll() doesn't work, in particular OS X.
888 PH/10 Added more information to debugging output for retry time not reached.
890 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
891 operations in malware.c.
893 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
896 PH/13 If write_rejectlog was set false when logging was sent to syslog with
897 syslog_duplication set false, log lines that would normally be written
898 both the the main log and to the reject log were not written to syslog at
901 PH/14 In the default configuration, change the use of "message" in ACL warn
902 statements to "add_header".
904 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
905 not followed by a command (e.g. "seen endif").
907 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
908 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
911 PH/17 Added extra conditions to the default value of delay_warning_condition
915 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
916 { match{$h_precedence:}{(?i)bulk|list|junk} } \
917 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
920 The Auto-Submitted: and various List- headers are standardised, whereas I
921 don't think Precedence: ever was.
923 PH/18 Refactored debugging code in route_finduser() to show more information,
924 in particular, the error code if getpwnam() issues one.
926 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
927 This is apparently needed in addition to the PH/07 change above to avoid
928 any possible encoding problems.
930 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
931 but not after initializing Perl.
933 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
934 output them only if debugging. By default they are written stderr,
935 apparently, which is not desirable.
937 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
940 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
943 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
945 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
946 authenticated or an ident call has been made. Suppress the default
947 values for $authenticated_id and $authenticated_sender (but permit -oMai
948 and -oMas) when testing with -bh.
950 PH/24 Re-jigged the order of the tests in the default configuration so that the
951 tests for valid domains and recipients precede the DNS black list and CSA
952 tests, on the grounds that those ones are more expensive.
954 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
955 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
956 This bug exists in every version of Exim that I still have, right back to
959 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
960 However, an attempt to turn on case-sensitivity in a regex key by
961 including (?-i) didn't work because the subject string was already
962 lowercased, and the effects were non-intuitive. It turns out that a
963 one-line patch can be used to allow (?-i) to work as expected.
969 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
970 other effects) broke the use of negated acl sub-conditions.
972 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
975 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
976 "Deny" causes Exim to reject the incoming connection with a 554 error.
977 Unfortunately, if there is a major crisis, such as a disk failure,
978 tcp-wrappers gives "deny", whereas what one would like would be some
979 kind of temporary error. A kludge has been added to help with this.
980 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
981 554 error is used if errno is still zero or contains ENOENT (which occurs
982 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
985 PH/03 Add -lutil to the default FreeBSD LIBS setting.
987 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
988 errors. Otherwise a message that provokes a temporary error (when other
989 messages do not) can cause a whole host to time out.
991 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
992 addresses were routed directly to files or pipes from a redirect router.
993 File deliveries just didn't batch; pipe deliveries might have suffered
996 PH/06 A failure to get a lock for a hints database would erroneously always say
997 "Failed to get write lock", even when it was really a read lock.
999 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1000 of 0600. This has been changed to use the value of the lockfile_mode
1001 option (which defaults to 0600).
1003 PH/08 Applied small patch from the Sieve maintainer.
1005 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1006 folder from quota calculations, a direct delivery into this folder messed
1007 up the contents of the maildirsize file. This was because the regex was
1008 used only to exclude .Trash (or whatever) when the size of the mailbox
1009 was calculated. There was no check that a delivery was happening into an
1010 excluded directory. This bug has been fixed by ignoring all quota
1011 processing for deliveries into excluded directories.
1013 PH/10 Added the maildirfolder_create_regex option to appendfile.
1019 PH/01 The code for finding all the local interface addresses on a FreeBSD
1020 system running IPv6 was broken. This may well have applied to all BSD
1021 systems, as well as to others that have similar system calls. The broken
1022 code found IPv4 interfaces correctly, but gave incorrect values for the
1023 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1024 that it would not match correctly against @[] and not recognize the IPv6
1027 PH/02 The ipliteral router was not recognizing addresses of the form user@
1028 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1030 PH/03 Added disable_ipv6.
1032 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1033 decoded form, because it is most often used to construct To: headers
1034 lines in autoreplies, and the decoded form may well be syntactically
1035 invalid. However, $reply_address has leading white space removed, and all
1036 newlines turned into spaces so that the autoreply transport does not
1039 PH/05 If group was specified without a user on a router, and no group or user
1040 was specified on a transport, the group from the router was ignored.
1042 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1043 for visible compile-time settings that can be used to change these
1044 numbers, for those that want even more. Backwards compatibility with old
1045 spool files has been maintained. However, going back to a previous Exim
1046 release will lost any variables that are in spool files.
1048 PH/07 Two small changes when running in the test harness: increase delay when
1049 passing a TCP/IP connection to a new process, in case the original
1050 process has to generate a bounce, and remove special handling of
1051 127.0.0.2 (sic), which is no longer necessary.
1053 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1054 be the same on different OS.
1056 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1059 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1060 whether --show-vars was specified or not
1062 JJ/02 exipick: Added support for new ACL variable spool format introduced
1065 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1066 syntactically invalid From: or Reply-to: line, and a filter used this to
1067 generate an autoreply, and therefore failed to obtain an address for the
1068 autoreply, Exim could try to deliver to a non-existent relative file
1069 name, causing unrelated and misleading errors. What now happens is that
1070 it logs this as a hard delivery error, but does not attempt to create a
1073 PH/11 The exinext utility has a -C option for testing purposes, but although
1074 the given file was scanned by exinext itself; it wasn't being passed on
1075 when Exim was called.
1077 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1078 an end-of-file indication when reading a command response.
1080 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1081 compiled. In many other places in Exim, IPv6 addresses are always
1082 recognized, so I have changed this. It also means that IPv4 domain
1083 literals of the form [IPV4:n.n.n.n] are now always recognized.
1085 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1086 used if the router is not running as root, for example, when verifying at
1087 ACL time, or when using -bh. The debugging output from this situation was
1088 non-existent - all you got was a failure to exec. I have made two
1091 (a) Failures to set uid/gid, the current directory, or a process leader
1092 in a subprocess such as that created by queryprogram now generate
1093 suitable debugging ouput when -d is set.
1095 (b) The queryprogram router detects when it is not running as root,
1096 outputs suitable debugging information if -d is set, and then runs
1097 the subprocess without attempting to change uid/gid.
1099 PH/15 Minor change to Makefile for building test_host (undocumented testing
1102 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1103 additional section of a DNS packet that returns MX or SRV records.
1104 Instead, it always explicitly searches for A/AAAA records. This avoids
1105 major problems that occur when a DNS server includes only records of one
1106 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1107 fixed another bug: if SRV records were looked up and the corresponding
1108 address records were *not* found in the additional section, the port
1109 values from the SRV records were lost.
1111 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1112 using the correct key (the original address) when searching the retry
1113 rules in order to find which one to use for generating the retry hint.
1115 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1116 adding the default one. Similarly, if it contains a Reply-To: header, the
1117 errors_reply_to option, if set, is not used.
1119 PH/19 When calculating a retry time, Exim used to measure the "time since
1120 failure" by looking at the "first failed" field in the retry record. Now
1121 it does not use this if it is later than than the arrival time of the
1122 message. Instead it uses the arrival time. This makes for better
1123 behaviour in cases where some deliveries succeed, thus re-setting the
1124 "first failed" field. An example is a quota failure for a huge message
1125 when small messages continue to be delivered. Without this change, the
1126 "time since failure" will always be short, possible causing more frequent
1127 delivery attempts for the huge message than are intended.
1128 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1130 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1131 $1, $2, $3) because the numerical variables can be reset during some
1132 expansion items (e.g. "match"), thereby losing the authentication data.
1134 PH/21 Make -bV show the size of off_t variables so that the test suite can
1135 decide whether to run tests for quotas > 2G.
1137 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1138 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1139 filecount value is greater than 2G or if a quota value is greater than 2G
1140 on a system where the size of off_t is not greater than 4, a panic error
1143 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1144 never match. The debug and -bh output now contains an explicit error
1145 message indicating a malformed IPv4 address or mask.
1147 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1148 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1149 PH/23 above applies.
1151 PH/25 Do not write to syslog when running in the test harness. The only
1152 occasion when this arises is a failure to open the main or panic logs
1153 (for which there is an explicit test).
1155 PH/26 Added the /no_tell option to "control=freeze".
1157 PH/27 If a host name lookup failed very early in a connection, for example, if
1158 the IP address matched host_lookup and the reverse lookup yielded a name
1159 that did not have a forward lookup, an error message of the form "no IP
1160 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1161 could be logged. Now it outputs the IP address instead of "NULL".
1163 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1164 allows the sender and the authenticated sender to be set when
1165 submitting a message from within Exim. Since child_open_exim() is
1166 documented for local_scan(), the new function should be too.
1168 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1169 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1170 results in an empty string is now treated as unset.
1172 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1174 PH/31 Added sender_verify_fail logging option.
1176 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1177 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1178 tidied the source and removed it altogether.
1180 PH/33 When a queue run was abandoned because the load average was too high, a
1181 log line was always written; now it is written only if the queue_run log
1182 selector is set. In addition, the log line for abandonment now contains
1183 information about the queue run such as the pid. This is always present
1184 in "start" and "stop" lines but was omitted from the "abandon" line.
1186 PH/34 Omit spaces between a header name and the colon in the error message that
1187 is given when verify = headers_syntax fails (if there are lots of them,
1188 the message gets confusing).
1190 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1191 names, as there are now some PTR records that contain slashes. This check
1192 is only to protect against broken name servers that fall over on strange
1193 characters, so the fact that it applies to all lookups doesn't matter.
1195 PH/36 Now that the new test suite is complete, we can remove some of the
1196 special code in Exim that was needed for the old test suite. For example,
1197 sorting DNS records because real resolvers return them in an arbitrary
1198 order. The new test suite's fake resolver always returns records in the
1201 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1202 bounces) except when queue_only is set, to avoid logging races between
1203 the different processes.
1205 PH/38 Panic-die if .include specifies a non-absolute path.
1207 PH/39 A tweak to the "H" retry rule from its user.
1209 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1210 a label. They prevented compilation on older perls.
1212 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1213 a warning to be raised on newish perls.
1215 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1216 on queue. Changes to match documented behaviour of showing count of
1217 messages matching specified criteria.
1219 PH/40 Changed the default ident timeout from 30s to 5s.
1221 PH/41 Added support for the use of login_cap features, on those BSD systems
1222 that have them, for controlling the resources used by pipe deliveries.
1224 PH/42 The content-scanning code uses fopen() to create files in which to put
1225 message data. Previously it was not paying any attention to the mode of
1226 the files. Exim runs with umask(0) because the rest of the code creates
1227 files with open(), and sets the required mode explicitly. Thus, these
1228 files were ending up world-writeable. This was not a big issue, because,
1229 being within the spool directory, they were not world-accessible. I have
1230 created a function called modefopen, which takes an additional mode
1231 argument. It sets umask(777), creates the file, chmods it to the required
1232 mode, then resets the umask. All the relevant calls to fopen() in the
1233 content scanning code have been changed to use this function.
1235 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1236 to 24 hours. This avoids potential overflow problems when processing G
1237 and H retry rules. I suspect nobody ever tinkers with this value.
1239 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1241 PH/45 When the plaintext authenticator is running as a client, the server's
1242 challenges are checked to ensure they are valid base64 strings. By
1243 default, the authentication attempt is cancelled if an invalid string is
1244 received. Setting client_ignore_invalid_base64 true ignores these errors.
1245 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1246 they are received. Thus, the responses can be made to depend on the
1247 challenges. If an invalid string is ignored, an empty string is placed in
1250 PH/46 Messages that are created by the autoreply transport now contains a
1251 References: header, in accordance with RFCs 2822 and 3834.
1253 PH/47 Added authenticated_sender_force to the smtp transport.
1255 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1257 PH/49 Installed latest patch from the Sieve maintainer.
1259 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1260 was also set, the appendfile transport was unnecessarily scanning a
1261 directory of message files (e.g. for maildir delivery) to find the count
1262 of files (along with the size), even though it did not need this
1263 information. It now does the scan only if it needs to find either the
1264 size of the count of files.
1266 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1268 PH/52 Two bugs concerned with error handling when the smtp transport is
1271 (i) Exim was not creating retry information for temporary errors given
1272 for individual recipients after the DATA command when the smtp transport
1273 was used in LMTP mode. This meant that they could be retried too
1274 frequently, and not timed out correctly.
1276 (ii) Exim was setting the flag that allows error details to be returned
1277 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1278 recipients that were returned after the DATA command.
1280 PH/53 This is related to PH/52, but is more general: for any failing address,
1281 when detailed error information was permitted to be returned to the
1282 sender, but the error was temporary, then after the final timeout, only
1283 "retry timeout exceeded" was returned. Now it returns the full error as
1284 well as "retry timeout exceeded".
1286 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1287 do this, and (what is worse) MTAs that accept it.
1289 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1290 will now be deprecated.
1292 PH/56 New os.c-cygwin from the Cygwin maintainer.
1294 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1295 formats (previously only available in exim formats via -bpr, -bpru,
1296 and -bpra. Now also available in native and exiqgrep formats)
1298 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1299 with very large, slow to parse queues
1301 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1303 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1305 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1306 responses to authentication challenges, though it was showing the
1307 challenges; (ii) I've removed the CR characters from the debug output for
1310 PH/58 Allow for the insertion of a newline as well as a space when a string
1311 is turned into more than one encoded-word during RFC 2047 encoding. The
1312 Sieve code now uses this.
1314 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1315 data_4xx, lost_connection, tls_required.
1317 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1318 message was being sent as an SMTP response.
1320 PH/61 Add -l and -k options to exicyclog.
1322 PH/62 When verifying, if an address was redirected to one new address, so that
1323 verification continued, and the new address failed or deferred after
1324 having set something in $address_data, the value of $address_data was not
1325 passed back to the ACL. This was different to the case when no
1326 redirection occurred. The value is now passed back in both cases.
1328 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1329 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1330 use login_cap.h, so on its own it isn't the distinguishing feature. The
1331 new name refers directly to the setclassresources() function.
1333 PH/65 Added configuration files for NetBSD3.
1335 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1337 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1338 is preferred over IPv4.
1340 PH/68 The bounce_return_message and bounce_return_body options were not being
1341 honoured for bounces generated during the reception of non-SMTP messages.
1342 In particular, this applied to messages rejected by the ACL. This bug has
1343 been fixed. However, if bounce_return_message is true and bounce_return_
1344 body is false, the headers that are returned for a non-SMTP message
1345 include only those that have been read before the error was detected.
1346 (In the case of an ACL rejection, they have all been read.)
1348 PH/69 The HTML version of the specification is now built in a directory called
1349 spec_html instead of spec.html, because the latter looks like a path with
1350 a MIME-type, and this confuses some software.
1352 PH/70 Catch two compiler warnings in sieve.c.
1354 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1355 function verify_get_ident() calls ip_connect() to connect a socket, but
1356 if the "connect()" function timed out, ip_connect() used to close the
1357 socket. However, verify_get_ident() also closes the socket later, and in
1358 between Exim writes to the log, which may get opened at this point. When
1359 the socket was closed in ip_connect(), the log could get the same file
1360 descriptor number as the socket. This naturally causes chaos. The fix is
1361 not to close the socket in ip_connect(); the socket should be closed by
1362 the function that creates it. There was only one place in the code where
1363 this was missing, in the iplookup router, which I don't think anybody now
1364 uses, but I've fixed it anyway.
1366 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1367 well as to direct DNS lookups. Otherwise the handling of names in host
1368 lists is inconsistent and therefore confusing.
1374 PH/01 Two changes to the default runtime configuration:
1376 (1) Move the checks for relay_from_hosts and authenticated clients from
1377 after to before the (commented out) DNS black list checks.
1379 (2) Add control=submission to the relay_from_hosts and authenticated
1380 clients checks, on the grounds that messages accepted by these
1381 statements are most likely to be submissions.
1383 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1385 (1) Generate an error if the third argument for the ${prvs expansion is
1388 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1391 (3) Reset the variables that are obtained from the first argument of
1392 ${prvscheck and used in the second argument before leaving the code,
1393 because their memory is reclaimed, so using them afterwards may do
1396 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1397 one (it's much easier than Tom thought :-).
1399 (5) Because of (4), we can now allow for the use of $prvscheck_result
1400 inside the third argument.
1402 PH/03 For some reason, the default setting of PATH when running a command from
1403 a pipe transport was just "/usr/bin". I have changed it to
1406 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1407 anything to be listed in the output from -bV.
1409 PH/05 When a filter generated an autoreply, the entire To: header line was
1410 quoted in the delivery log line, like this:
1412 => >A.N.Other <ano@some.domain> <original@ddress> ...
1414 This has been changed so that it extracts the operative address. There
1415 may be more than one such address. If so, they are comma-separated, like
1418 => >ano@some.domain,ona@other.domain <original@ddress> ...
1420 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1421 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1422 not being looked up in the rDNS to get a host name, Exim was showing the
1423 IP address twice in Received: lines, even though the IP addresses were
1424 identical. For example:
1426 Received: from [1.2.3.4] (helo=[1.2.3.4])
1428 However, if the real host name was known, it was omitting the HELO data
1429 if it matched the actual IP address. This has been tidied up so that it
1430 doesn't show the same IP address twice.
1432 PH/07 When both +timestamp and +memory debugging was on, the value given by
1433 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1434 called by the debug printing, thereby overwriting the timestamp buffer.
1435 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1437 PH/08 When the original message was included in an autoreply transport, it
1438 always said "this is a copy of the message, including all the headers",
1439 even if body_only or headers_only was set. It now gives an appropriate
1442 PH/09 Applied a patch from the Sieve maintainer which:
1444 o fixes some comments
1445 o adds the (disabled) notify extension core
1446 o adds some debug output for the result of if/elsif tests
1447 o points to the current vacation draft in the documentation
1448 and documents the missing references header update
1452 o fixes a bug in processing the envelope test (when testing
1453 multiple envelope elements, the last element determinted the
1456 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1457 Electronic Mail") by including:
1459 Auto-submitted: auto-generated
1461 in the messages that it generates (bounce messages and others, such as
1462 warnings). In the case of bounce messages for non-SMTP mesages, there was
1463 also a typo: it was using "Auto_submitted" (underscore instead of
1464 hyphen). Since every message generated by Exim is necessarily in response
1465 to another message, thes have all been changed to:
1467 Auto-Submitted: auto-replied
1469 in accordance with these statements in the RFC:
1471 The auto-replied keyword:
1473 - SHOULD be used on messages sent in direct response to another
1474 message by an automatic process,
1476 - MUST NOT be used on manually-generated messages,
1478 - MAY be used on Delivery Status Notifications (DSNs) and Message
1479 Disposition Notifications (MDNs),
1481 - MUST NOT be used on messages generated by automatic or periodic
1482 processes, except for messages which are automatic responses to
1485 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1486 to the default Received: header definition.
1488 PH/12 Added log selector acl_warn_skipped (default on).
1490 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1491 variables because (a) they are in the wrong storage pool and (b) even if
1492 they were copied, it wouldn't work properly because of the caching.
1494 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1495 checking when decoding. Apparently there are clients that generate
1496 overlong encoded strings. Why am I not surprised?
1498 PH/15 If the first argument of "${if match_address" was not empty, but did not
1499 contain an "@" character, Exim crashed. Now it writes a panic log message
1500 and treats the condition as false.
1502 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1504 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1505 extension "envelope-auth". The code is finished and in agreement with
1506 other implementations, but there is no documentation so far and in fact,
1507 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1508 not changing the active code.
1510 Print executed "if" and "elsif" statements when debugging is used. This
1511 helps a great deal to understand what a filter does.
1513 Document more things not specified clearly in RFC3028. I had all this
1514 sorted out, when out of a sudden new issues came to my mind. Oops."
1516 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1519 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1520 canonical form (as documented). However, after a host name lookup from
1521 the IP address, check_host() was doing a simple string comparison with
1522 addresses acquired from the DNS when checking that the found name did
1523 have the original IP as one of its addresses. Since any found IPv6
1524 addresses are likely to be in abbreviated form, the comparison could
1525 fail. Luckily, there already exists a function for doing the comparison
1526 by converting both addresses to binary, so now that is used instead of
1527 the text comparison.
1529 PH/20 There was another similar case to PH/19, when a complete host name was
1530 given in a host list; looking up its IP address could give an abbreviated
1531 form, whereas the current host's name might or might not be abbreviated.
1532 The same fix has been applied.
1538 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1539 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1542 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1543 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1545 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1547 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1548 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1549 Also, for "Auto-Submitted", treat anything other than "no" as
1550 non-personal, in accordance with RFC 3834. (Previously it treated
1551 anything starting "auto-" as non-personal.)
1553 TF/01 The control=submission/name=... option had a problem with syntax
1554 errors if the name included a slash character. The /name= option
1555 now slurps the rest of the string, so it can include any characters
1556 but it must come last in the list of options (after /sender_retain
1559 PH/05 Some modifications to the interface to the fake nameserver for the new
1567 TK/01 Added the "success_on_redirect" address verification option. See
1568 NewStuff for rationale and an example.
1570 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1572 PH/02 Patch to exigrep to allow it to work on syslog lines.
1574 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1575 fread() to skip over the body file's header line, because in Cygwin the
1576 header line is locked and is inaccessible.
1578 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1579 co-exist for some time) to make it clear that it is the Exim ID that is
1580 referenced, not the Message-ID: header line.
1582 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1583 string_format() function, because snprintf() does not exist on all
1586 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1587 ${stat: expansion item.
1589 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1590 protocol synchronization error", to keep the pedants happy.
1592 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1593 well as for IRIX systems, when gcc is being used. See the host.c source
1596 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1598 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1601 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1602 transport and to the smtp transport in LMTP mode.
1604 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1606 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1607 run a filter in a subprocess. This could lead to confusion in subsequent
1608 lookups in the parent process. There should also be a search_tidyup() at
1609 the end of the subprocess.
1611 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1612 only if the host matched helo_try_verify_hosts, which caused the
1613 verification to occur when the EHLO/HELO command was issued. The ACL just
1614 tested the remembered result. Now, if a previous verification attempt has
1615 not happened, "verify = helo" does it there and then.
1617 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1619 TK/03 Fix log output including CR from clamd.
1621 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1622 exist provoked a memory error which could cause a segfault.
1624 PH/15 Installed PCRE 6.2
1626 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1628 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1629 of the problem. Specifically, suggested +O2 rather than +O1 for the
1632 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1634 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1635 record, the port was not correctly logged when the outgoing_port log
1636 selector was set (it logged the transort's default port).
1638 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1639 fallback_hosts, and "hosts" in the smtp transport.
1641 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1642 host errors such as "Connection refused".
1644 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1645 authentication with radiusclient 0.4.9:
1647 - Error returned from rc_read_config was caught wrongly
1648 - Username/password not passed on to radius server due to wrong length.
1650 The presumption is that some radiusclient API changes for 4.51/PH/17
1651 were not taken care of correctly. The code is still untested by me (my
1652 Linux distribution still has 0.3.2 of radiusclient), but it was
1653 contributed by a Radius user.
1655 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1656 expanding the "port" option of the smtp transport.
1658 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1659 while reading a MIME header. Thanks to Tom Hughes for a patch.
1661 PH/24 Include config.h inside local_scan.h so that configuration settings are
1664 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1665 that in an ACL for RCPT (for example), you can examine exactly what was
1668 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1669 commands, but it was not correctly comparing the address with the actual
1670 client host address. Thus, it would show the EHLO address in Received:
1671 header lines when this was not necessary.
1673 PH/27 Added the % operator to ${eval:}.
1675 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1676 it should be ignoring failures (because with -C, for example, it has lost
1677 privilege). It wasn't ignoring creation failures other than "already
1680 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1683 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1684 because an input line was too long, either on its own, or by virtue of
1685 too many continuations, the temporary file was not being removed, and the
1686 return code was incorrect.
1688 PH/31 Missing "BOOL" in function definition in filtertest.c.
1690 PH/32 Applied Sieve patches from the maintainer.
1692 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1694 PH/33 Added "verify = not_blind".
1696 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1697 Local/Makefile (with some defaults set). These are used in built scripts
1698 such as exicyclog, but they have never been used in the exim_install
1699 script (though there are many overriding facilities there). I have
1700 arranged that the exim_install script now takes note of these two
1703 PH/35 Installed configuration files for Dragonfly.
1705 PH/36 When a locally submitted message by a trusted user did not contain a
1706 From: header, and the sender address was obtained from -f or from an SMTP
1707 MAIL command, and the trusted user did not use -F to supply a sender
1708 name, $originator_name was incorrectly used when constructing a From:
1709 header. Furthermore, $originator_name was used for submission mode
1710 messages from external hosts without From: headers in a similar way,
1711 which is clearly wrong.
1713 PH/37 Added control=suppress_local_fixups.
1715 PH/38 When log_selector = +received_sender was set, and the addition of the
1716 sender made the log line's construction buffer exactly full, or one byte
1717 less than full, an overflow happened when the terminating "\n" was
1720 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1721 when the result of a list match is failure because a DNS lookup failed.
1723 PH/40 RM_COMMAND is now used in the building process.
1725 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1726 the "build-* directories that it finds.
1728 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1729 address in a domain literal was a prefix of an interface address.
1731 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1732 when verifying a sender address, unless rewrite_headers is false.
1734 PH/44 Wrote a long comment about why errors_to addresses are verified as
1735 recipients, not senders.
1737 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1738 the ratelimit ACL was added.
1740 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1742 PH/46 Added extra information about PostgreSQL errors to the error string.
1744 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1745 avoiding the need to install special zones in a real server. This is
1746 backwards compatible; if it can't find the fake resolver, it drops back.
1747 Thus, both old and new test suites can be run.
1749 TF/02 Added util/ratelimit.pl
1751 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1752 clock is set back in time.
1754 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1755 Candler <B.Candler@pobox.com>.
1757 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1758 OK for addresses that are the result of redirections.
1760 PH/48 A number of further additions for the benefit of the new test suite,
1761 including a fake gethostbyname() that interfaces to the fake DNS resolver
1764 TF/06 The fix for widen_domains has also been applied to qualify_single and
1765 search_parents which are the other dnslookup options that can cause
1768 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1771 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1773 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1774 service" if it encountered an MX record with an empty target hostname.
1775 The message is now "an MX or SRV record indicated no SMTP service".
1777 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1778 if the DNS of the sending site is misconfigured. This is quite a
1779 common situation. This change restores the behaviour of treating a
1780 helo verification defer as a failure.
1782 PH/51 If self=fail was set on a router, the bounce message did not include the
1783 actual error message.
1789 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1791 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1792 command itself ended in error, the underlying message about the transport
1793 filter timeout was being overwritten with the pipe command error. Now the
1794 underlying error message should be appended to the second error message.
1796 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1798 PH/02 Reduce the amount of output that "make" produces by default. Full output
1799 can still be requested.
1801 PH/03 The warning log line about a condition test deferring for a "warn" verb
1802 was being output only once per connection, rather than after each
1803 occurrence (because it was using the same function as for successful
1804 "warn" verbs). This seems wrong, so I have changed it.
1806 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1807 it should not have, which might have caused a crash in the right
1808 circumstances, but probably never did.
1810 PH/04 Installed a modified version of Tony Finch's patch to make submission
1811 mode fix the return path as well as the Sender: header line, and to
1812 add a /name= option so that you can make the user's friendly name appear
1815 TF/03 Added the control = fakedefer ACL modifier.
1817 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1818 Mark Lowes for thorough testing.
1820 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1822 TK/03 Merged latest SRS patch from Miles Wilton.
1824 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1825 with the definition in sysexits.h (which is #included earlier).
1826 Fortunately, Exim does not actually use EX_OK. The code used to try to
1827 preserve the sysexits.h value, by assumimg that macro definitions were
1828 scanned for macro replacements. I have been disabused of this notion,
1829 so now the code just undefines EX_OK before #including unistd.h.
1831 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1832 in the smtp transport. When a block could not be written in a single
1833 write() function, the timeout was being re-applied to each part-write.
1834 This seems wrong - if the receiver was accepting one byte at a time it
1835 would take for ever. The timeout is now adjusted when this happens. It
1836 doesn't have to be particularly precise.
1838 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1839 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1841 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1842 without a domain if the check to <postmaster@domain> fails.
1844 SC/01 Eximstats: added -xls and the ability to specify output files
1845 (patch written by Frank Heydlauf).
1847 SC/02 Eximstats: use FileHandles for outputing results.
1849 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1851 SC/04 Eximstats: fixed display of large numbers with -nvr option
1853 SC/05 Eximstats: fixed merging of reports with empty tables.
1855 SC/06 Eximstats: added the -include_original_destination flag
1857 SC/07 Eximstats: removed tabs and trailing whitespace.
1859 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1861 TK/06 MBOX spool code: Add real "From " MBOX separator line
1862 so the .eml file is really in mbox format (even though
1863 most programs do not really care). Patch from Alex Miller.
1865 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1866 The latter is generated from $received_to and is only set if the
1867 message has one envelope recipient. SA can use these headers,
1868 obviously out-of-the-box. Patch from Alex Miller.
1870 PH/08 The ${def test on a variable was returning false if the variable's
1871 value was "0", contrary to what the specification has always said!
1872 The result should be true unless the variable is empty.
1874 PH/09 The syntax error of a character other than { following "${if
1875 def:variable_name" (after optional whitespace) was not being diagnosed.
1876 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1877 accidental colon was present, for example, could give incorrect results.
1879 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1880 result is used (not including appendfile, where other changes are about
1883 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1884 This involved changing a lot of size variables from int to off_t. It
1885 should work with maildirs and everything.
1887 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1888 spamd dying while we are connected to it.
1890 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1893 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1894 function for BDB 4.3.
1896 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1898 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1899 the sources for PCRE have been re-arranged and more files are now
1902 PH/15 The code I had for printing potentially long long variables in PH/11
1903 above was not the best (it lost precision). The length of off_t variables
1904 is now inspected at build time, and an appropriate printing format (%ld
1905 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1906 to be "long long int" or "long int". This is needed for the internal
1907 formatting function string_vformat().
1909 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1910 the configuration file to be ":syslog", then the script "guesses" where
1911 the logs files are, rather than using the compiled in default. In our
1912 case the guess is not the same as the compiled default, so the script
1913 suddenly stopped working when I started to use syslog. The patch checks
1914 to see if log_file_path is "". If so, it attempts to read it from exim
1915 with no configuration file to get the compiled in version, before it
1916 falls back to the previous guessing code."
1918 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1919 implementing BATV in an Exim configuration. See NewStuff for the gory
1922 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1923 Makefile that are specific to HP-UX.
1925 PH/18 If the "use_postmaster" option was set for a recipient callout together
1926 with the "random" option, the postmaster address was used as the MAIL
1927 FROM address for the random test, but not for the subsequent recipient
1928 test. It is now used for both.
1930 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1931 patch removes a few documentation additions to RFC 3028, because the
1932 latest draft now contains them. It adds the new en;ascii-case comparator
1933 and a new error check for 8bit text in MIME parts. Comparator and
1934 require names are now matched exactly. I enabled the subaddress
1935 extension, but it is not well tested yet (read: it works for me)."
1937 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1938 rework some of the code of TK/09 above to avoid the hardwired use of
1939 "%lld" and "long long". Replaced the call to snprintf() with a call to
1942 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1943 records point to non-existent hosts", "retry timeout exceeded", and
1944 "retry time not reached for any host after a long failure period".
1946 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1947 experimental DomainKeys support:
1949 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1950 (2) On an error such as an illegally used "control", the wrong name for
1951 the control was given.
1953 These problems did NOT occur unless DomainKeys support was compiled.
1955 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1957 PH/24 Added ${if match_ip condition.
1959 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1960 databases so that it will be absolutely obvious if a crash occurs in the
1961 DB library. This is a regular occurrence (often caused by mis-matched
1964 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1965 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1966 picked up on a user's system that detects such things. There doesn't seem
1967 to be a gcc warning option for this - only an attribute that has to be
1968 put on the function's prototype. It seems that in Fedora Core 4 they have
1969 set this on a number of new functions. No doubt there will be more in due
1972 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1973 specify a transport. However, if an address that was verified by such a
1974 router was the subject of a callout, Exim crashed because it tried to
1975 read the rcpt_include_affixes from the non-existent transport. Now it
1976 just assumes that the setting of that option is false. This bug was
1977 introduced by 4.51/PH/31.
1979 PH/28 Changed -d+all to exclude +memory, because that information is very
1980 rarely of interest, but it makes the output a lot bigger. People tend to
1981 do -d+all out of habit.
1983 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1984 code in os-type was giving problems when libc.so lives in lib64, like on
1987 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1988 aren't the modern standard, and it seems that some systems' include files
1989 don't always have them. Exim was already checking for some of the newer
1990 ones like T_AAAA, and defining it itself. I've added checks for all the
1991 record types that Exim uses.
1993 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1994 not automatically generating a new one, as it is supposed to. This
1995 prevented TLS from working. If the file did exist, but contained invalid
1996 data, a new version was generated, as expected. It was only the case of a
1997 non-existent file that was broken.
1999 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2000 with a change in libdomainkeys > 0.64.
2002 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2003 from DNS. If the selector record carries the flag, it now has
2004 precedence over the domain-wide flag.
2006 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2008 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2009 the use of a port name that isn't defined in /etc/services) occurred, the
2010 message was deferred as in a normal delivery, and thus remained on the
2011 spool, instead of being failed because of the mua_wrapper setting. This
2012 is now fixed, and I tidied up some of the mua_wrapper messages at the
2015 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2016 the messages in a hash of arrays rather than using individual hashes.
2017 This is a bit cleaner and results in dramatic memory savings, albeit
2018 at a slight CPU cost.
2020 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2021 as requested by Marc Sherman.
2023 SC/10 Eximstats: added histograms for user specified patterns as requested
2026 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2028 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2029 fopen() in the content-scanning modules that did not already have it.
2035 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2036 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2038 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2040 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2042 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2043 to test Sieve filters that use "vacation".
2045 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2046 that changes the way the GnuTLS parameters are stored in the cache file.
2047 The new format can be generated externally. For backward compatibility,
2048 if the data in the cache doesn't make sense, Exim assumes it has read an
2049 old-format file, and it generates new data and writes a new file. This
2050 means that you can't go back to an older release without removing the
2053 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2054 work if there are any delivery delays because "one_time" forces the
2055 parent to be marked "delivered", so its unseen clone is never tried
2056 again. For this reason, Exim now forbids the simultaneous setting of
2059 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2060 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2061 entry if you want to know the details. The fix, however, neglected to
2062 consider the case when local delivery batching is involved. The test for
2063 "previously delivered" was not happening when checking to see if an
2064 address could be batched with a previous (undelivered) one; under
2065 certain circumstances this could lead to multiple deliveries to the same
2068 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2069 in its include files, and this causes problems building Exim.
2071 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2072 header_syntax) but Exim was just ignoring anything given after a slash.
2073 In particular, this caused confusion with an attempt to use "verify =
2074 reverse_host_lookup/defer_ok". An error is now given when options are
2075 supplied for verify items that do not have them. (Maybe reverse_host_
2076 lookup should have a defer_ok option, but that's a different point.)
2078 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2079 defined by RFC 821) to 2048, because there were problems with some AUTH
2080 commands, and RFC 1869 says the size should be increased for extended
2081 SMTP commands that take arguments.
2083 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2086 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2087 "unknown" error; now it says that the functionality isn't in the binary.
2089 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2090 an address' error message when a string expansion fails (syntax or
2091 whatever). Otherwise the password may appear in the log. Following change
2092 PH/42 below, there is no longer a chance of it appearing in a bounce
2095 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2097 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2098 of its addresses was ever tried. (Bugzilla bug #2.)
2100 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2101 the result incorrectly in the debug output. (It correctly added a newline
2102 to what was transported.)
2104 TF/01 Added $received_time.
2106 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2107 commented out examples of how to interface to a virus scanner and to
2108 SpamAssassin. Also added commented examples of av_scanner and
2109 spamd_address settings.
2111 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2112 and controls are allowed in which ACLs. There were a couple of minor
2113 errors. Some of the entries in the conditions table (which is a table of
2114 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2115 negation of where the condition IS allowed.
2117 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2119 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2120 header file does not have a version number, so I've had to invent a new
2121 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2122 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2123 radiusclient), but it was contributed by a Radius user.
2125 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2126 files or directories, for OpenSSL.
2128 PH/19 When an Exim process that is running as root has to create an Exim log
2129 file, it does so in a subprocess that runs as exim:exim so as to get the
2130 ownership right at creation (otherwise, other Exim processes might see
2131 the file with the wrong ownership). There was no test for failure of this
2132 fork() call, which would lead to the process getting stuck as it waited
2133 for a non-existent subprocess. Forks do occasionally fail when resources
2134 run out. I reviewed all the other calls to fork(); they all seem to check
2137 PH/20 When checking for unexpected SMTP input at connect time (before writing
2138 the banner), Exim was not dealing correctly with a non-positive return
2139 from the read() function. If the client had disconnected by this time,
2140 the result was a log entry for a synchronization error with an empty
2141 string after "input=" when read() returned zero. If read() returned -1
2142 (an event I could not check), uninitialized data bytes were printed.
2143 There were reports of junk text (parts of files, etc) appearing after
2146 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2148 PH/22 Added support for macro redefinition, and (re)definition in between
2149 driver and ACL definitions.
2151 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2152 forgetting to use the resulting value; it was using the unexpanded value.
2154 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2155 hadn't been configured. The fix is from Juergen Kreileder, who
2156 understands it better than I do:
2158 "Here's what I see happening with three configured cyrus_sasl
2159 authenticators configured (plain, login, cram-md5):
2161 On startup auth_cyrus_sasl_init() gets called for each of these.
2162 This means three calls to sasl_listmech() without a specified mech_list.
2163 => SASL tests which mechs of all available mechs actually work
2164 => three warnings about OTP not working
2165 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2167 With the patch, sasl_listmech() also gets called three times. But now
2168 SASL's mech_list option is set to the server_mech specified in the the
2169 authenticator. Or in other words, the answer from sasl_listmech()
2170 gets limited to just the mech you're testing for (which is different
2172 => the return list contains just 'plain' or 'login', 'cram-md5' or
2173 nothing depending on the value of ob->server_mech.
2175 I've just tested the patch: Authentication still works fine,
2176 unavailable mechs specified in the exim configuration are still
2177 caught, and the auth.log warnings about OTP are gone."
2179 PH/25 When debugging is enabled, the contents of the command line are added
2180 to the debugging output, even when log_selector=+arguments is not
2183 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2184 answer is "GNU", and only if the return is "GNU/something" is the answer
2187 PH/27 $acl_verify_message is now set immediately after the failure of a
2188 verification in an ACL, and so is available in subsequent modifiers. In
2189 particular, the message can be preserved by coding like this:
2191 warn !verify = sender
2192 set acl_m0 = $acl_verify_message
2194 Previously, $acl_verify_message was set only while expanding "message"
2195 and "log_message" when a very denied access.
2197 PH/28 Modified OS/os.c-Linux with
2199 -#ifndef OS_LOAD_AVERAGE
2200 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2202 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2203 nomenclature these days.)
2205 PH/29 Installed patch from the Sieve maintainer that adds the options
2206 sieve_useraddress and sieve_subaddress to the redirect router.
2208 PH/30 In these circumstances:
2209 . Two addresses routed to the same list of hosts;
2210 . First host does not offer TLS;
2211 . First host accepts first address;
2212 . First host gives temporary error to second address;
2213 . Second host offers TLS and a TLS session is established;
2214 . Second host accepts second address.
2215 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2216 and peerdn, if requested) that were in fact used only for the second
2219 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2220 attention to any local part prefix or suffix that was matched by the
2221 router that accepted the address. It now behaves in the same way as it
2222 does for delivery: the affixes are removed from the local part unless
2223 rcpt_include_affixes is set on the transport.
2225 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2226 timeout during the DATA phase of an incoming message.
2228 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2229 applied a patch sanctioned by the Sieve maintainer.
2231 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2232 the uid or gid is negative. A case of a negative gid caused this to be
2233 noticed. The fix allows for either to be negative.
2235 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2236 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2239 PH/36 The change PH/12 above was broken. Fixed it.
2241 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2242 the grounds that routing the same address twice would always produce the
2243 same answer. This might have been true once, but it is certainly no
2244 longer true now. Routing a child address may depend on the previous
2245 routing that produced that child. Some complicated redirection strategies
2246 went wrong when messages had multiple recipients, and made Exim's
2247 behaviour dependent on the order in which the addresses were given.
2249 I have moved the duplicate checking until after the routing is complete.
2250 Exim scans the addresses that are assigned to local and remote
2251 transports, and removes any duplicates. This means that more work will be
2252 done, as duplicates will always all be routed, but duplicates are
2253 presumably rare, so I don't expect this is of any significance.
2255 For deliveries to pipes, files, and autoreplies, the duplicate checking
2256 still happens during the routing process, since they are not going to be
2259 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2260 It corrects a timeout issue with spamd. This is Ian's comment: "The
2261 background is that sometimes spamd either never reads data from a
2262 connection it has accepted, or it never writes response data. The exiscan
2263 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2264 blindly assumes that writes won't block so it may never time out."
2266 PH/39 Allow G after quota size as well as K and M.
2268 PH/40 The value set for $authenticated_id in an authenticator may not contain
2269 binary zeroes or newlines because the value is written to log lines and
2270 to spool files. There was no check on this. Now the value is run through
2271 the string_printing() function so that such characters are converted to
2272 printable escape sequences.
2274 PH/41 $message_linecount is a new variable that contains the total number of
2275 lines in the message. Compare $body_linecount, which is the count for the
2278 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2279 bounce and delay warning messages, except in certain special cases, which
2282 (a) An SMTP error message from a remote host;
2283 (b) A message specified in a :fail: redirection;
2284 (c) A message specified in a "fail" command in a system filter;
2285 (d) A message specified in a FAIL return from the queryprogram router;
2286 (e) A message specified by the cannot_route_message router option.
2288 In these cases only, Exim does include the error details in bounce and
2289 warning messages. There are also a few cases where bland messages such
2290 as "unrouteable address" or "local delivery error" are given.
2292 PH/43 $value is now also set for the "else" part of a ${run expansion.
2294 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2295 being worked on, but at least Exim now implements the latest version to
2298 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2299 process to complete was treated as a delivery failure, a timeout while
2300 writing the message to the pipe was logged, but erroneously treated as a
2301 successful delivery. Such timeouts include transport filter timeouts. For
2302 consistency with the overall process timeout, these timeouts are now
2303 treated as errors, giving rise to delivery failures by default. However,
2304 there is now a new Boolean option for the pipe transport called
2305 timeout_defer, which, if set TRUE, converts the failures into defers for
2306 both kinds of timeout. A transport filter timeout is now identified in
2309 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2310 systems where "make" and "gmake" are different, calling "gmake" at top
2311 level broke things. I've arranged for the value of $(MAKE) to be passed
2312 from the Makefile to this script so that it can call the same version of
2316 A note about Exim versions 4.44 and 4.50
2317 ----------------------------------------
2319 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2320 changes of various kinds. As a consequence, a big documentation update was
2321 needed. This delayed the release for rather longer than seemed good, especially
2322 in the light of a couple of (minor) security issues. Therefore, the changes
2323 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2324 release. So 4.44 and 4.50 are in effect two different branches that both start
2327 I have left the 4.50 change log unchanged; it contains all the changes since
2328 4.43. The change log for 4.44 is below; many of its items are identical to
2329 those for 4.50. This seems to be the most sensible way to preserve the
2330 historical information.
2336 1. Minor wording change to the doc/README.SIEVE file.
2338 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2339 computation of the current number of files was incorrect.
2341 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2342 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2345 4. Give more explanation in the error message when the command for a transport
2346 filter fails to execute.
2348 5. There are several places where Exim runs a non-Exim command in a
2349 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2350 was being done only for the command run by the queryprogram router. It is
2351 now done for all such subprocesses. The other cases are: ${run, transport
2352 filters, and the commands run by the lmtp and pipe transports.
2354 6. Added CONFIGURE_GROUP build-time option.
2356 7. Some older OS have a limit of 256 on the maximum number of file
2357 descriptors. Exim was using setrlimit() to set 1000 as a large value
2358 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2359 systems. I've change it so that if it can't get 1000, it tries for 256.
2361 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2362 was an oversight, and furthermore, ever since the addition of extra
2363 controls (e.g. 4.43/32), the checks on when to allow different forms of
2364 "control" were broken. There should now be diagnostics for all cases when a
2365 control that does not make sense is encountered.
2367 9. Added the /retain_sender option to "control=submission".
2369 10. $recipients is now available in the predata ACL (oversight).
2371 11. Tidy the search cache before the fork to do a delivery from a message
2372 received from the command line. Otherwise the child will trigger a lookup
2373 failure and thereby defer the delivery if it tries to use (for example) a
2374 cached ldap connection that the parent has called unbind on.
2376 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2377 of $address_data from the recipient verification was clobbered by the
2378 sender verification.
2380 13. The value of address_data from a sender verification is now available in
2381 $sender_address_data in subsequent conditions in the ACL statement.
2383 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2385 15. Added a new option "connect=<time>" to callout options, to set a different
2388 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2389 was its contents. (It was OK if the option was not defined at all.)
2391 17. A "Completed" log line is now written for messages that are removed from
2392 the spool by the -Mrm option.
2394 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2395 information about exactly what failed.
2397 19. Added -dd to debug only the daemon process.
2399 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2400 handles timeouts, both on the server side and network timeouts. Renamed the
2401 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2403 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2404 It is now set to "smtps".
2406 22. $host_address is now set to the target address during the checking of
2407 ignore_target_hosts.
2409 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2410 being passed; this would have caused $sender_host_name to have been used if
2411 matching the list had actually called for a host name (not very likely,
2412 since this list is usually IP addresses). A host name is now passed as
2415 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2416 code that specifies a non-restarting handler (typically sigaction() in
2417 modern systems) in an attempt to fix a rare and obscure crash bug.
2419 25. Narrowed the window for a race in the daemon that could cause it to ignore
2420 SIGCHLD signals. This is not a major problem, because they are used only to
2421 wake it up if nothing else does.
2423 26. A malformed maildirsize file could cause Exim to calculate negative values
2424 for the mailbox size or file count. Odd effects could occur as a result.
2425 The maildirsize information is now recalculated if the size or filecount
2428 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2429 support for a long time. Removed HAVE_SYS_VFS_H.
2431 28. Installed the latest version of exipick from John Jetmore.
2433 29. In an address list, if the pattern was not a regular expression, an empty
2434 subject address (from a bounce message) matched only if the pattern was an
2435 empty string. Non-empty patterns were not even tested. This was the wrong
2436 because it is perfectly reasonable to use an empty address as part of a
2437 database query. An empty address is now tested by patterns that are
2438 lookups. However, all the other forms of pattern expect the subject to
2439 contain a local part and a domain, and therefore, for them, an empty
2440 address still always fails if the pattern is not itself empty.
2442 30. Exim went into a mad DNS loop when attempting to do a callout where the
2443 host was specified on an smtp transport, and looking it up yielded more
2444 than one IP address.
2446 31. Re-factored the code for checking spool and log partition space into a
2447 function that finds that data and another that does the check. The former
2448 is then used to implement four new variables: $spool_space, $log_space,
2449 $spool_inodes, and $log_inodes.
2451 32. The RFC2047 encoding function was originally intended for short strings
2452 such as real names; it was not keeping to the 75-character limit for
2453 encoded words that the RFC imposes. It now respects the limit, and
2454 generates multiple encoded words if necessary. To be on the safe side, I
2455 have increased the buffer size for the ${rfc2047: expansion operator from
2458 33. It is now permitted to omit both strings after an "if" condition; if the
2459 condition is true, the result is "true". As before, when the second string
2460 is omitted, a false condition yields an empty string. This makes it less
2461 cumbersome to write custom ACL and router conditions.
2463 34. Failure to deliver a bounce message always caused it to be frozen, even if
2464 there was an errors_to setting on the router. The errors_to setting is now
2467 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2468 canonical form (fully expanded) before being placed in
2469 $sender_host_address.
2471 36. The table in the code that translates DNS record types into text (T_A to
2472 "A" for instance) was missing entries for NS and CNAME. It is just possible
2473 that this could have caused confusion if both these types were looked up
2474 for the same domain, because the text type is used as part of Exim's
2475 per-process caching. But the chance of anyone hitting this buglet seems
2478 37. The dnsdb lookup has been extended in a number of ways.
2480 (1) There is a new type, "zns", which walks up the domain tree until it
2481 finds some nameserver records. It should be used with care.
2483 (2) There is a new type, "mxh", which is like "mx" except that it returns
2484 just the host names, not the priorities.
2486 (3) It is now possible to give a list of domains (or IP addresses) to be
2487 looked up. The behaviour when one of the lookups defers can be
2488 controlled by a keyword.
2490 (4) It is now possible to specify the separator character for use when
2491 multiple records are returned.
2493 38. The dnslists ACL condition has been extended: it is now possible to supply
2494 a list of IP addresses and/or domains to be looked up in a particular DNS
2497 39. Added log_selector=+queue_time_overall.
2499 40. When running the queue in the test harness, wait just a tad after forking a
2500 delivery process, to get repeatability of debugging output.
2502 41. Include certificate and key file names in error message when GnuTLS fails
2503 to set them up, because the GnuTLS error message doesn't include the name
2504 of the failing file when there is a problem reading it.
2506 42. Allow both -bf and -bF in the same test run.
2508 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2510 44. The "Exiscan patch" is now merged into the mainline Exim source.
2512 45. Sometimes the final signoff response after QUIT could fail to get
2513 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2514 < 0 before doing a fflush(). This bug looks as though it goes back to the
2515 introduction of TLS in release 3.20, but "sometimes" must have been rare
2516 because the tests only now provoked it.
2518 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2519 (this can affect the format of dates).
2521 47. exim_tidydb, when checking for the continued existence of a message for
2522 which it has found a message-specific retry record, was not finding
2523 messages that were in split spool directories. Consequently, it was
2524 deleting retry records that should have stayed in existence.
2526 48. Steve fixed some bugs in eximstats.
2528 49. The SPA authentication driver was not abandoning authentication and moving
2529 on to the next authenticator when an expansion was forced to fail,
2530 contradicting the general specification for all authenticators. Instead it
2531 was generating a temporary error. It now behaves as specified.
2533 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2534 (the order specifies the preference for clients). The order is now AES256,
2535 AES128, 3DES, ARCFOUR128.
2537 51. Small patch to Sieve code - explicitly set From: when generating an
2540 52. Exim crashed if a remote delivery caused a very long error message to be
2541 recorded - for instance if somebody sent an entire SpamAssassin report back
2542 as a large number of 550 error lines. This bug was coincidentally fixed by
2543 increasing the size of one of Exim's internal buffers (big_buffer) that
2544 happened as part of the Exiscan merge. However, to be on the safe side, I
2545 have made the code more robust (and fixed the comments that describe what
2548 53. Now that there can be additional text after "Completed" in log lines (if
2549 the queue_time_overall log selector is set), a one-byte patch to exigrep
2550 was needed to allow it to recognize "Completed" as not the last thing in
2553 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2554 patch that reportedly fixes this has been added. I am not expert enough to
2555 create a test for it. This is what the patch creator wrote:
2557 "I found a little strange behaviour of ldap code when working with
2558 Windows 2003 AD Domain, where users was placed in more than one
2559 Organization Units. When I tried to give exim partial DN, the exit code
2560 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2561 But simultaneously result of request was absolutely normal ldap result,
2562 so I produce this patch..."
2564 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2565 REFERENCE, so I have modified the code to exclude the patch when that macro
2568 55. Some experimental protocols are using DNS PTR records for new purposes. The
2569 keys for these records are domain names, not reversed IP addresses. The
2570 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2571 leaves it alone. Component reversal etc. now happens only for IP addresses.
2574 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2576 57. Double the size of the debug message buffer (to 2048) so that more of very
2577 long debug lines gets shown.
2579 58. The exicyclog utility now does better if the number of log files to keep
2580 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2582 59. Two changes related to the smtp_active_hostname option:
2584 (1) $smtp_active_hostname is now available as a variable.
2585 (2) The default for smtp_banner uses $smtp_active_hostname instead
2586 of $primary_hostname.
2588 60. The host_aton() function is supposed to be passed a string that is known
2589 to be a valid IP address. However, in the case of IPv6 addresses, it was
2590 not checking this. This is a hostage to fortune. Exim now panics and dies
2591 if the condition is not met. A case was found where this could be provoked
2592 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2593 components; fortuitously, this particular loophole had already been fixed
2594 by change 4.50/55 above.
2596 If there are any other similar loopholes, the new check in host_aton()
2597 itself should stop them being exploited. The report I received stated that
2598 data on the command line could provoke the exploit when Exim was running as
2599 exim, but did not say which command line option was involved. All I could
2600 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2601 running as the user.
2604 61. There was a buffer overflow vulnerability in the SPA authentication code
2605 (which came originally from the Samba project). I have added a test to the
2606 spa_base64_to_bits() function which I hope fixes it.
2609 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2610 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2612 63. The daemon start-up calls getloadavg() while still root for those OS that
2613 need the first call to be done as root, but it missed one case: when
2614 deliver_queue_load_max is set with deliver_drop_privilege. This is
2615 necessary for the benefit of the queue runner, because there is no re-exec
2616 when deliver_drop_privilege is set.
2618 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2619 This has been fixed.
2621 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2622 was in use, was not putting the data itself into the right store pool;
2623 consequently, it could be overwritten for a subsequent message in the same
2624 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2627 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2629 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2630 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2631 the function were treating the return as a boolean value, which happened to
2632 work because 0=false and not-0=true, but is not correct code.
2634 68. The host_aton() function was not handling scoped IPv6 addresses (those
2635 with, for example, "%eth0" on the end) correctly.
2637 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2638 negated items (that is, ~something) in unsigned ints. Some compilers
2639 apparently mutter when there is no cast.
2641 70. If an address verification called from an ACL failed, and did not produce a
2642 user-specific message (i.e. there was only a "system" message), nothing was
2643 put in $acl_verify_message. In this situation, it now puts the system
2646 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2647 session; change 4.31/43 added the unwanted input to the log line - except
2648 that it did not do this in the start of session case. It now does.
2650 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2651 This is wrong and can cause the other end to generate a synchronization
2652 error if it is another Exim or anything else that does the synchronization
2653 check. A QUIT command is no longer sent after a timeout.
2655 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2656 during host lookups.
2658 74. The defer_ok option of callout verification was not working if it was used
2659 when verifying addresses in header lines, that is, for this case:
2661 verify = header_sender/callout=defer_ok
2663 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2664 those file descriptors could be used for SMTP connections. If anything
2665 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2666 could be sent to the SMTP client, causing chaos. The daemon now opens
2667 stdin, stdout, and stderr to /dev/null when it puts itself into the
2670 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2671 log by default. The user can override this with suitable Perl magic.
2673 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2674 the log message when discard triggers, was not working for the DATA ACL or
2675 for the non-SMTP ACL.
2677 78. Error message wording change in sieve.c.
2679 79. If smtp_accept_max_per_host was set, the number of connections could be
2680 restricted to fewer than expected, because the daemon was trying to set up
2681 a new connection before checking whether the processes handling previous
2682 connections had finished. The check for completed processes is now done
2683 earlier. On busy systems, this bug wouldn't be noticed because something
2684 else would have woken the daemon, and it would have reaped the completed
2687 80. If a message was submitted locally by a user whose login name contained one
2688 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2689 It caused a spool format error. I have fixed the spool reading code. A
2690 related problem was that the "from" clause in the Received: line became
2691 illegal because of the space(s). It is now covered by ${quote_local_part.
2693 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2696 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2698 83. Patch from Sieve maintainer to fix unterminated string problem in
2699 "vacation" handling.
2701 84. Some minor changes to the Linux configuration files to help with other
2702 OS variants using glibc.
2704 85. One more patch for Sieve to update vacation handling to latest spec.
2707 ----------------------------------------------------
2708 See the note above about the 4.44 and 4.50 releases.
2709 ----------------------------------------------------
2715 1. Change 4.43/35 introduced a bug that caused file counts to be
2716 incorrectly computed when quota_filecount was set in an appendfile
2719 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2720 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2723 3. Give more explanation in the error message when the command for a transport
2724 filter fails to execute.
2726 4. There are several places where Exim runs a non-Exim command in a
2727 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2728 was being done only for the command run by the queryprogram router. It is
2729 now done for all such subprocesses. The other cases are: ${run, transport
2730 filters, and the commands run by the lmtp and pipe transports.
2732 5. Some older OS have a limit of 256 on the maximum number of file
2733 descriptors. Exim was using setrlimit() to set 1000 as a large value
2734 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2735 systems. I've change it so that if it can't get 1000, it tries for 256.
2737 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2738 was an oversight, and furthermore, ever since the addition of extra
2739 controls (e.g. 4.43/32), the checks on when to allow different forms of
2740 "control" were broken. There should now be diagnostics for all cases when a
2741 control that does not make sense is encountered.
2743 7. $recipients is now available in the predata ACL (oversight).
2745 8. Tidy the search cache before the fork to do a delivery from a message
2746 received from the command line. Otherwise the child will trigger a lookup
2747 failure and thereby defer the delivery if it tries to use (for example) a
2748 cached ldap connection that the parent has called unbind on.
2750 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2751 of $address_data from the recipient verification was clobbered by the
2752 sender verification.
2754 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2755 was its contents. (It was OK if the option was not defined at all.)
2757 11. A "Completed" log line is now written for messages that are removed from
2758 the spool by the -Mrm option.
2760 12. $host_address is now set to the target address during the checking of
2761 ignore_target_hosts.
2763 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2764 being passed; this would have caused $sender_host_name to have been used if
2765 matching the list had actually called for a host name (not very likely,
2766 since this list is usually IP addresses). A host name is now passed as
2769 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2770 code that specifies a non-restarting handler (typically sigaction() in
2771 modern systems) in an attempt to fix a rare and obscure crash bug.
2773 15. Narrowed the window for a race in the daemon that could cause it to ignore
2774 SIGCHLD signals. This is not a major problem, because they are used only to
2775 wake it up if nothing else does.
2777 16. A malformed maildirsize file could cause Exim to calculate negative values
2778 for the mailbox size or file count. Odd effects could occur as a result.
2779 The maildirsize information is now recalculated if the size or filecount
2782 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2783 support for a long time. Removed HAVE_SYS_VFS_H.
2785 18. Updated exipick to current release from John Jetmore.
2787 19. Allow an empty sender to be matched against a lookup in an address list.
2788 Previously the only cases considered were a regular expression, or an
2791 20. Exim went into a mad DNS lookup loop when doing a callout where the
2792 host was specified on the transport, if the DNS lookup yielded more than
2795 21. The RFC2047 encoding function was originally intended for short strings
2796 such as real names; it was not keeping to the 75-character limit for
2797 encoded words that the RFC imposes. It now respects the limit, and
2798 generates multiple encoded words if necessary. To be on the safe side, I
2799 have increased the buffer size for the ${rfc2047: expansion operator from
2802 22. Failure to deliver a bounce message always caused it to be frozen, even if
2803 there was an errors_to setting on the router. The errors_to setting is now
2806 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2807 canonical form (fully expanded) before being placed in
2808 $sender_host_address.
2810 24. Updated eximstats to version 1.33
2812 25. Include certificate and key file names in error message when GnuTLS fails
2813 to set them up, because the GnuTLS error message doesn't include the name
2814 of the failing file when there is a problem reading it.
2816 26. Expand error message when OpenSSL has problems setting up cert/key files.
2819 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2820 (this can affect the format of dates).
2822 28. exim_tidydb, when checking for the continued existence of a message for
2823 which it has found a message-specific retry record, was not finding
2824 messages that were in split spool directories. Consequently, it was
2825 deleting retry records that should have stayed in existence.
2827 29. eximstats updated to version 1.35
2828 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2829 1.35 - bugfix such that pie charts by volume are generated correctly
2831 30. The SPA authentication driver was not abandoning authentication and moving
2832 on to the next authenticator when an expansion was forced to fail,
2833 contradicting the general specification for all authenticators. Instead it
2834 was generating a temporary error. It now behaves as specified.
2836 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2837 (the order specifies the preference for clients). The order is now AES256,
2838 AES128, 3DES, ARCFOUR128.
2840 31. Small patch to Sieve code - explicitly set From: when generating an
2843 32. Exim crashed if a remote delivery caused a very long error message to be
2844 recorded - for instance if somebody sent an entire SpamAssassin report back
2845 as a large number of 550 error lines. This bug was coincidentally fixed by
2846 increasing the size of one of Exim's internal buffers (big_buffer) that
2847 happened as part of the Exiscan merge. However, to be on the safe side, I
2848 have made the code more robust (and fixed the comments that describe what
2851 33. Some experimental protocols are using DNS PTR records for new purposes. The
2852 keys for these records are domain names, not reversed IP addresses. The
2853 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2854 leaves it alone. Component reversal etc. now happens only for IP addresses.
2857 34. The host_aton() function is supposed to be passed a string that is known
2858 to be a valid IP address. However, in the case of IPv6 addresses, it was
2859 not checking this. This is a hostage to fortune. Exim now panics and dies
2860 if the condition is not met. A case was found where this could be provoked
2861 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2862 components; fortuitously, this particular loophole had already been fixed
2863 by change 4.50/55 or 4.44/33 above.
2865 If there are any other similar loopholes, the new check in host_aton()
2866 itself should stop them being exploited. The report I received stated that
2867 data on the command line could provoke the exploit when Exim was running as
2868 exim, but did not say which command line option was involved. All I could
2869 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2870 running as the user.
2873 35. There was a buffer overflow vulnerability in the SPA authentication code
2874 (which came originally from the Samba project). I have added a test to the
2875 spa_base64_to_bits() function which I hope fixes it.
2878 36. The daemon start-up calls getloadavg() while still root for those OS that
2879 need the first call to be done as root, but it missed one case: when
2880 deliver_queue_load_max is set with deliver_drop_privilege. This is
2881 necessary for the benefit of the queue runner, because there is no re-exec
2882 when deliver_drop_privilege is set.
2884 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2885 was in use, was not putting the data itself into the right store pool;
2886 consequently, it could be overwritten for a subsequent message in the same
2887 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2890 38. Sometimes the final signoff response after QUIT could fail to get
2891 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2892 < 0 before doing a fflush(). This bug looks as though it goes back to the
2893 introduction of TLS in release 3.20, but "sometimes" must have been rare
2894 because the tests only now provoked it.
2900 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2901 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2902 Now it returns an integer. A number of calls were still expecting a T/F
2903 return. Fortuitously, in all cases, the tests worked in OK situations,
2904 which is the norm. However, things would have gone wrong on any write
2905 failures on the smtp file descriptor. This function is used when sending
2906 messages over SMTP and also when doing verify callouts.
2908 2. When Exim is called to do synchronous delivery of a locally submitted
2909 message (the -odf or -odi options), it no longer closes stderr before doing
2912 3. Implemented the mua_wrapper option.
2914 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2916 5. Implemented the functions header_remove(), header_testname(),
2917 header_add_at_position(), and receive_remove_recipient(), and exported them
2920 6. If an ACL "warn" statement specified the addition of headers, Exim already
2921 inserted X-ACL-Warn: at the start if there was no header name. However, it
2922 was not making this test for the second and subsequent header lines if
2923 there were newlines in the string. This meant that an invalid header could
2924 be inserted if Exim was badly configured.
2926 7. Allow an ACL "warn" statement to add header lines at the start or after all
2927 the Received: headers, as well as at the end.
2929 8. Added the rcpt_4xx retry error code.
2931 9. Added postmaster_mailfrom=xxx to callout verification option.
2933 10. Added mailfrom=xxxx to the callout verification option, for verify=
2936 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2937 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2939 12. Inserted some casts to stop certain compilers warning when using pointer
2940 differences as field lengths or precisions in printf-type calls (mostly
2941 affecting debugging statements).
2943 13. Added optional readline() support for -be (dynamically loaded).
2945 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2946 same clock tick as a message's arrival, so that its received time was the
2947 same as the "first fail" time on the retry record, and that message
2948 remained on the queue past the ultimate address timeout, every queue runner
2949 would try a delivery (because it was past the ultimate address timeout) but
2950 after another failure, the ultimate address timeout, which should have then
2951 bounced the address, did not kick in. This was a "< instead of <=" error;
2952 in most cases the first failure would have been in the next clock tick
2953 after the received time, and all would be well.
2955 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2956 being recognized when the domain list was tested by the match_domain
2957 condition in an expansion string.
2959 16. Added the ${str2b64: operator.
2961 17. Exim was always calling setrlimit() to set a large limit for the number of
2962 processes, without checking whether the existing limit was already
2963 adequate. (It did check for the limit on file descriptors.) Furthermore,
2964 errors from getrlimit() and setrlimit() were being ignored. Now they are
2965 logged to the main and panic logs, but Exim does carry on, to try to do its
2966 job under whatever limits there are.
2968 18. Imported PCRE 5.0.
2970 19. Trivial typo in log message " temporarily refused connection" (the leading
2973 20. If the log selector return_path_on_delivery was set and an address was
2974 redirected to /dev/null, the delivery process crashed because it assumed
2975 that a return path would always be set for a "successful" delivery. In this
2976 case, the whole delivery is bypassed as an optimization, and therefore no
2979 21. Internal re-arrangement: the function for sending a challenge and reading
2980 a response while authentication was assuming a zero-terminated challenge
2981 string. It's now changed to take a pointer and a length, to allow for
2982 binary data in such strings.
2984 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2986 23. Exim was not respecting finduser_retries when seeking the login of the
2987 uid under which it was called; it was always trying 10 times. (The default
2988 setting of finduser_retries is zero.) Also, it was sleeping after the final
2989 failure, which is pointless.
2991 24. Implemented tls_on_connect_ports.
2993 25. Implemented acl_smtp_predata.
2995 26. If the domain in control=submission is set empty, Exim assumes that the
2996 authenticated id is a complete email address when it generates From: or
2997 Sender: header lines.
2999 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3000 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3001 chown and chgrp in /bin and hostname in /usr/bin.
3003 28. Exim was keeping the "process log" file open after each use, just as it
3004 does for the main log. This opens the possibility of it remaining open for
3005 long periods when the USR1 signal hits a daemon. Occasional processlog
3006 errors were reported, that could have been caused by this. Anyway, it seems
3007 much more sensible not to leave this file open at all, so that is what now
3010 29. The long-running daemon process does not normally write to the log once it
3011 has entered its main loop, and it closes the log before doing so. This is
3012 so that log files can straightforwardly be renamed and moved. However,
3013 there are a couple of unusual error situations where the daemon does write
3014 log entries, and I had neglected to close the log afterwards.
3016 30. The text of an SMTP error response that was received during a remote
3017 delivery was being truncated at 512 bytes. This is too short for some of
3018 the long messages that one sometimes sees. I've increased the limit to
3021 31. It is now possible to make retry rules that apply only when a message has a
3022 specific sender, in particular, an empty sender.
3024 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3025 it possible to be selective about when SMTP synchronization is enforced.
3027 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3029 32. Implemented hosts_connection_nolog.
3031 33. Added an ACL for QUIT.
3033 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3036 35. Added mailbox_size and mailbox_filecount to appendfile.
3038 36. Added control = no_multiline_responses to ACLs.
3040 37. There was a bug in the logic of the code that waits for the clock to tick
3041 in the case where the clock went backwards by a substantial amount such
3042 that the microsecond fraction of "now" was more than the microsecond
3043 fraction of "then" (but the whole seconds number was less).
3045 38. Added support for the libradius Radius client library this is found on
3046 FreeBSD (previously only the radiusclient library was supported).
3052 1. When certain lookups returned multiple values in the form name=value, the
3053 quoting of the values was not always being done properly. Specifically:
3054 (a) If the value started with a double quote, but contained no whitespace,
3056 (b) If the value contained whitespace other than a space character (i.e.
3057 tabs or newlines or carriage returns) it was not quoted.
3058 This fix has been applied to the mysql and pgsql lookups by writing a
3059 separate quoting function and calling it from the lookup code. The fix
3060 should probably also be applied to nisplus, ibase and oracle lookups, but
3061 since I cannot test any of those, I have not disturbed their existing code.
3063 2. A hit in the callout cache for a specific address caused a log line with no
3064 reason for rejecting RCPT. Now it says "Previous (cached) callout
3065 verification failure".
3067 3. There was an off-by-one bug in the queryprogram router. An over-long
3068 return line was truncated at 256 instead of 255 characters, thereby
3069 overflowing its buffer with the terminating zero. As well as fixing this, I
3070 have increased the buffer size to 1024 (and made a note to document this).
3072 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3073 when Exim is waiting for an SMTP response from a remote server, Exim
3074 restarts its select() call on the socket, thereby resetting its timeout.
3075 This is not a problem when such interrupts are rare. Somebody set up a cron
3076 job to run exiwhat every 2 minutes, which is less than the normal select()
3077 timeout (5 or 10 minutes). This meant that the select() timeout never
3078 kicked in because it was always reset. I have fixed this by comparing the
3079 time when an interrupt arrives with the time at the start of the first call
3080 to select(). If more time than the timeout has elapsed, the interrupt is
3081 treated as a timeout.
3083 5. Some internal re-factoring in preparation for the addition of Sieve
3084 extensions (by MH). In particular, the "personal" test is moved to a
3085 separate function, and given an option for scanning Cc: and Bcc: (which is
3086 not set for Exim filters).
3088 6. When Exim created an email address using the login of the caller as the
3089 local part (e.g. when creating a From: or Sender: header line), it was not
3090 quoting the local part when it contained special characters such as @.
3092 7. Installed new OpenBSD configuration files.
3094 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3095 try to make them clearer.
3097 9. Callout options, other than the timeout value, were being ignored when
3098 verifying sender addresses in header lines. For example, when using
3100 verify = header_sender/callout=no_cache
3102 the cache was (incorrectly) being used.
3104 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3105 only the Cygwin environment.
3107 11. When return_path_on_delivery was set as a log selector, if different remote
3108 addresses in the same message used different return paths and parallel
3109 remote delivery occurred, the wrong values would sometimes be logged.
3110 (Whenever a remote delivery process finished, the return path value from
3111 the most recently started remote delivery process was logged.)
3113 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3114 lines when AUTH and/or TLS are in use. This is the "received protocol"
3115 field. Exim used to use "asmtp" for authenticated SMTP, without any
3116 indication (in the protocol name) for TLS use. Now it follows the RFC and
3117 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3118 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3119 names appear in log lines as well as in Received: header lines.
3121 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3122 extensions, and comparison tests, and to fix some bugs.
3124 14. Changes to the "personal" filter test:
3126 (1) The test was buggy in that it was just doing the equivalent of
3127 "contains" tests on header lines. For example, if a user's address was
3128 anne@some.where, the "personal" test would incorrectly be true for
3130 To: susanne@some.where
3132 This test is now done by extracting each address from the header in turn,
3133 and checking the entire address. Other tests that are part of "personal"
3134 are now done using regular expressions (for example, to check local parts
3135 of addresses in From: header lines).
3137 (2) The list of non-personal local parts in From: addresses has been
3138 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3139 taken from the Sieve specification recommendations.
3141 (3) If the message contains any header line starting with "List-" it is
3142 treated as non-personal.
3144 (4) The test for "circular" in the Subject: header line has been removed
3145 because it now seems ill-conceived.
3147 15. Minor typos in src/EDITME comments corrected.
3149 16. Installed latest exipick from John Jetmore.
3151 17. If headers_add on a router specified a text string that was too long for
3152 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3153 of string_sprintf() is now avoided.
3155 18. $message_body_size was not set (it was always zero) when running the DATA
3156 ACL and the local_scan() function.
3158 19. For the "mail" command in an Exim filter, no default was being set for
3159 the once_repeat time, causing a random time value to be used if "once" was
3160 specified. (If the value happened to be <= 0, no repeat happened.) The
3161 default is now 0s, meaning "never repeat". The "vacation" command was OK
3162 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3163 (I found it when inspecting the code).
3165 20. There is now an overall timeout for performing a callout verification. It
3166 defaults to 4 times the callout timeout, which applies to individual SMTP
3167 commands during the callout. The overall timeout applies when there is more
3168 than one host that can be tried. The timeout is checked before trying the
3169 next host. This prevents very long delays if there are a large number of
3170 hosts and all are timing out (e.g. when the network connections are timing
3171 out). The value of the overall timeout can be changed by specifying an
3172 additional sub-option for "callout", called "maxwait". For example:
3174 verify = sender/callout=5s,maxwait=20s
3176 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3177 to the end before writing, but this should make it even safer).
3179 22. Exim was forgetting that it had advertised PIPELINING for the second and
3180 subsequent messages on an SMTP connection. It was also not resetting its
3181 memory on STARTTLS and an internal HELO.
3183 23. When Exim logs an SMTP synchronization error within a session, it now
3184 records whether PIPELINING has been advertised or not.
3186 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3187 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3188 rather than long int.
3190 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3192 26. Added the never_mail option to autoreply.
3198 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3199 crash if the getsockname() call failed; this can happen if a connection is
3200 closed very soon after it is established. The problem was simply in the
3201 order in which certain operations were done, causing Exim to try to write
3202 to the SMTP stream before it had set up the file descriptor. The bug has
3203 been fixed by making things happen in the correct order.
3209 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3210 before the connection was closed, thus losing the rejection response.
3212 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3213 some early Solaris releases, but causes trouble in current releases where
3214 socklen_t is defined.
3216 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3219 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3222 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3223 that releasing the top of it at the end releases what was used for sub-
3224 expansions (unless the block got too big). However, discard this block if
3225 the first thing is a variable or header, so that we can use its block when
3226 it is dynamic (useful for very large $message_headers, for example).
3228 6. Lookups now cache *every* query, not just the most recent. A new, separate
3229 store pool is used for this. It can be recovered when all lookup caches are
3230 flushed. Lookups now release memory at the end of their result strings.
3231 This has involved some general refactoring of the lookup sources.
3233 7. Some code has been added to the store_xxx() functions to reduce the amount
3234 of flapping under certain conditions.
3236 8. log_incoming_interface used to affect only the <= reception log lines. Now
3237 it causes the local interface and port to be added to several more SMTP log
3238 lines, for example "SMTP connection from", and rejection lines.
3240 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3242 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3244 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3245 could be overwritten at the end of the current message (or the start of a
3246 new message if it was set in a HELO ACL). The value is now preserved for
3247 the duration of the SMTP connection.
3249 12. If a transport had a headers_rewrite setting, and a matching header line
3250 contained an unqualified address, that address was qualified, even if it
3251 did not match any rewriting rules. The underlying bug was that the values
3252 of the flags that permit the existence of unqualified sender and recipient
3253 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3254 non-local messages, and by -bnq for local messages) were not being
3255 preserved with the message after it was received.
3257 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3258 "next input=" as part of the text comprising the host identity instead of
3259 the correct text. The code was using the same buffer for two different
3260 strings. However, depending on which order the printing function evaluated
3261 its arguments, the bug did not always show up. Under Linux, for example, my
3262 test suite worked just fine.
3264 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3265 doesn't work with some older versions of Perl. It has been changed to "my",
3266 which in any case is probably the better facility to use.
3268 15. A really picky compiler found some instances of statements for creating
3269 error messages that either had too many or two few arguments for the format
3272 16. The size of the buffer for calls to the DNS resolver has been increased
3273 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3274 for addresses that have a lot of PTR records. This alleviates a problem; it
3275 does not fully solve it.
3277 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3278 buffer now truncates the list and logs the incident, which is the same
3279 action as happens when Exim is looking up a host name and its aliases.
3280 Previously in this situation something unpredictable would happen;
3281 sometimes it was "internal error: store_reset failed".
3283 18. If a server dropped the connection unexpectedly when an Exim client was
3284 using GnuTLS and trying to read a response, the client delivery process
3285 crashed while trying to generate an error log message.
3287 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3288 string, for example:
3290 warn message = H1: something\nH2: something
3292 the text was added as a single header line from Exim's point of view
3293 though it ended up OK in the delivered message. However, searching for the
3294 second and subsequent header lines using $h_h2: did not work. This has been
3295 fixed. Similarly, if a system filter added multiple headers in this way,
3296 the routers could not see them.
3298 20. Expanded the error message when iplsearch is called with an invalid key to
3299 suggest using net-iplsearch in a host list.
3301 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3302 ACLs are no longer actually imposed (and a message to that effect is
3305 22. If a "gecos" field in a passwd entry contained escaped characters, in
3306 particular, if it contained a \" sequence, Exim got it wrong when building
3307 a From: or a Sender: header from that name. A second bug also caused
3308 incorrect handling when an unquoted " was present following a character
3309 that needed quoting.
3311 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3312 was not being matched caselessly.
3314 24. Arranged for all hyphens in the exim.8 source to be escaped with
3317 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3318 properly. Recipient callout cache records were still being keyed to include
3319 the sender, even when use_sender was set false. This led to far more
3320 callouts that were necessary. The sender is no longer included in the key
3321 when use_sender is false.
3323 26. Added "control = submission" modifier to ACLs.
3325 27. Added the ${base62d: operator to decode base 62 numbers.
3327 28. dnsdb lookups can now access SRV records.
3329 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3330 the configuration file.
3332 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3333 (-v) mode. This makes the output for a verbose queue run more intelligible.
3335 31. Added a use_postmaster feature to recipient callouts.
3337 32. Added the $body_zerocount variable, containing the number of binary zero
3338 bytes in the message body.
3340 33. The time of last modification of the "new" subdirectory is now used as the
3341 "mailbox time last read" when there is a quota error for a maildir
3344 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3346 35. Added +ignore_unknown as a special item in host lists.
3348 36. Code for decoding IPv6 addresses in host lists is now included, even if
3349 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3350 address was recognized as an IP address, but was then not correctly decoded
3351 into binary, causing unexpected and incorrect effects when compared with
3358 1. Very minor rewording of debugging text in manualroute to say "list of
3359 hosts" instead of "hostlist".
3361 2. If verify=header_syntax was set, and a header line with an unqualified
3362 address (no domain) and a large number of spaces between the end of the
3363 name and the colon was received, the reception process suffered a buffer
3364 overflow, and (when I tested it) crashed. This was caused by some obsolete
3365 code that should have been removed. The fix is to remove it!
3367 3. When running in the test harness, delay a bit after writing a bounce
3368 message to get a bit more predictability in the log output.
3370 4. Added a call to search_tidyup() just before forking a reception process. In
3371 theory, someone could use a lookup in the expansion of smtp_accept_max_
3372 per_host which, without the tidyup, could leave open a database connection.
3374 5. Added the variables $recipient_data and $sender_data which get set from a
3375 lookup success in an ACL "recipients" or "senders" condition, or a router
3376 "senders" option, similar to $domain_data and $local_part_data.
3378 6. Moved the writing of debug_print from before to after the "senders" test
3381 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3382 problems for message scanning, either using a data ACL, or using
3383 local_scan() because the Received: header was not generated till after they
3384 were called (in order to set the time as the time of reception completion).
3385 I have revised the way this works. The header is now generated after the
3386 body is received, but before the ACL or local_scan() are called. After they
3387 are run, the timestamp in the header is updated.
3393 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3394 before starting a queue runner without re-exec. This happened only when
3395 deliver_drop_privilege was set or when the Exim user was set to root. The
3396 effect of the bug was that timeouts during subsequent deliveries caused
3397 crashes instead of being properly handled. The handler is now left at its
3398 default (and expected) setting.
3400 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3401 message, again when deliver_drop_privilege is set or Exim is run as root.
3402 The bug described in (1) was not present in this case, but the tidying up
3403 of the other signals was missing. I have made the two cases consistent.
3405 3. The ignore_target_hosts setting on a manualroute router was being ignored
3406 for hosts that were looked up using the /MX notation.
3408 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3411 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3412 operated on the sender address. After changing the $sender_address to <>
3413 for the sender address verify, Exim was re-instated it as the original
3414 (before rewriting) address, but remembering that it had rewritten it, so it
3415 wasn't rewriting it again. This bug also had the effect of breaking the
3416 sender address verification caching when the sender address was rewritten.
3418 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3419 This has been changed so that if the ip literal address matches
3420 ignore_target_hosts, the router declines.
3422 7. Added expansion conditions match_domain, match_address, and match_local_
3423 part (NOT match_host).
3425 8. The placeholder for the Received: header didn't have a length field set.
3427 9. Added code to Exim itself and to exim_lock to test for a specific race
3428 condition that could lead to file corruption when using MBX delivery. The
3429 issue is with the lockfile that is created in /tmp. If this file is removed
3430 after a process has opened it but before that process has acquired a lock,
3431 there is the potential for a second process to recreate the file and also
3432 acquire a lock. This could lead to two Exim processes writing to the file
3433 at the same time. The added code performs the same test as UW imapd; it
3434 checks after acquiring the lock that its file descriptor still refers to
3435 the same named file.
3437 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3438 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3441 11. Added the smtp_active_hostname option. If used, this will typically be made
3442 to depend on the incoming interface address. Because $interface_address is
3443 not set up until the daemon has forked a reception process, error responses
3444 that can happen earlier (such as "too many connections") no longer contain
3447 12. If an expansion in a condition on a "warn" statement fails because a lookup
3448 defers, the "warn" statement is abandoned, and the next ACL statement is
3449 processed. Previously this caused the whole ACL to be aborted.
3451 13. Added the iplsearch lookup type.
3453 14. Added ident_timeout as a log selector.
3455 15. Added tls_certificate_verified as a log selector.
3457 16. Added a global option tls_require_ciphers (compare the smtp transport
3458 option of the same name). This controls incoming TLS connections.
3460 17. I finally figured out how to make tls_require_ciphers do a similar thing
3461 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3462 before starting the TLS session.
3464 18. Tabs are now shown as \t in -bP output.
3466 19. If the log selector return_path_on_delivery was set, Exim crashed when
3467 bouncing a message because it had too many Received: header lines.
3469 20. If two routers both had headers_remove settings, and the first one included
3470 a superfluous trailing colon, the final name in the first list and the
3471 first name in the second list were incorrectly joined into one item (with a
3472 colon in the middle).
3478 1. Added -C and -D options to the exinext utility, mainly to make it easier
3479 to include in the automated testing, but these could be helpful when
3480 multiple configurations are in use.
3482 2. The exinext utility was not formatting the output nicely when there was
3483 an alternate port involved in the retry record key, nor when there was a
3484 message id as well (for retries that were specific to a specific message
3485 and a specific host). It was also confused by IPv6 addresses, because of
3486 the additional colons they contain. I have fixed the IPv4 problem, and
3487 patched it up to do a reasonable job for IPv6.
3489 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3490 delivery, the log line now contains "pipelined" if PIPELINING was used.
3492 4. An SMTP transport process used to panic and die if the bind() call to set
3493 an explicit outgoing interface failed. This has been changed; it is now
3494 treated in the same way as a connect() failure.
3496 5. A reference to $sender_host_name in the part of a conditional expansion
3497 that was being skipped was still causing a DNS lookup. This no longer
3500 6. The def: expansion condition was not recognizing references to header lines
3501 that used bh_ and bheader_.
3503 7. Added the _cache feature to named lists.
3505 8. The code for checking quota_filecount in the appendfile transport was
3506 allowing one more file than it should have been.
3508 9. For compatibility with Sendmail, the command line option
3516 and sets the incoming protocol and host name (for trusted callers). The
3517 host name and its colon can be omitted when only the protocol is to be set.
3518 Note the Exim already has two private options, -pd and -ps, that refer to
3519 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3520 "s", but I don't think that's a major issue.
3522 10. A number of refactoring changes to the code, none of which should affect
3525 (a) The number of logging options was getting close to filling up the
3526 32-bit word that was used as a bit map. I have split them into two classes:
3527 those that are passed in the argument to log_write(), and those that are
3528 only ever tested independently outside of that function. These are now in
3529 separate 32-bit words, so there is plenty of room for expansion again.
3530 There is no change in the user interface or the logging behaviour.
3532 (b) When building, for example, log lines, the code previously used a
3533 macro that called string_cat() twice, in order to add two strings. This is
3534 not really sufficiently general. Furthermore, there was one instance where
3535 it was actually wrong because one of the argument was used twice, and in
3536 one call a function was used. (As it happened, calling the function twice
3537 did not affect the overall behaviour.) The macro has been replaced by a
3538 function that can join an arbitrary number of extra strings onto a growing
3541 (c) The code for expansion conditions now uses a table and a binary chop
3542 instead of a serial search (which was left over from when there were very
3543 few conditions). Also, it now recognizes conditions like "pam" even when
3544 the relevant support is not compiled in: a suitably worded error message is
3545 given if an attempt is made to use such a condition.
3547 11. Added ${time_interval:xxxxx}.
3549 12. A bug was causing one of the ddress fields not to be passed back correctly
3550 from remote delivery subprocesses. The field in question was not being
3551 subsequently used, so this caused to problems in practice.
3553 13. Added new log selectors queue_time and deliver_time.
3555 14. Might have fixed a bug in maildirsizefile handling that threw up
3556 "unexpected character" debug warnings, and recalculated the data
3557 unnecessarily. In any case, I expanded the warning message to give more
3560 15. Added the message "Restricted characters in address" to the statements in
3561 the default ACL that block characters like @ and % in local parts.
3563 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3564 Three changes have been made:
3566 (a) There was a serious bug; a negative response to MAIL caused the whole
3567 recipient domain to be cached as invalid, thereby blocking all messages
3568 to all local parts at the same domain, from all senders. This bug has
3569 been fixed. The domain is no longer cached after a negative response to
3570 MAIL if the sender used is not empty.
3572 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3575 (c) A new callout option, "use_sender" has been added for people who want
3576 the modified behaviour.
3582 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3585 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3586 indeed breaks things for older releases.
3588 3. Added additional logging to the case where there is a problem reading data
3589 from a filter that is running in a subprocess using a pipe, in order to
3590 try to track down a specific problem.
3592 4. Testing facility fudge: when running in the test harness and attempting
3593 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3594 getting "No route to host". Convert this to a timeout.
3596 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3599 6. Some OS don't have socklen_t but use size_t instead. This affects the
3600 fifth argument of getsockopt() amongst other things. This is now
3601 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3602 can be set for individual OS. I have set it for SunOS5, OSF1, and
3603 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3604 some earlier ones do not.
3606 7. Change 4.30/15 was not doing the test caselessly.
3608 8. The standard form for an IPv6 address literal was being rejected by address
3609 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3610 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3611 this, as well as the form without the "IPv6" on the front (but only when
3612 address literals are enabled, of course).
3614 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3616 10. Exim crashed if a message with an empty sender address specified by -f
3617 encountered a router with an errors_to setting. This could be provoked only
3618 by a command such as
3622 where an empty string was supplied; "<>" did not hit this bug.
3624 11. Installed PCRE release 4.5.
3626 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3627 remained set. It is now erased.
3629 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3630 times from message ids (which are base 36 rather than the normal 62).
3632 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3633 were being counted as actual protocol errors, and logged if the log
3634 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3635 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3636 and DATA following a set of rejected RCPTs do not count as protocol errors.
3637 In other words, Exim assumes they were pipelined, though this may not
3638 actually be the case. Of course, in all cases the client gets an
3639 appropriate error code.
3641 15. If a lookup fails in an ACL condition, a message about the failure may
3642 be available; it is used if testing the ACL cannot continue, because most
3643 such messages specify what the cause of the deferral is. However, some
3644 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3645 that caused an old message to be retained and used if a later statement
3646 caused a defer, replacing the real cause of the deferral.
3648 16. If an IP address had so many PTR records that the DNS lookup buffer
3649 was not large enough to hold them, Exim could crash while trying to process
3650 the truncated data. It now detects and logs this case.
3652 17. Further to 4.21/58, another change has been made: if (and only if) the
3653 first line of a message (the first header line) ends with CRLF, a bare LF
3654 in a subsequent header line has a space inserted after it, so as not to
3655 terminate the header.
3657 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3658 unnecessarily, used atoi() instead of strtol(), and didn't check the
3659 termination when getting file sizes from file names by regex.
3661 19. Completely re-implemented the support for maildirsize files, in the light
3662 of a number of problems with the previous contributed implementation
3663 (4.30/29). In particular:
3665 . If the quota is zero, the maildirsize file is maintained, but no quota is
3668 . If the maildir directory does not exist, it is created before any attempt
3669 to write a maildirsize file.
3671 . The quota value in the file is just a cache; if the quota is changed in
3672 the transport, the new value overrides.
3674 . A regular expression is available for excluding directories from the
3677 20. The autoreply transport checks the characters in options that define the
3678 message's headers; it allows continued headers, but it was checking with
3679 isspace() after an embedded newline instead of explicitly looking for a
3682 21. If all the "regular" hosts to which an address was routed had passed their
3683 expiry times, and had not reached their retry times, the address was
3684 bounced, even if fallback hosts were defined. Now Exim should go on to try
3687 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3688 equivalent code in the SMTP transport. Some hosts send humungous responses
3689 to HELO/EHLO, more than 1024 it seems.
3691 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3692 gives compiler warnings in some environments. I've now done it "properly",
3695 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3696 (because of problems with the built-in one) was declared to return uschar *
3697 instead of char *, causing compiler failure.
3699 25. Fixed a file descriptor leak when processing alias/forward files.
3701 26. Fixed a minor format string issue in dbfn.c.
3703 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3705 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3706 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3709 29. When a sender address is verified, it is cached, to save repeating the test
3710 when there is more than one recipient in a message. However, when the
3711 verification involves a callout, it is possible for different callout
3712 options to be set for different recipients. It is too complicated to keep
3713 track of this in the cache, so now Exim always runs a verification when a
3714 callout is required, relying on the callout cache for the optimization.
3715 The overhead is duplication of the address routing, but this should not be
3718 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3719 to be verified with callout=postmaster, and the main callout worked but the
3720 postmaster check failed, the verification correctly failed. However, if a
3721 subsequent RCPT command asked for sender verification *without* the
3722 postmaster check, incorrect caching caused this verification also to fail,
3725 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3726 it was not caching the DNS options (qualify_single, search_parents) that
3727 were used when the lookup failed. A subsequent lookup with different
3728 options therefore always gave the same answer, though there were cases
3729 where it should not have. (Example: a "domains = !$mx_any" option on a
3730 dnslookup router: the "domains" option is always processed without any
3731 widening, but the router might have qualify_single set.) Now Exim uses the
3732 cached value only when the same options are set.
3734 32. Added John Jetmore's "exipick" utility to the distribution.
3736 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3737 than a timeout (e.g. a certificate is required, and is not provided), an
3738 Exim server now closes the connection immediately. Previously it waited for
3739 the client to close - but if the client is SSL, it seems that they each
3740 wait for each other, leading to a delay before one of them times out.
3742 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3743 maintained 0.8.x compatibility because I don't think many are using it, and
3744 it is clearly obsolete.
3746 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3749 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3750 client certificate was expired. A simple patch fixes this, though I don't
3751 understand the full logic of why the verify callback is called multiple
3754 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3755 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3756 which causes problems with some clients (such as the Certicom SSL Plus
3757 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3758 disables the coutermeasure allowing Eudora to connect."
3760 38. Exim was not checking that a write() to a log file succeeded. This could
3761 lead to Bad Things if a log got too big, in particular if it hit a file
3762 size limit. Exim now panics and dies if it cannot write to a log file, just
3763 as it does if it cannot open a log file.
3765 39. Modified OS/Makefile-Linux so that it now contains
3767 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3769 The two -D definitions ensure that Exim is compiled with large file
3770 support, which makes it possible to handle log files that are bigger than
3773 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3774 instance) a domain was checked against a named list that involved a lookup,
3775 causing $domain_data to be set, then another domain was checked against the
3776 same list, then the first domain was re-checked, the value of $domain_data
3777 after the final check could be wrong. In particular, if the second check
3778 failed, it could be set empty. This bug probably also applied to
3781 41. The strip_trailing_dot option was not being applied to the address given
3782 with the -f command-line option.
3784 42. The code for reading a message's header from the spool was incrementing
3785 $received_count, but never initializing it. This meant that the value was
3786 incorrect (doubled) while delivering a message in the same process in which
3787 it was received. In the most common configuration of Exim, this never
3788 happens - a fresh exec is done - but it can happen when
3789 deliver_drop_privilege is set.
3791 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3792 it now includes up to 150 characters of the unexpected data in the log
3795 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3796 and building data strings. The size of both of these buffers was 10 000
3797 bytes - far larger than anybody would *ever* want, thought I. Needless to
3798 say, somebody hit the limit. I have increased the maximum line length to
3799 20 000 and the maximum data length of concatenated lines to 100 000. I have
3800 also fixed two bugs, because there was no checking on these buffers. Tsk,
3801 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3802 buffer is too small.
3804 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3805 lsearch lookups. Now it does.
3807 46. When parsing a route_list item in a manualroute router, a fixed-length
3808 buffer was used for the list of hosts. I made this 1024 bytes long,
3809 thinking that nobody would ever have a list of hosts that long. Wrong.
3810 Somebody had a whole pile of complicated expansion conditions, and the
3811 string was silently truncated, leading to an expansion error. It turns out
3812 that it is easier to change to an unlimited length (owing to other changes
3813 that have happened since this code was originally written) than to build
3814 structure for giving a limitation error. The length of the item that
3815 expands into the list of hosts is now unlimited.
3817 47. The lsearch lookup could not handle data where the length of text line was
3818 more than 4095 characters. Such lines were truncated, leading to shortened
3819 data being returned. It should now handle lines of any length.
3821 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3822 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3825 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3826 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3827 no longer happens in comment lines. A list of replacements is now placed
3828 at the head of all of the source files, except those whose only change is
3829 to replace PERL_COMMAND in the very first #! line.
3831 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3832 messages on the queue, with a bottom-up merge sort, using code contributed
3833 by Michael Haardt. This should make operations like -bp somewhat faster on
3834 large queues. It won't affect queue runners, except when queue_run_in_order
3837 51. Installed eximstats 1.31 in the distribution.
3839 52. Added support for SRV lookups to the dnslookup router.
3841 53. If an ACL referred to $message_body or $message_body_end, the value was not
3842 reset for any messages that followed in the same SMTP session.
3844 54. The store-handling optimization for building very long strings was not
3845 differentiating between the different store pools. I don't think this
3846 actually made any difference in practice, but I've tidied it.
3848 55. While running the routers to verify a sender address, $sender_address
3849 was still set to the sender address. This is wrong, because when routing to
3850 send a bounce to the sender, it would be empty. Therefore, I have changed
3851 it so that, while verifying a sender address, $sender_address is set to <>.
3852 (There is no change to what happens when verifying a recipient address.)
3854 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3855 target A or AAAA records (if not already returned) without resetting the
3856 qualify_single or search_parents options of the DNS resolver. These are
3857 inappropriate in this case because the targets of MX and SRV records must
3858 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3859 target that, when qualified, matched something in the local domain. These
3860 two options are now turned off when doing these lookups.
3862 57. It seems that at least some releases of Reiserfs (which does not have the
3863 concept of a fixed number of inodes) returns zero and not -1 for the
3864 number of available inodes. This interacted badly with check_spool_inodes,
3865 which assumed that -1 was the "no such thing" setting. What I have done is
3866 to check that the total number of inodes is greater than zero before doing
3867 the test of how many are available.
3869 58. When a "warn" ACL statement has a log_message modifier, the message is
3870 remembered, and not repeated. This is to avoid a lot of repetition when a
3871 message has many recipients that cause the same warning to be written.
3872 Howewer, Exim was preserving the list of already written lines for an
3873 entire SMTP session, which doesn't seem right. The memory is now reset if a
3874 new message is started.
3876 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3877 debugging output unless log_rewrite was also set.
3879 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3880 of (int)(handle) when we know that handle contains (void *)(-1).
3882 61. The Exim daemon panic-logs an error return when it closes the incoming
3883 connection. However "connection reset by peer" seems to be common, and
3884 isn't really an error worthy of noting specially, so that particular error
3887 62. When Exim is trying to find all the local interfaces, it used to panic and
3888 die if the ioctl to get the interface flags failed. However, it seems that
3889 on at least one OS (Solaris 9) it is possible to have an interface that is
3890 included in the list of interfaces, but for which you get a failure error
3891 for this call. This happens when the interface is not "plumbed" into a
3892 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3893 failure of the "get flags" call assumes that the interface is down.
3895 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3896 makes life easier for people who are doing arithmetic on fields extracted
3897 from dates, where you often get leading zeros that should not be
3898 interpreted as octal.
3900 64. Added qualify_domain to the redirect router, to override the global
3903 65. If a pathologically long header line contained very many addresses (the
3904 report of this problem mentioned 10 000) and each of them was rewritten,
3905 Exim could use up a very large amount of memory. (It kept on making new
3906 copies of the header line as it rewrote, and never released the old ones.)
3907 At the expense of a bit more processing, the header rewriting function has
3908 been changed so that it no longer eats memory in this way.
3910 66. The generation of the Received: header has been moved from the time that a
3911 message starts to be received, to the time that it finishes. The timestamp
3912 in the Received: header should now be very close to that of the <= log
3913 line. There are two side-effects of this change:
3915 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3916 logged header lines no longer include the local Received: line, because
3917 it has not yet been created. The same applies to a copy of the message
3918 that is returned to a non-SMTP sender when a message is rejected.
3920 (b) When a filter file is tested using -bf, no additional Received: header
3921 is added to the test message. After some thought, I decided that this
3924 This change does not affect the value of $received_for. It is still set
3925 after address rewriting, but before local_scan() is called.
3927 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3929 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3930 gave an unhelpful panic error message, and a defer error. I have managed to
3931 change this behaviour so that it now rejects any supplied certificate,
3932 which seems right, as the list of acceptable certificates is empty.
3934 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3935 gave an unhelpful defer error. I have not managed to make this reject any
3936 supplied certificates, but the error message it gives is "no certificate
3937 supplied", which is not helpful.
3939 70. exigrep's output now also includes lines that are not associated with any
3940 message, but which match the given pattern. Implemented by a patch from
3941 Martin Sluka, which also tidied up the Perl a bit.
3943 71. Recipient callout verification, like sender verification, was using <> in
3944 the MAIL FROM command. This isn't really the right thing, since the actual
3945 sender may affect whether the remote host accepts the recipient or not. I
3946 have changed it to use the actual sender in the callout; this means that
3947 the cache record is now keyed on a recipient/sender pair, not just the
3948 recipient address. There doesn't seem to be a real danger of callout loops,
3949 since a callout by the remote host to check the sender would use <>.
3950 [SEE ABOVE: changed after hitting problems.]
3952 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3953 temporary errors. However, in the case of such a code being given after
3954 the end of a data transmission (i.e. after ".") Exim was failing to write
3955 a retry record for the message. (Yes, there was some broken host that was
3956 actually sending 8xx at this point.)
3958 73. An unknown lookup type in a host list could cause Exim to panic-die when
3959 the list was checked. (An example that provoked this was putting <; in the
3960 middle of a list instead of at the start.) If this happened during a DATA
3961 ACL check, a -D file could be left lying around. This kind of configuration
3962 error no longer causes Exim to die; instead it causes a defer errror. The
3963 incident is still logged to the main and panic logs.
3965 74. Buglet left over from Exim 3 conversion. The message "too many messages
3966 in one connection" was written to the rejectlog but not the mainlog, except
3967 when address rewriting (yes!) was being logged.
3969 75. Added write_rejectlog option.
3971 76. When a system filter was run not as root (that is, when system_filter_user
3972 was set), the values of the $n variables were not being returned to the
3973 main process; thus, they were not subsequently available in the $sn
3976 77. Added +return_path_on_delivery log selector.
3978 78. A connection timeout was being treated differently from recipients deferred
3979 when testing hosts_max_try with a message that was older than the host's
3980 retry timeout. (The host should not be counted, thus allowing all hosts to
3981 be tried at least once before bouncing.) This may have been the cause of an
3982 occasionally reported bug whereby a message would remain on the queue
3983 longer than the retry timeout, but would be bounced if a delivery was
3984 forced. I say "may" because I never totally pinned down the problem;
3985 setting up timeout/retry tests is difficult. See also the next item.
3987 79. The ultimate address timeout was not being applied to errors that involved
3988 a combination of host plus message (for example, a timeout on a MAIL
3989 command). When an address resolved to a number of possible hosts, and they
3990 were not all tried for each delivery (e.g. because of hosts_max_try), a
3991 message could remain on the queue longer than the retry timeout.
3993 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3996 81. Fixed an obscure SMTP outgoing bug which required at least the following
3997 conditions: (a) there was another message waiting for the same server;
3998 (b) the server returned 5xx to all RCPT commands in the first message so
3999 that the message was not completed; (c) the server dropped the connection
4000 or gave a negative response to the RSET that Exim sends to abort the
4001 transaction. The observed case was a dropped connection after DATA that had
4002 been sent in pipelining mode. That is, the server had advertised PIPELINING
4003 but was not implementing it correctly. The effect of the bug was incorrect
4004 behaviour, such as trying another host, and this could lead to a crash.
4010 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4011 and daemon.c were passed as pointers to ints; they should have been
4012 pointers to socklen_t variables (which are typically unsigned ints).
4014 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4017 3. Fixed a really odd bug that affected only the testing scheme; patching a
4018 certain fixed string in the binary changed the value of another string that
4019 happened to be identical to the end of the original first string.
4021 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4022 name", it returns that address as the IP address. On some operating
4023 systems (e.g. Solaris), it also passes back the IP address string as the
4024 "host name". However, on others (e.g. Linux), it passes back an empty
4025 string. Exim wasn't checking for this, and was changing the host name to an
4026 empty string, assuming it had been canonicized.
4028 5. Although rare, it is permitted to have more than one PTR record for a given
4029 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4030 all the names associated with an address, because they do in Solaris.
4031 However, it seems that they do not in Linux for data that comes from the
4032 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4033 I found this out when I moved to a new Linux workstation and tried to run
4034 the Exim test suite.
4036 To get round this problem I have changed the code so that it now does its
4037 own call to the DNS to look up PTR records when searching for a host name.
4038 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4039 addresses that are only in /etc/hosts are still found.
4041 This behaviour is, however, controlled by an option called host_lookup_
4042 order, which defaults to "bydns:byaddr". If people want to use the other
4043 order, or indeed, just use one or the other means of lookup, they can
4044 specify it in this variable.
4046 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4047 some operating systems, this comes back from gethostbyaddr() as an empty
4048 string, and this is what Exim used to test for. However, it seems that in
4049 other systems, "." is yielded. Exim now tests for this case too.
4051 7. The values of check_spool_space and check_log_space are now held internally
4052 as a number of kilobytes instead of an absolute number of bytes. If a
4053 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4054 kilobyte. This means that much larger values can be stored.
4056 8. Exim monitor: an attempt to get the action menu when not actually pointing
4057 at a message produces an empty menu entitled "No message selected". This
4058 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4059 no entries in it ("Shell widget menu has zero width and/or height"). So I
4060 have added a single, blank menu entry in this case.
4062 9. Added ${quote_local_part.
4064 10. MIME decoding is now applied to the contents of Subject: header lines when
4067 11. Now that a reference to $sender_host_address automatically causes a reverse
4068 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4069 host lookup before query-style lookups in lists that might use this
4070 variable. This has therefore been abolished, and the "net-" prefix is no
4071 longer necessary for query-style lookups.
4073 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4074 have been a typo for LFLAGS, so it has been changed.
4076 13. The install script calls Exim with "-C /dev/null" in order to find the
4077 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4078 to be output. Howeve, since Exim outputs its version number before the
4079 error, it didn't break the script. It just looked ugly. I fixed this by
4080 always allowing "-C /dev/null" if the caller is root.
4082 14. Ignore overlarge ACL variable number when reading spool file - insurance
4083 against a later release with more variables having written the file.
4085 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4086 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4087 this, as well as the form without the "IPv6" on the front.
4089 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4090 OS/Makefile-Darwin file.
4092 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4093 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4095 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4096 variable. However, some OS use ints for inodes. I've added cast to long int
4097 to get rid of the compiler warning.
4099 19. I had forgotten to lock out "/../" in configuration file names when
4100 ALT_CONFIG_PREFIX was set.
4102 20. Routers used for verification do not need to specify transports. However,
4103 if such a router generated a host list, and callout was configured, Exim
4104 crashed, because it could not find a port number from the (non-existent)
4105 transport. It now assumes port 25 in this circumstance.
4107 21. Added the -t option to exigrep.
4109 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4110 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4111 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4112 them is set, LOOKUP_LSEARCH is forced.
4114 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4116 24. Added sender and host information to the "rejected by local_scan()" log
4117 line; previously there was no indication of these.
4119 25. Added .include_if_exists.
4121 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4122 Linux. It turns out that not all file systems support this. Apparently some
4123 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4124 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4125 not supported on the file descriptor, is now ignored when Exim is trying to
4126 sync a directory. This applies only to Linux.
4128 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4130 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4131 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4134 29. Added support for maildirsize files from supplied patch (modified a bit).
4136 30. The use of :fail: followed by an empty string could lead Exim to respond to
4137 sender verification failures with (e.g.):
4139 550 Verification failed for <xxx>
4140 550 Sender verify failed
4142 where the first response line was missing the '-' that indicates it is not
4143 the final line of the response.
4145 31. The loop for finding the name of the user that called Exim had a hardwired
4146 limit of 10; it now uses the value of finduser_retries, which is used for
4147 all other user lookups.
4149 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4152 33. Exim was neglecting to zero errno before one call of strtol() when
4153 expanding a string and expecting an integer value. On some systems this
4154 resulted in spurious "integer overflow" errors. Also, it was casting the
4155 result into an int without checking.
4157 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4158 did not work. The code looks as if it has *never* worked, though it appears
4159 to have been documented since at least releast 1.62. I have made it work.
4161 35. The "timeout_DNS" error in retry rules, also documented since at least
4162 1.62, also never worked. As it isn't clear exactly what this means, and
4163 clearly it isn't a major issue, I have abolished the feature by treating it
4164 as "timeout", and writing a warning to the main and panic logs.
4166 36. The display of retry rules for -brt wasn't always showing the error code
4169 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4170 timeout_connect_A, timeout_connect_MX.
4172 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4173 to the empty sender.
4175 39. The daemon was not analysing the content of -oX till after it had closed
4176 stderr and disconnected from the controlling terminal. This meant that any
4177 syntax errors were only noted on the panic log, and the return code from
4178 the command was 0. By re-arranging the code a little, I've made the
4179 decoding happen first, so such errors now appear on stderr, and the return
4180 code is 1. However, the actual setting up of the sockets still happens in
4181 the disconnected process, so errors there are still only recorded on the
4184 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4185 connections (as happens on some IP stacks) was logged at start up time as
4186 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4187 it from "IPv6 and IPv4", which means that two separate sockets are being
4190 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4191 says whether AF_INET or AF_INET6 was passed as an argument.
4193 42. Exiwhat output was messed up when time zones were included in log
4196 43. Exiwhat now gives more information about the daemon's listening ports,
4197 and whether -tls-on-connect was used.
4199 44. The "port" option of the smtp transport is now expanded.
4201 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4202 silently ignored. Now an error message is written to the main and panic
4205 46. There's a new ACL modifier called "logwrite" which writes to a log file
4206 as soon as it is encountered.
4208 47. Added $local_user_uid and $local_user_gid at routing time.
4210 48. Exim crashed when trying to verify a sender address that was being
4213 49. Exim was recognizing only a space character after ".include". It now also
4214 recognizes a tab character.
4216 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4217 extracting the relevant information from the specification. The man page no
4218 longer contains scrambled data for the -d option, and I've added a section
4219 at the front about calling Exim under different names.
4221 51. Added "extra_headers" argument to the "mail" command in filter files.
4223 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4226 53. Installed eximstats 1.29.
4228 54. Added transport_filter_timeout as a generic transport option.
4230 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4231 Cc: header lines. This was required by RFC 822, but it not required by RFC
4234 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4235 incoming messages that did not have them. Now it does so only if the
4236 message originates locally, that is, if there is no associated remote host
4237 address. When Resent- header lines are present, this applies to the Resent-
4238 lines rather than the non-Resent- lines.
4240 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4241 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4243 58. Messages for configuration errors now include the name of the main
4244 configuration file - useful now that there may be more than one file in a
4245 list (.included file names were always shown).
4247 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4248 for those rare installations that do not start the daemon as root or run it
4249 setuid root. I've cut out the call to initgroups() if the daemon is not
4252 60. The Exim user and group can now be bound into the binary as text strings
4253 that are looked up at the start of Exim's processing.
4255 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4257 62. Added $mailstore_basename variable.
4259 63. Installed patch to sieve.c from Michael Haardt.
4261 64. When Exim failed to open the panic log after failing to open the main log,
4262 the original message it was trying to log was written to stderr and debug
4263 output, but if they were not available (the usual case in production), it
4264 was lost. Now it is written to syslog before the two lines that record the
4265 failures to open the logs.
4267 65. Users' Exim filters run in subprocesses under the user's uid. It is
4268 possible for a "deliver" command or an alias in a "personal" command to
4269 provoke an address rewrite. If logging of address rewriting is configured,
4270 this fails because the process is not running as root or exim. There may be
4271 a better way of dealing with this, but for the moment (because 4.30 needs
4272 to be released), I have disabled address rewrite logging when running a
4273 filter in a non-root, non-exim process.
4279 1. The buildconfig auxiliary program wasn't quoting the value set for
4280 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4281 not defined. This bug was masked in 4.22 by the effect that was fixed in
4284 2. Some messages that were rejected after a message id was allocated were
4285 shown as "incomplete" by exigrep. It no longer does this for messages that
4286 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4288 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4289 have allow_domain_literals set, the ID did not get logged in the <= line.
4290 Domain literals are now always recognized in Message-ID: header lines.
4292 4. The first argument for a ${extract expansion item is the key name or field
4293 number. Leading and trailing spaces in this item were not being ignored,
4294 causing some misleading effects.
4296 5. When deliver_drop_privilege was set, single queue runner processes started
4297 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4298 same command in the process it spins off) were not dropping privilege.
4300 6. When the daemon running as "exim" started a queue runner, it always
4301 re-executed Exim in the spun-off process. This is a waste of effort when
4302 deliver_drop_privilege is set. The new process now just calls the
4303 queue-runner function directly.
4309 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4312 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4313 ignored. Though the use of -oP was forcing the writing of a pid file, it
4314 was always written to the default place.
4316 3. If the message "no IP address found for host xxxx" is generated during
4317 incoming verification, it is now followed by identification of the incoming
4318 connection (so you can more easily find what provoked it).
4320 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4322 5. Added some features to "harden" Exim a bit more against certain attacks:
4324 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4325 be put in Local/Makefile. This is like the never_users runtime option,
4326 but it cannot be overridden. The default setting is "root".
4328 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4329 prefix string with which any file named in a -C command line option
4332 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4333 is retained for -C and -D only if the caller of Exim is root. Without
4334 it, the exim user may also use -C and -D and retain privilege.
4336 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4337 command line option is disabled.
4339 6. Macro names set by the -D option must start with an upper case letter, just
4340 like macro names defined in the configuration file.
4342 7. Added "dereference=" facility to LDAP.
4344 8. Two instances of the typo "uknown" in the source files are fixed.
4346 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4347 the Configure-Makefile script screwed up while processing it.
4349 10. Incorporated PCRE 4.4.
4351 11. The SMTP synchronization check was not operating right at the start of an
4352 SMTP session. For example, it could not catch a HELO sent before the client
4353 waited for the greeting. There is now a check for outstanding input at the
4354 point when the greeting is written. Because of the duplex, asynchronous
4355 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4356 way, but not yet received, when the check is performed.
4358 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4359 on TCP/IP sockets, because this apparently causes some broken clients to
4362 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4363 unchanged) from the Cygwin maintainer.
4365 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4366 was supported instead of testing for mbx. Effectively a typo.
4368 15. The spa authenticator server code was not checking that the input it
4369 received was valid base64.
4371 16. The debug output line for the "set" modifier in ACLs was not showing the
4372 name of the variable that was being set.
4374 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4376 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4377 lookup on its own. Something else was needed to trigger the lookup. For
4378 example, a match in host_lookup or the need for a host name in a host list.
4379 Now, if $sender_host_name is referenced and the host name has not yet been
4380 looked up, a lookup is performed. If the lookup fails, the variable remains
4381 empty, and $host_lookup_failed is set to "1".
4383 19. Added "eqi" as a case-independent comparison operator.
4385 20. The saslauthd authentication condition could segfault if neither service
4386 nor realm was specified.
4388 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4389 error message that was logged was misleading, and incoming SMTP
4390 connections were dropped. The message is now more accurate, and temporary
4391 errors are given to SMTP connections.
4393 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4394 (except RSET) with a 421 error, until QUIT is received. However, it was
4395 failing to send a response to QUIT.
4397 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4398 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4399 ACL is now run after the helo_try_verify_hosts code.
4401 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4402 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4403 case-independent, so other case variants are also recognized). Apparently
4404 some systems use these upper case variants.
4406 25. If more than two messages were waiting for the same host, and a transport
4407 filter was specified for the transport, Exim sent two messages over the
4408 same TCP/IP connection, and then failed with "socket operation on non-
4409 socket" when it tried to send the third.
4411 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4413 27. The extern definition of crypt16() in expand.c was not being excluded when
4414 the OS had its own crypt16() function.
4416 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4417 as a preferred synonym for return_size_limit, both as an option and as an
4420 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4422 30. Changed the default configuration ACL to relax the local part checking rule
4423 for addresses that are not in any local domains. For these addresses,
4424 slashes and pipe symbols are allowed within local parts, but the sequence
4425 /../ is explicitly forbidden.
4427 31. SPA server authentication was not clearing the challenge buffer before
4430 32. log_message in a "warn" ACL statement was writing to the reject log as
4431 well as to the main log, which contradicts the documentation and doesn't
4432 seem right (because no rejection is happening). So I have stopped it.
4434 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4435 However, I am unable to do any testing of this.
4437 34. Fixed an infelicity in the appendfile transport. When checking directories
4438 for a mailbox, to see if any needed to be created, it was accidentally
4439 using path names with one or more superfluous leading slashes; tracing
4440 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4442 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4443 contents are added to the log line that is written for every discarded
4444 recipient. (Previously a log_message setting was ignored.)
4446 36. The ${quote: operator now quotes the string if it is empty.
4448 37. The install script runs exim in order to find its version number. If for
4449 some reason other than non-existence or emptiness, which it checks, it
4450 could not run './exim', it was installing it with an empty version number,
4451 i.e. as "exim-". This error state is now caught, and the installation is
4454 38. An argument was missing from the function that creates an error message
4455 when Exim fails to connect to the socket for saslauthd authentication.
4456 This could cause Exim to crash, or give a corrupted message.
4458 39. Added isip, isip4, and isip6 to ${if conditions.
4460 40. The ACL variables $acl_xx are now saved with the message, and can be
4461 accessed later in routers, transports, and filters.
4463 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4464 strings in the file are not string-expanded.
4466 42. If a MAIL command specified a SIZE value that was too large to fit into an
4467 int variable, the check against message_size_limit failed. Such values are
4468 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4469 day this will have to be increased, but I don't think I want to be around
4470 when emails are that large.
4477 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4478 iconv() is not standard in FreeBSD.
4480 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4481 IPv6 enabled. The observed symptom was a segmentation fault on return from
4482 the function os_common_find_running_interfaces() in src/os.c.
4484 3. In the check_special_case() function in daemon.c I had used "errno" as an
4485 argument name, which causes warnings on some systems. This was basically a
4486 typo, since it was named "eno" in the comments!
4488 4. The code that waits for the clock to tick (at a resolution of some fraction
4489 of a second) so as to ensure message-id uniqueness was always waiting for
4490 at least one whole tick, when it could have waited for less. [This is
4491 almost certainly not relevant at current processor speeds, where it is
4492 unlikely to ever wait at all. But we try to future-proof.]
4494 5. The function that sleeps for a time interval that includes fractions of a
4495 second contained a race. It did not block SIGALRM between setting the
4496 timer, and suspending (a couple of lines later). If the interval was short
4497 and the sigsuspend() was delayed until after it had expired, the suspension
4498 never ended. On busy systems this could lead to processes getting stuck for
4501 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4502 process, before it forks any delivery processes. The open lookup caching
4503 mechanism meant that the open file or database connection was passed into
4504 the delivery process. The problem was that delivery processes always tidy
4505 up cached lookup data. This could cause a problem for the next delivery
4506 process started by the queue runner, because the external queue runner
4507 process does not know about the closure. So the next delivery process
4508 still has data in the lookup cache. In the case of a file lookup, there was
4509 no problem because closing a file descriptor in a subprocess doesn't affect
4510 the parent. However, if the lookup was caching a connection to a database,
4511 the connection was closed, and the second delivery process was likely to
4512 see errors such as "PGSQL: query failed: server closed the connection
4513 unexpectedly". The problem has been fixed by closing all cached lookups
4514 in a queue runner before running a delivery process.
4516 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4517 seem to have the "const" qualifier which it has on other OS. I've
4520 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4521 *of the same type* (client or server) with the same public name that an
4522 error should be diagnosed.
4524 9. When Exim looked up a host name for an IP address, but failed to find the
4525 original IP address when looking up the host name (a safety check), it
4526 output the message "<ip address> does not match any IP for NULL", which was
4527 confusing, to say the least. The bug was that the host name should have
4528 appeared instead of "NULL".
4530 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4531 user that is built into the binary, and the -C or -D options is used, root
4532 privilege is dropped before the configuration file is read. In addition,
4533 logging is switched to stderr instead of the normal log files. If the
4534 configuration then re-defines the Exim user, the unprivileged environment
4535 is probably not what is expected, so Exim logs a panic warning message (but
4538 However, if deliver_drop_privilege is set, the unprivileged state may well
4539 be exactly what is intended, so the warning has been cut out in that case,
4540 and Exim is allowed to try to write to its normal log files.
4546 1. smtp_return_error_details was not giving details for temporary sender
4547 or receiver verification errors.
4549 2. Diagnose a configuration error if two authenticators have the same public
4552 3. Exim used not to create the message log file for a message until the first
4553 delivery attempt. This could be confusing when incoming messages were held
4554 for policy or load reasons. The message log file is now created at the time
4555 the message is received, and an initial "Received" line is written to it.
4557 4. The automatically generated man page for command line options had a minor
4558 bug that caused no ill effects; however, a more serious problem was that
4559 the procedure for building the man page automatically didn't always
4560 operate. Consequently, release 4.20 contains an out-of-date version. This
4561 shouldn't happen again.
4563 5. When building Exim with embedded Perl support, the script that builds the
4564 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4565 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4567 6. The freeze_tell option was not being used for messages that were frozen on
4568 arrival, either by an ACL or by local_scan().
4570 7. Added the smtp_incomplete_transaction log selector.
4572 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4573 was accepting AUTH without a new EHLO.
4575 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4576 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4579 10. Logging of TCP/IP connections (when configured) now happens in the main
4580 daemon process instead of the child process, so that the TCP/IP connection
4581 count is more accurate (but it can never be perfect).
4583 11. The use of "drop" in a nested ACL was not being handled correctly in the
4584 outer ACL. Now, if condition failure induced by the nested "drop" causes
4585 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4586 or "require"), the connection is dropped.
4588 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4589 that yield "discard" can now be used with an "accept" or a "discard" verb,
4590 but an error is generated for any others (because I can't see a useful way
4591 to define what should happen).
4593 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4594 are now processed in the same way as lines in the Exim configuration file.
4595 In particular, continuation lines are supported.
4597 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4599 15. Added -ti meaning -t -i.
4601 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4602 domains, and warn by logging if others are found.
4604 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4605 returned from GIFCONF when Exim is trying to find the list of interfaces on
4606 a host. The code in os.c has been modified to copy each ifreq to an aligned
4607 structure in all cases.
4609 Also, in some cases, the returned ifreq's were being copied to a 'struct
4610 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4611 means the last couple of bytes of an IPv6 address could be chopped if the
4612 ifreq contained only a normal sockaddr (14 bytes storage).
4614 18. Named domain lists were not supported in the hosts_treat_as_local option.
4615 An entry such as +xxxx was not recognized, and was treated as a literal
4618 19. Ensure that header lines added by a DATA ACL are included in the reject log
4619 if the ACL subsequently rejects the message.
4621 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4622 MD5 (which is deprecated).
4624 21. When testing a filter file using -bf, Exim was writing a message when it
4625 took the sender from a "From " line in the message, but it was not doing so
4626 when it took $return_path from a Return-Path: header line. It now does.
4628 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4629 with a valid header line field name (a series of printing characters
4630 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4632 23. Changed "disc" in the source to "disk" to conform to the documentation and
4633 the book and for uniformity.
4635 24. Ignore Sendmail's -Ooption=value command line item.
4637 25. When execve() failed while trying to run a command in a pipe transport,
4638 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4639 could be confused with a return value of 69 from the command itself. This
4640 has been changed to 127, the value the shell returns if it is asked to run
4641 a non-existent command. The wording for the related log line suggests a
4642 non-existent command as the problem.
4644 26. If received_header_text expands to an empty string, do not add a Received:
4645 header line to the message. (Well, it adds a token one on the spool, but
4646 marks it "old" so that it doesn't get used or transmitted.)
4648 27. Installed eximstats 1.28 (addition of -nt option).
4650 28. There was no check for failure on the call to getsockname() in the daemon
4651 code. This can fail if there is a shortage of resources on the system, with
4652 ENOMEM, for example. A temporary error is now given on failure.
4654 29. Contrary to the C standard, it seems that in some environments, the
4655 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4656 program. Exim now does this explicitly; it affects the formatting of
4657 timestamps using strftime().
4659 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4660 complaints. I've now initialized all the variables, to avoid this.
4662 32. Header lines added by a system filter were not being "seen" during
4663 transport-time rewrites.
4665 33. The info_callback() function passed to OpenSSL is set up with type void
4666 (*)(SSL *, int, int), as described somewhere. However, when calling the
4667 function (actually a macro) that sets it up, the type void(*)() is
4668 expected. I've put in a cast to prevent warnings from picky compilers.
4670 34. If a DNS black list lookup found a CNAME record, but there were no A
4671 records associated with the domain it pointed at, Exim crashed.
4673 35. If a DNS black list lookup returned more than one A record, Exim ignored
4674 all but the first. It now scans all returned addresses if a particular IP
4675 value is being sought. In this situation, the contents of the
4676 $dnslist_value variable are a list of all the addresses, separated by a
4679 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4680 to accept a host name and all its aliases if the forward lookup for any of
4681 them yielded the IP address of the incoming connection. Now it accepts only
4682 those names whose forward lookup yields the correct IP address. Any other
4683 names are discarded. This closes a loophole whereby a rogue DNS
4684 administrator could create reverse DNS records to break through a
4685 wildcarded host restriction in an ACL.
4687 37. If a user filter or a system filter that ran in a subprocess used any of
4688 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4689 the wrong values were passed to the pipe command ($thisaddress had the
4690 value of $0, $0 had the value of $1, etc). This bug was introduced by
4691 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4693 38. Improved the line breaking for long SMTP error messages from ACLs.
4694 Previously, if there was no break point between 40 and 75 characters, Exim
4695 left the rest of the message alone. Two changes have been made: (a) I've
4696 reduced the minimum length to 35 characters; (b) if it can't find a break
4697 point between 35 and 75 characters, it looks ahead and uses the first one
4698 that it finds. This may give the occasional overlong line, but at least the
4699 remaining text gets split now.
4701 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4702 file descriptors might be low, and that setting 1000 would always raise it.
4703 It turns out that in some environments, the limit is already over 1000 and
4704 that lowering it causes trouble. So now Exim takes care not to decrease it.
4706 40. When delivering a message, the value of $return_path is set to $sender_
4707 address at the start of routing (routers may change the value). By an
4708 oversight, this default was not being set up when an address was tested by
4709 -bt or -bv, which affected the outcome if any router or filter referred to
4712 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4713 log only when "message" or "log_message" are set. However, if one of the
4714 conditions was an address verification, or a call to a nested ACL, the
4715 messages generated by the underlying test were being passed through. This
4716 no longer happens. The underlying message is available in $acl_verify_
4717 message for both "message" and "log_message" expansions, so it can be
4718 passed through if needed.
4720 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4721 new expansion $bh_ to give the encoded byte string without charset
4722 translation. Translation happens only if iconv() is available; HAVE_ICONV
4723 indicates this at build time. HEADERS_CHARSET gives the charset to
4724 translate to; headers_charset can change it in the configuration, and
4725 "headers charset" can change it in an individual filter file.
4727 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4728 that creates RFC 2047 encoded "words" labels them as that charset instead
4729 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4730 expansion operator; (ii) when Exim creates a From: line for a local
4731 message; (iii) when a header line is rewritten to include a "phrase" part.
4733 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4734 buggy, causing it to skip the first lines of messages whose message ID
4735 ended in 'D'. This would not have bitten before Exim release 4.14, because
4736 message IDs were unlikely to end in 'D' before then. The effect was to have
4737 incorrect size information for certain domains.
4739 45. #include "config.h" was missing at the start of the crypt16.c module. This
4740 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4743 46. If there was a timeout during a "random" callout check, Exim treated it as
4744 a failure of the random address, and carried on sending RSET and the real
4745 address. If the delay was just some slowness somewhere, the response to the
4746 original RCPT would be taken as a response to RSET and so on, causing
4747 mayhem of various kinds.
4749 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4750 when I implemented it. It didn't allow for the fact that some option values
4751 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4752 the right test for positive values.
4754 48. Domain names in DNS records are case-independent. Exim always looks them up
4755 in lower case. Some resolvers return domain names in exactly the case they
4756 appear in the zone file, that is, they may contain uppercase letters. Not
4757 all resolvers do this - some return always lower case. Exim was treating a
4758 change of case by a resolver as a change of domain, similar to a widening
4759 of a domain abbreviation. This triggered its re-routing code and so it was
4760 trying to route what was effectively the same domain again. This normally
4761 caused routing to fail (because the router wouldn't handle the domain
4762 twice). Now Exim checks for this case specially, and just changes the
4763 casing of the domain that it ultimately uses when it transmits the message
4766 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4769 50. If a filter generated a file delivery with a non-absolute name (possible if
4770 no home directory exists for the router), the forbid_file option was not
4773 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4774 the existing equality matching.
4776 52. Exim was using ints instead of ino_t variables in some places where it was
4777 dealing with inode numbers.
4779 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4780 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4781 called TMPDIR, and if it finds it is different, it changes its value.
4783 54. The smtp_printf() function is now made available to local_scan() so
4784 additional output lines can be written before returning. There is also an
4785 smtp_fflush() function to enable the detection of a dropped connection.
4786 The variables smtp_input and smtp_batched_input are exported to
4789 55. Changed the default runtime configuration: the message "Unknown user"
4790 has been removed from the ACL, and instead placed on the localuser router,
4791 using the cannot_route_message feature. This means that any verification
4792 failures that generate their own messages won't get overridden. Similarly,
4793 the "Unrouteable address" message that was in the ACL for unverifiable
4794 relay addresses has also been removed.
4796 56. Added hosts_avoid_esmtp to the smtp transport.
4798 57. The exicyclog script was not checking for the esoteric option
4799 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4800 will work only if exicyclog is run under the appropriate euid.
4802 58. Following a discussion on the list, the rules by which Exim recognises line
4803 endings on incoming messages have been changed. The -dropcr and drop_cr
4804 options are now no-ops, retained only for backwards compatibility. The
4805 following line terminators are recognized: LF CRLF CR. However, special
4806 processing applies to CR:
4808 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4809 nor a local message in the state where . is a terminator.
4811 (ii) If a bare CR is encountered in a header line, an extra space is added
4812 after the line terminator so as not to end the header. The reasoning
4813 behind this is that bare CRs in header lines are most likely either
4814 to be mistakes, or people trying to play silly games.
4816 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4817 was being incorrectly given as 18 bytes larger than it should have been.
4818 This is a VOB (very old bug).
4820 60. This may never have affected anything current, but just in case it has:
4821 When the local host is found other than at the start of a list of hosts,
4822 the local host, those with the same MX, and any that follow, are discarded.
4823 When the list in question was part of a longer list of hosts, the following
4824 hosts (not currently being processed) were also being discarded. This no
4825 longer happens. I'm not sure if this situation could ever has previously
4828 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4831 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4832 header. This is something that is recommended in a new Internet Draft, and
4833 is something that is documented as being done by Sendmail. There are two
4834 possible values. For messages generated by the autoreply transport, Exim
4837 Auto-Submitted: auto-replied
4839 whereas for all other generated messages (e.g. bounces) it adds
4841 Auto-Submitted: auto-generated
4843 63. The "personal" condition in filters now includes a test for the
4844 Auto-Submitted: header. If it contains the string "auto-" the message it
4845 not considered personal.
4847 64. Added rcpt_include_affixes as a generic transport option.
4849 65. Added queue_only_override (default true).
4851 66. Added the syslog_duplication option.
4853 67. If what should have been the first header line of a message consisted of
4854 a space followed by a colon, Exim was mis-interpreting it as a header line.
4855 It isn't of course - it is syntactically invalid and should therefore be
4856 treated as the start of the message body. The misbehaviour could have
4857 caused a number of strange effects, including loss of data in subsequent
4858 header lines, and spool format errors.
4860 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4861 client host had authenticated. This control can now be exercised by an ACL
4862 for more flexibility.
4864 69. By default, callouts do not happen when testing with -bh. There is now a
4865 variant, -bhc, which does actually run the callout code, including
4866 consulting and updating the callout cache.
4868 70. Added support for saslauthd authentication, courtesy of Alexander
4871 71. If statvfs() failed on the spool or log directories while checking their
4872 size for availability, Exim confusingly gave the error "space shortage".
4873 Furthermore, in debugging mode it crashed with a floating point exception.
4874 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4875 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4876 serious problem, Exim now writes to the main and panic logs when this
4877 happens, with details of the failure. It then refuses to accept the
4878 incoming message, giving the message "spool directory problem" or "log
4879 directory problem" with a 421 code for SMTP messages.
4881 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4882 0, 1, and 2 exist, because some OS complain for execs without them (see
4883 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4884 descriptors. However, the code omitted to check that the open succeeded,
4885 causing mysterious errors if for some reason the permissions on /dev/null
4886 got screwed. Now Exim writes a message to the main and panic logs, and
4887 bombs out if it can't open /dev/null.
4889 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4890 interact so that it is all more flexible. It is supposed to remain
4891 backwards compatible. Also added extra_local_interfaces.
4893 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4894 to bomb out with an assertion failure - to the client this appears as a
4895 connection drop. This problem occurs in the part of the code that was taken
4896 from the Samba project. Fortunately, the assertion is in a very simple
4897 function, so I have fixed this by reproducing the function inline in the
4898 one place where it is called, and arranging for authentication to fail
4899 instead of killing the process with assert().
4901 75. The SPA client code was not working when the server requested OEM rather
4902 than Unicode encoding.
4904 76. Added code to make require_files with a specific uid setting more usable in
4905 the case where statting the file as root fails - usually a non-root-mounted
4906 NFS file system. When this happens and the failure is EACCES, Exim now
4907 forks a subprocess and does the per-uid checking as the relevant uid.
4909 77. Added process_log_path.
4911 78. If log_file_path was not explicitly set, a setting of check_log_space or
4912 check_log_inodes was ignored.
4914 79. If a space check for the spool or log partitions fails, the incident is now
4915 logged. Of course, in the latter case the data may get lost...
4917 80. Added the %p formatting code to string_format() so that it can be used to
4918 print addresses in debug_print(). Adjusted all the address printing in the
4919 debugging in store.c to use %p rather than %d.
4921 81. There was a concern that a line of code in smtp_in.c could overflow a
4922 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4923 initially expressed, the concern was not well-founded, because trailing
4924 spaces are removed early. However, if the trailing spaces were followed by
4925 a NULL, they did not get removed, so the overflow was possible. Two fixes
4928 (a) I re-wrote the offending code in a cleaner fashion.
4929 (b) If an incoming SMTP command contains a NULL character, it is rejected
4932 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4933 runs initgroups(), so that if the Exim user is in any additional groups,
4934 they will be used during message reception.
4940 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob