1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.559 2008/12/12 14:51:47 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
80 NM/09 Bugzilla 787: Potential buffer overflow in string_format
81 Patch provided by Eugene Bujak
83 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to accept()
84 Patch provided by Maxim Dounin
86 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
87 Patch provided by Phil Pennock
93 TK/01 Add preliminary DKIM support. Currently requires a forked version of
94 ALT-N's libdkim that I have put here:
95 http://duncanthrax.net/exim-experimental/
97 Note to Michael Haardt: I had to rename some vars in sieve.c. They
98 were called 'true' and it seems that C99 defines that as a reserved
99 keyword to be used with 'bool' variable types. That means you could
100 not include C99-style headers which use bools without triggering
101 build errors in sieve.c.
103 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
104 as mailq or other aliases. Changed the --help handling significantly
105 to do whats expected. exim_usage() emits usage/help information.
107 SC/01 Added the -bylocaldomain option to eximstats.
109 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
111 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
113 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
119 PH/01 Another patch from the Sieve maintainer.
121 PH/02 When an IPv6 address is converted to a string for single-key lookup
122 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
123 dots are used instead of colons so that keys in lsearch files need not
124 contain colons. This was done some time before quoting was made available
125 in lsearch files. However, iplsearch files do require colons in IPv6 keys
126 (notated using the quote facility) so as to distinguish them from IPv4
127 keys. This meant that lookups for IP addresses in host lists did not work
128 for iplsearch lookups.
130 This has been fixed by arranging for IPv6 addresses to be expressed with
131 colons if the lookup type is iplsearch. This is not incompatible, because
132 previously such lookups could never work.
134 The situation is now rather anomolous, since one *can* have colons in
135 ordinary lsearch keys. However, making the change in all cases is
136 incompatible and would probably break a number of configurations.
138 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
141 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
142 conversion specification without a maximum field width, thereby enabling
143 a rogue spamd server to cause a buffer overflow. While nobody in their
144 right mind would setup Exim to query an untrusted spamd server, an
145 attacker that gains access to a server running spamd could potentially
146 exploit this vulnerability to run arbitrary code as the Exim user.
148 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
149 $primary_hostname instead of what libspf2 thinks the hosts name is.
151 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
152 a directory entry by the name of the lookup key. Previously, if a
153 symlink pointed to a non-existing file or a file in a directory that
154 Exim lacked permissions to read, a lookup for a key matching that
155 symlink would fail. Now it is enough that a matching directory entry
156 exists, symlink or not. (Bugzilla 503.)
158 PH/03 The body_linecount and body_zerocount variables are now exported in the
161 PH/04 Added the $dnslist_matched variable.
163 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
164 This means they are set thereafter only if the connection becomes
167 PH/06 Added the client_condition to authenticators so that some can be skipped
168 by clients under certain conditions.
170 PH/07 The error message for a badly-placed control=no_multiline_responses left
171 "_responses" off the end of the name.
173 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
175 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
176 (without spaces) instead of just copying the configuration text.
178 PH/10 Added the /noupdate option to the ratelimit ACL condition.
180 PH/11 Added $max_received_linelength.
182 PH/12 Added +ignore_defer and +include_defer to host lists.
184 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
185 way in which PCRE > 7.0 is built.
187 PH/14 Implemented queue_only_load_latch.
189 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
190 MAIL command. The effect was to mangle the value on 64-bit systems.
192 PH/16 Another patch from the Sieve maintainer.
194 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
196 PH/18 If a system quota error occurred while trying to create the file for
197 a maildir delivery, the message "Mailbox is full" was not appended to the
198 bounce if the delivery eventually timed out. Change 4.67/27 below applied
199 only to a quota excession during the actual writing of the file.
201 PH/19 It seems that peer DN values may contain newlines (and other non-printing
202 characters?) which causes problems in log lines. The DN values are now
203 passed through string_printing() before being added to log lines.
205 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
206 and InterBase are left for another time.)
208 PH/21 Added message_body_newlines option.
210 PH/22 Guard against possible overflow in moan_check_errorcopy().
212 PH/23 POSIX allows open() to be a macro; guard against that.
214 PH/24 If the recipient of an error message contained an @ in the local part
215 (suitably quoted, of course), incorrect values were put in $domain and
216 $local_part during the evaluation of errors_copy.
222 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
223 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
226 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
227 issue a MAIL command.
229 PH/02 In an ACL statement such as
231 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
233 if a client was not listed at all, or was listed with a value other than
234 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
235 the condition was not true (as it should be), so access was not denied.
236 The bug was that the ! inversion was incorrectly passed on to the second
237 item. This has been fixed.
239 PH/03 Added additional dnslists conditions == and =& which are different from
240 = and & when the dns lookup returns more than one IP address.
242 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
243 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
245 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
246 FSYNC, which compiles an option called disable_fsync that allows for
247 bypassing fsync(). The documentation is heavily laced with warnings.
249 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
251 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
252 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
253 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
254 including adding "make clean"; (3) Added -fPIC when compiling the test
255 dynamically loaded module, to get rid of a warning.
257 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
258 message fails, move_frozen_messages = true and ignore_bounce_errors_after
259 = 0s. The bug is otherwise harmless.
261 PH/07 There was a bug in the dovecot authenticator such that the value of
262 $auth1 could be overwritten, and so not correctly preserved, after a
263 successful authentication. This usually meant that the value preserved by
264 the server_setid option was incorrect.
266 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
268 PH/09 Installed PCRE release 7.0.
270 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
271 run for batched SMTP input. It is now run at the start of every message
272 in the batch. While fixing this I discovered that the process information
273 (output by running exiwhat) was not always getting set for -bs and -bS
274 input. This is fixed, and it now also says "batched" for BSMTP.
276 PH/11 Added control=no_pipelining.
278 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
279 patch, slightly modified), and move the expansion of helo_data till after
280 the connection is made in the smtp transport (so it can use these
283 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
285 PH/14 Added log_selector = +pid.
287 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
289 PH/16 Add ${if forany and ${if forall.
291 PH/17 Added dsn_from option to vary the From: line in DSNs.
293 PH/18 Flush SMTP output before performing a callout, unless control =
294 no_callout_flush is set.
296 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
297 was true (the default) a successful delivery failed to delete the retry
298 item, thus causing premature timeout of the address. The bug is now
301 PH/20 Added hosts_avoid_pipelining to the smtp transport.
303 PH/21 Long custom messages for fakedefer and fakereject are now split up
304 into multiline reponses in the same way that messages for "deny" and
305 other ACL rejections are.
307 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
308 with slight modification.
310 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
311 draft, changing the syntax and factoring some duplicate code".
313 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
314 for deliveries of the second and subsequent messages over the same SMTP
317 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
318 ${reduce, with only minor "tidies".
320 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
322 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
323 expansion side effects.
325 PH/27 When a message times out after an over-quota error from an Exim-imposed
326 quota, the bounce message says "mailbox is full". This message was not
327 being given when it was a system quota that was exceeded. It now should
330 MH/03 Made $recipients available in local_scan(). local_scan() already has
331 better access to the recipient list through recipients_list[], but
332 $recipients can be useful in postmaster-provided expansion strings.
334 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
335 in the case of a MAIL command with additional options following the
336 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
337 were accidentally chopped off.
339 PH/29 SMTP synchronization checks are implemented when a command is read -
340 there is a check that no more input is waiting when there shouldn't be
341 any. However, for some commands, a delay in an ACL can mean that it is
342 some time before the response is written. In this time, more input might
343 arrive, invalidly. So now there are extra checks after an ACL has run for
344 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
345 pipelining has not been advertised.
347 PH/30 MH's patch to allow iscntrl() characters to be list separators.
349 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
350 returned in the SMTP response when smtp_return_error_details was false.
353 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
354 instead of the C I/O that was originally supplied, because problems were
357 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
358 Exim which did not show up earlier: it was assuming that a call to
359 SSL_CTX_set_info_callback() might give an error value. In fact, there is
360 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
361 was a macro that became an assignment, so it seemed to work. This has
362 changed to a proper function call with a void return, hence the compile
363 error. Exim's code has been fixed.
365 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
368 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
370 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
372 PH/37 If a message is not accepted after it has had an id assigned (e.g.
373 because it turns out to be too big or there is a timeout) there is no
374 "Completed" line in the log. When some messages of this type were
375 selected by exigrep, they were listed as "not completed". Others were
376 picked up by some special patterns. I have improved the selection
377 criteria to be more general.
379 PH/38 The host_find_failed option in the manualroute router can now be set
380 to "ignore", to completely ignore a host whose IP address cannot be
381 found. If all hosts are ignored, the behaviour is controlled by the new
382 host_all_ignored option.
384 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
385 homing or because of multiple MX records with /mx) generated more than
386 one IP address, and the following item turned out to be the local host,
387 all the secondary addresses of the first item were incorrectly removed
388 from the list, along with the local host and any following hosts (which
389 is what is supposed to happen).
391 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
392 whoever called Exim into the -H file. In the case of the daemon it was
393 behaving confusingly. When first started, it used values for whoever
394 started the daemon, but after a SIGHUP it used the Exim user (because it
395 calls itself on a restart). I have changed the code so that it now always
398 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
399 message are rejected with the same error (e.g. no authentication or bad
400 sender address), and a DATA command is nevertheless sent (as can happen
401 with PIPELINING or a stupid MUA), the error message that was given to the
402 RCPT commands is included in the rejection of the DATA command. This is
403 intended to be helpful for MUAs that show only the final error to their
406 PH/42 Another patch from the Sieve maintainer.
408 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
409 Eximstats - Fixed some broken HTML links and added missing column headers
411 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
412 columns for Rejects, Temp Rejects, Ham, and Spam rows.
414 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
416 PH/43 Yet another patch from the Sieve maintainer.
418 PH/44 I found a way to check for a TCP/IP connection going away before sending
419 the response to the final '.' that terminates a message, but only in the
420 case where the client has not sent further data following the '.'
421 (unfortunately, this is allowed). However, in many cases there won't be
422 any further data because there won't be any more messages to send. A call
423 to select() can be used: if it shows that the input is "ready", there is
424 either input waiting, or the socket has been closed. An attempt to read
425 the next input character can distinguish the two cases. Previously, Exim
426 would have sent an OK response which the client would never have see.
427 This could lead to message repetition. This fix should cure that, at
428 least in a lot of common cases.
430 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
431 advertised in response to EHLO.
437 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
438 fixed by 4.65/MH/01 (is this a record?) are fixed:
440 (i) An empty string was always treated as zero by the numeric comparison
441 operators. This behaviour has been restored.
443 (ii) It is documented that the numeric comparison operators always treat
444 their arguments as decimal numbers. This was broken in that numbers
445 starting with 0 were being interpreted as octal.
447 While fixing these problems I realized that there was another issue that
448 hadn't been noticed. Values of message_size_limit (both the global option
449 and the transport option) were treated as octal if they started with 0.
450 The documentation was vague. These values are now always treated as
451 decimal, and I will make that clear in the documentation.
457 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
458 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
461 MH/01 Don't check that the operands of numeric comparison operators are
462 integers when their expansion is in "skipping" mode (fixes bug
463 introduced by 4.64-PH/07).
465 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
466 child addresses, Exim now panics and dies. Previously, because the count
467 is held in a short int, deliveries were likely to be lost. As such a
468 large number of recipients for a single message is ridiculous
469 (performance will be very, very poor), I have chosen to impose a limit
470 rather than extend the field.
476 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
477 leftover -K file (the existence of which was triggered by #402).
478 While we were at it, introduced process PID as part of the -K
479 filename. This should rule out race conditions when creating
482 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
483 processing considerably. Previous code took too long for large mails,
484 triggering a timeout which in turn triggers #401.
486 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
487 in the DK code in transports.c. sendfile() is not really portable,
488 hence the _LINUX specificness.
490 TF/01 In the add_headers option to the mail command in an Exim filter,
491 there was a bug that Exim would claim a syntax error in any
492 header after the first one which had an odd number of characters
495 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
496 callout verification, Exim cached a "reject" for the entire domain. This
497 is correct for most verifications, but it is not correct for a recipient
498 verification with use_sender or use_postmaster set, because in that case
499 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
500 case of MAIL FROM:<> rejection from other early rejections (e.g.
501 rejection of HELO). When verifying a recipient using a non-null MAIL
502 address, the cache is ignored if it shows MAIL FROM:<> rejection.
503 Whatever the result of the callout, the value of the domain cache is
504 left unchanged (for any other kind of callout, getting as far as trying
505 RCPT means that the domain itself is ok).
507 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
510 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
511 manifest itself as EPIPE rather than ECONNECT. When tidying away a
512 session, the daemon ignores ECONNECT errors and logs others; it now
513 ignores EPIPE as well.
515 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
516 (quoted-printable decoding).
518 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
519 later the small subsequent patch to fix an introduced bug.
521 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
523 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
525 PH/08 An error is now given if message_size_limit is specified negative.
527 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
528 to be given (somewhat) arbitrary names.
530 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
533 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
534 miscellaneous code fixes
536 PH/10 Added the log_reject_target ACL modifier to specify where to log
539 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
540 hostname. This is wrong, because it relates to the incoming message (and
541 probably the interface on which it is arriving) and not to the outgoing
542 callout (which could be using a different interface). This has been
543 changed to use the value of the helo_data option from the smtp transport
544 instead - this is what is used when a message is actually being sent. If
545 there is no remote transport (possible with a router that sets up host
546 addresses), $smtp_active_hostname is used.
548 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
549 tweaks were necessary in order to get it to work (see also 21 below):
550 (a) The code assumed that strncpy() returns a negative number on buffer
551 overflow, which isn't the case. Replaced with Exim's string_format()
553 (b) There were several signed/unsigned issues. I just did the minimum
554 hacking in of casts. There is scope for a larger refactoring.
555 (c) The code used strcasecmp() which is not a standard C function.
556 Replaced with Exim's strcmpic() function.
557 (d) The code set only $1; it now sets $auth1 as well.
558 (e) A simple test gave the error "authentication client didn't specify
559 service in request". It would seem that Dovecot has changed its
560 interface. Fortunately there's a specification; I followed it and
561 changed what the client sends and it appears to be working now.
563 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
566 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
567 address A is aliased to B and C, where B exists and C does not. Without
568 -v the output is "A verified" because verification stops after a
569 successful redirection if more than one address is generated. However,
570 with -v the child addresses are also verified. Exim was outputting "A
571 failed to verify" and then showing the successful verification for C,
572 with its parentage. It now outputs "B failed to verify", showing B's
573 parentage before showing the successful verification of C.
575 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
576 look up a TXT record in a specific list after matching in a combined
579 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
580 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
581 they consult the DNS. I had assumed they would set it the way they
582 wanted; and indeed my experiments on Linux seem to show that in some
583 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
584 To be on the safe side, however, I have now made the interface to
585 host_find_byname() similar to host_find_bydns(), with an argument
586 containing the DNS resolver options. The host_find_byname() function now
587 sets these options at its start, just as host_find_bydns() does. The smtp
588 transport options dns_qualify_single and dns_search_parents are passed to
589 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
590 of host_find_byname() use the default settings of RES_DEFNAMES
591 (qualify_single) but not RES_DNSRCH (search_parents).
593 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
594 spool_read_header() do less string testing, by means of a preliminary
595 switch on the second character of optional "-foo" lines. (This is
596 overdue, caused by the large number of possibilities that now exist.
597 Originally there were few.) While I was there, I also converted the
598 str(n)cmp tests so they don't re-test the leading "-" and the first
599 character, in the hope this might squeeze out yet more improvement.
601 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
602 flag allowing group syntax was set by the header_syntax check but not
603 turned off, possible causing trouble later; (2) The flag was not being
604 set at all for the header_verify test, causing "group"-style headers to
605 be rejected. I have now set it in this case, and also caused header_
606 verify to ignore an empty address taken from a group. While doing this, I
607 came across some other cases where the code for allowing group syntax
608 while scanning a header line wasn't quite right (mostly, not resetting
609 the flag correctly in the right place). These bugs could have caused
610 trouble for malformed header lines. I hope it is now all correct.
612 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
613 with the "reply" argument non-NULL. The code, however (which originally
614 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
615 but it didn't always do it. This confused somebody who was copying the
616 code for some other use. I have removed all the tests.
618 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
619 feature that was used to support insecure browsers during the U.S. crypto
620 embargo. It requires special client support, and Exim is probably the
621 only MTA that supported it -- and would never use it because real RSA is
622 always available. This code has been removed, because it had the bad
623 effect of slowing Exim down by computing (never used) parameters for the
624 RSA_EXPORT functionality.
626 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
627 authenticator to fail if there's a tab character in the incoming data
628 (there should never be unless someone is messing about, as it's supposed
629 to be base64-encoded). Also added, on Timo's advice, the "secured" option
630 if the connection is using TLS or if the remote IP is the same as the
631 local IP, and the "valid-client-cert option" if a client certificate has
634 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
635 authenticators. This can be used for authorization after authentication
636 succeeds. (In the case of plaintext, it servers for both authentication
639 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
640 if any retry times were supplied.
642 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
643 connection, where there is no client IP address to check. In this
644 situation, the verify now always succeeds.
646 PH/25 Applied John Jetmore's -Mset patch.
648 PH/26 Added -bem to be like -Mset, but loading a message from a file.
650 PH/27 In a string expansion for a processed (not raw) header when multiple
651 headers of the same name were present, leading whitespace was being
652 removed from all of them, but trailing whitespace was being removed only
653 from the last one. Now trailing whitespace is removed from each header
654 before concatenation. Completely empty headers in a concatenation (as
657 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
658 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
660 PH/29 [Removed. This was a change that I later backed out, and forgot to
661 correct the ChangeLog entry (that I had efficiently created) before
662 committing the later change.]
664 PH/30 Exim was sometimes attempting to deliver messages that had suffered
665 address errors (4xx response to RCPT) over the same connection as other
666 messages routed to the same hosts. Such deliveries are always "forced",
667 so retry times are not inspected. This resulted in far too many retries
668 for the affected addresses. The effect occurred only when there were more
669 hosts than the hosts_max_try setting in the smtp transport when it had
670 the 4xx errors. Those hosts that it had tried were not added to the list
671 of hosts for which the message was waiting, so if all were tried, there
672 was no problem. Two fixes have been applied:
674 (i) If there are any address or message errors in an SMTP delivery, none
675 of the hosts (tried or untried) are now added to the list of hosts
676 for which the message is waiting, so the message should not be a
677 candidate for sending over the same connection that was used for a
678 successful delivery of some other message. This seems entirely
679 reasonable: after all the message is NOT "waiting for some host".
680 This is so "obvious" that I'm not sure why it wasn't done
681 previously. Hope I haven't missed anything, but it can't do any
682 harm, as the worst effect is to miss an optimization.
684 (ii) If, despite (i), such a delivery is accidentally attempted, the
685 routing retry time is respected, so at least it doesn't keep
686 hammering the server.
688 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
689 in ${readsocket because some servers need this prod.
691 PH/32 Added some extra debug output when updating a wait-xxx database.
693 PH/33 The hint "could be header name not terminated by colon", which has been
694 given for certain expansion errors for a long time, was not being given
695 for the ${if def:h_colon_omitted{... case.
697 PH/34 The spec says: "With one important exception, whenever a domain list is
698 being scanned, $domain contains the subject domain." There was at least
699 one case where this was not true.
701 PH/35 The error "getsockname() failed: connection reset by peer" was being
702 written to the panic log as well as the main log, but it isn't really
703 panic-worthy as it just means the connection died rather early on. I have
704 removed the panic log writing for the ECONNRESET error when getsockname()
707 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
708 runs only) independently of the message's sender address. This meant
709 that, if the 4xx error was in fact related to the sender, a different
710 message to the same recipient with a different sender could confuse
711 things. In particualar, this can happen when sending to a greylisting
712 server, but other circumstances could also provoke similar problems.
713 I have changed the default so that the retry time for these errors is now
714 based a combination of the sender and recipient addresses. This change
715 can be overridden by setting address_retry_include_sender=false in the
718 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
719 remote server are returned as part of bounce messages. This was not
720 happening for LMTP over a pipe (the lmtp transport), but now it is the
721 same for both kinds of LMTP.
723 PH/38 Despite being documented as not happening, Exim was rewriting addresses
724 in header lines that were in fact CNAMEs. This is no longer the case.
726 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
727 and queue runs started by the daemon processed all messages. This has
728 been fixed so that -R and -S can now usefully be given with -q<time>.
730 PH/40 Import PCRE release 6.7 (fixes some bugs).
732 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
734 PH/42 Give an error if -q is specified more than once.
736 PH/43 Renamed the variables $interface_address and $interface_port as
737 $received_ip_address and $received_port, to make it clear that these
738 values apply to message reception, and not to the outgoing interface when
739 a message is delivered. (The old names remain recognized, of course.)
741 PH/44 There was no timeout on the connect() call when using a Unix domain
742 socket in the ${readsocket expansion. There now is.
744 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
745 be meaningful with "accept".
747 SC/01 Eximstats V1.43
748 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
750 SC/02 Eximstats V1.44
751 Use a glob alias rather than an array ref in the generated
752 parser. This improves both readability and performance.
754 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
755 Collect SpamAssassin and rejection statistics.
756 Don't display local sender or destination tables unless
757 there is data to show.
758 Added average volumes into the top table text output.
760 SC/04 Eximstats V1.46
761 Collect data on the number of addresses (recipients)
762 as well as the number of messages.
764 SC/05 Eximstats V1.47
765 Added 'Message too big' to the list of mail rejection
766 reasons (thanks to Marco Gaiarin).
768 SC/06 Eximstats V1.48
769 Mainlog lines which have GMT offsets and are too short to
770 have a flag are now skipped.
772 SC/07 Eximstats V1.49 (Alain Williams)
773 Added the -emptyok flag.
775 SC/08 Eximstats V1.50
776 Fixes for obtaining the IP address from reject messages.
778 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
779 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
780 whitesspace changes from 4.64-PH/27
782 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
785 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
786 are found, allow negative numbers in numeric criteria)
788 JJ/06 exipick.20061117.2, added new $message_body_missing variable
790 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
791 to match changes made in 4.64-PH/43
793 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
795 PH/47 Put in an explicit test for a DNS lookup of an address record where the
796 "domain" is actually an IP address, and force a failure. This locks out
797 those revolvers/nameservers that support "A-for-A" lookups, in
798 contravention of the specifications.
800 PH/48 When a host name was looked up from an IP address, and the subsequent
801 forward lookup of the name timed out, the host name was left in
802 $sender_host_name, contrary to the specification.
804 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
805 restricted to single-key lookups, Exim was not diagnosing an error if
806 * or *@ was used with a query-style lookup.
808 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
810 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
811 long ago, but noone interested enough thought of it. Let's just say that
812 the "1.1" means that there are some new functions that weren't there at
813 some point in the past.
815 PH/51 Error processing for expansion failure of helo_data from an smtp
816 transport during callout processing was broken.
818 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
819 tested/used via the -bh/-bhc/-bs options.
821 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
822 bug, fixed in subsequent PCRE releases).
824 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
825 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
827 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
833 SC/01 Use a glob alias rather than an array ref in eximstats generated
834 parser. This improves both readability and performance.
836 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
837 Don't display local sender or destination tables in eximstats unless
838 there is data to show.
839 Added average volumes into the eximstats top table text output.
841 SC/03 Collect data on the number of addresses (recipients) as well
842 as the number of messages in eximstats.
844 TF/01 Correct an error in the documentation for the redirect router. Exim
845 does (usually) call initgroups() when daemonizing.
847 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
848 with consistent privilege compared to when running as a daemon.
850 TF/03 Note in the spec that $authenticated_id is not set for local
851 submissions from trusted users.
853 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
854 Thanks to Dean Brooks <dean@iglou.com> for the patch.
856 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
857 by adding some example configuration directives to the default
858 configuration file. A little bit of work is required to uncomment the
859 directives and define how usernames and passwords are checked, but
860 there is now a framework to start from.
862 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
863 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
864 without this. I don't know how relevant this is to other LDAP libraries.
866 PH/02 Add the verb name to the "unknown ACL verb" error.
868 PH/03 Magnus Holmgren's patch for filter_prepend_home.
870 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
872 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
873 directory not expanded when it should be if an expanded home directory
874 was set for the address (which is overridden by the transport).
876 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
879 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
880 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
881 because it is too late at that time, and has no effect.
883 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
884 security issue with \' (bugzilla #107). I could not use the
885 PQescapeStringConn() function, because it needs a PGconn value as one of
888 PH/08 When testing addresses using -bt, indicate those final addresses that
889 are duplicates that would not cause an additional delivery. At least one
890 person was confused, thinking that -bt output corresponded to deliveries.
891 (Suppressing duplicates isn't a good idea as you lose the information
892 about possibly different redirections that led to the duplicates.)
894 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
895 systems where poll() doesn't work, in particular OS X.
897 PH/10 Added more information to debugging output for retry time not reached.
899 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
900 operations in malware.c.
902 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
905 PH/13 If write_rejectlog was set false when logging was sent to syslog with
906 syslog_duplication set false, log lines that would normally be written
907 both the the main log and to the reject log were not written to syslog at
910 PH/14 In the default configuration, change the use of "message" in ACL warn
911 statements to "add_header".
913 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
914 not followed by a command (e.g. "seen endif").
916 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
917 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
920 PH/17 Added extra conditions to the default value of delay_warning_condition
924 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
925 { match{$h_precedence:}{(?i)bulk|list|junk} } \
926 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
929 The Auto-Submitted: and various List- headers are standardised, whereas I
930 don't think Precedence: ever was.
932 PH/18 Refactored debugging code in route_finduser() to show more information,
933 in particular, the error code if getpwnam() issues one.
935 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
936 This is apparently needed in addition to the PH/07 change above to avoid
937 any possible encoding problems.
939 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
940 but not after initializing Perl.
942 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
943 output them only if debugging. By default they are written stderr,
944 apparently, which is not desirable.
946 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
949 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
952 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
954 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
955 authenticated or an ident call has been made. Suppress the default
956 values for $authenticated_id and $authenticated_sender (but permit -oMai
957 and -oMas) when testing with -bh.
959 PH/24 Re-jigged the order of the tests in the default configuration so that the
960 tests for valid domains and recipients precede the DNS black list and CSA
961 tests, on the grounds that those ones are more expensive.
963 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
964 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
965 This bug exists in every version of Exim that I still have, right back to
968 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
969 However, an attempt to turn on case-sensitivity in a regex key by
970 including (?-i) didn't work because the subject string was already
971 lowercased, and the effects were non-intuitive. It turns out that a
972 one-line patch can be used to allow (?-i) to work as expected.
978 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
979 other effects) broke the use of negated acl sub-conditions.
981 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
984 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
985 "Deny" causes Exim to reject the incoming connection with a 554 error.
986 Unfortunately, if there is a major crisis, such as a disk failure,
987 tcp-wrappers gives "deny", whereas what one would like would be some
988 kind of temporary error. A kludge has been added to help with this.
989 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
990 554 error is used if errno is still zero or contains ENOENT (which occurs
991 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
994 PH/03 Add -lutil to the default FreeBSD LIBS setting.
996 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
997 errors. Otherwise a message that provokes a temporary error (when other
998 messages do not) can cause a whole host to time out.
1000 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1001 addresses were routed directly to files or pipes from a redirect router.
1002 File deliveries just didn't batch; pipe deliveries might have suffered
1005 PH/06 A failure to get a lock for a hints database would erroneously always say
1006 "Failed to get write lock", even when it was really a read lock.
1008 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1009 of 0600. This has been changed to use the value of the lockfile_mode
1010 option (which defaults to 0600).
1012 PH/08 Applied small patch from the Sieve maintainer.
1014 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1015 folder from quota calculations, a direct delivery into this folder messed
1016 up the contents of the maildirsize file. This was because the regex was
1017 used only to exclude .Trash (or whatever) when the size of the mailbox
1018 was calculated. There was no check that a delivery was happening into an
1019 excluded directory. This bug has been fixed by ignoring all quota
1020 processing for deliveries into excluded directories.
1022 PH/10 Added the maildirfolder_create_regex option to appendfile.
1028 PH/01 The code for finding all the local interface addresses on a FreeBSD
1029 system running IPv6 was broken. This may well have applied to all BSD
1030 systems, as well as to others that have similar system calls. The broken
1031 code found IPv4 interfaces correctly, but gave incorrect values for the
1032 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1033 that it would not match correctly against @[] and not recognize the IPv6
1036 PH/02 The ipliteral router was not recognizing addresses of the form user@
1037 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1039 PH/03 Added disable_ipv6.
1041 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1042 decoded form, because it is most often used to construct To: headers
1043 lines in autoreplies, and the decoded form may well be syntactically
1044 invalid. However, $reply_address has leading white space removed, and all
1045 newlines turned into spaces so that the autoreply transport does not
1048 PH/05 If group was specified without a user on a router, and no group or user
1049 was specified on a transport, the group from the router was ignored.
1051 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1052 for visible compile-time settings that can be used to change these
1053 numbers, for those that want even more. Backwards compatibility with old
1054 spool files has been maintained. However, going back to a previous Exim
1055 release will lost any variables that are in spool files.
1057 PH/07 Two small changes when running in the test harness: increase delay when
1058 passing a TCP/IP connection to a new process, in case the original
1059 process has to generate a bounce, and remove special handling of
1060 127.0.0.2 (sic), which is no longer necessary.
1062 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1063 be the same on different OS.
1065 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1068 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1069 whether --show-vars was specified or not
1071 JJ/02 exipick: Added support for new ACL variable spool format introduced
1074 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1075 syntactically invalid From: or Reply-to: line, and a filter used this to
1076 generate an autoreply, and therefore failed to obtain an address for the
1077 autoreply, Exim could try to deliver to a non-existent relative file
1078 name, causing unrelated and misleading errors. What now happens is that
1079 it logs this as a hard delivery error, but does not attempt to create a
1082 PH/11 The exinext utility has a -C option for testing purposes, but although
1083 the given file was scanned by exinext itself; it wasn't being passed on
1084 when Exim was called.
1086 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1087 an end-of-file indication when reading a command response.
1089 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1090 compiled. In many other places in Exim, IPv6 addresses are always
1091 recognized, so I have changed this. It also means that IPv4 domain
1092 literals of the form [IPV4:n.n.n.n] are now always recognized.
1094 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1095 used if the router is not running as root, for example, when verifying at
1096 ACL time, or when using -bh. The debugging output from this situation was
1097 non-existent - all you got was a failure to exec. I have made two
1100 (a) Failures to set uid/gid, the current directory, or a process leader
1101 in a subprocess such as that created by queryprogram now generate
1102 suitable debugging ouput when -d is set.
1104 (b) The queryprogram router detects when it is not running as root,
1105 outputs suitable debugging information if -d is set, and then runs
1106 the subprocess without attempting to change uid/gid.
1108 PH/15 Minor change to Makefile for building test_host (undocumented testing
1111 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1112 additional section of a DNS packet that returns MX or SRV records.
1113 Instead, it always explicitly searches for A/AAAA records. This avoids
1114 major problems that occur when a DNS server includes only records of one
1115 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1116 fixed another bug: if SRV records were looked up and the corresponding
1117 address records were *not* found in the additional section, the port
1118 values from the SRV records were lost.
1120 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1121 using the correct key (the original address) when searching the retry
1122 rules in order to find which one to use for generating the retry hint.
1124 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1125 adding the default one. Similarly, if it contains a Reply-To: header, the
1126 errors_reply_to option, if set, is not used.
1128 PH/19 When calculating a retry time, Exim used to measure the "time since
1129 failure" by looking at the "first failed" field in the retry record. Now
1130 it does not use this if it is later than than the arrival time of the
1131 message. Instead it uses the arrival time. This makes for better
1132 behaviour in cases where some deliveries succeed, thus re-setting the
1133 "first failed" field. An example is a quota failure for a huge message
1134 when small messages continue to be delivered. Without this change, the
1135 "time since failure" will always be short, possible causing more frequent
1136 delivery attempts for the huge message than are intended.
1137 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1139 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1140 $1, $2, $3) because the numerical variables can be reset during some
1141 expansion items (e.g. "match"), thereby losing the authentication data.
1143 PH/21 Make -bV show the size of off_t variables so that the test suite can
1144 decide whether to run tests for quotas > 2G.
1146 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1147 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1148 filecount value is greater than 2G or if a quota value is greater than 2G
1149 on a system where the size of off_t is not greater than 4, a panic error
1152 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1153 never match. The debug and -bh output now contains an explicit error
1154 message indicating a malformed IPv4 address or mask.
1156 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1157 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1158 PH/23 above applies.
1160 PH/25 Do not write to syslog when running in the test harness. The only
1161 occasion when this arises is a failure to open the main or panic logs
1162 (for which there is an explicit test).
1164 PH/26 Added the /no_tell option to "control=freeze".
1166 PH/27 If a host name lookup failed very early in a connection, for example, if
1167 the IP address matched host_lookup and the reverse lookup yielded a name
1168 that did not have a forward lookup, an error message of the form "no IP
1169 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1170 could be logged. Now it outputs the IP address instead of "NULL".
1172 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1173 allows the sender and the authenticated sender to be set when
1174 submitting a message from within Exim. Since child_open_exim() is
1175 documented for local_scan(), the new function should be too.
1177 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1178 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1179 results in an empty string is now treated as unset.
1181 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1183 PH/31 Added sender_verify_fail logging option.
1185 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1186 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1187 tidied the source and removed it altogether.
1189 PH/33 When a queue run was abandoned because the load average was too high, a
1190 log line was always written; now it is written only if the queue_run log
1191 selector is set. In addition, the log line for abandonment now contains
1192 information about the queue run such as the pid. This is always present
1193 in "start" and "stop" lines but was omitted from the "abandon" line.
1195 PH/34 Omit spaces between a header name and the colon in the error message that
1196 is given when verify = headers_syntax fails (if there are lots of them,
1197 the message gets confusing).
1199 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1200 names, as there are now some PTR records that contain slashes. This check
1201 is only to protect against broken name servers that fall over on strange
1202 characters, so the fact that it applies to all lookups doesn't matter.
1204 PH/36 Now that the new test suite is complete, we can remove some of the
1205 special code in Exim that was needed for the old test suite. For example,
1206 sorting DNS records because real resolvers return them in an arbitrary
1207 order. The new test suite's fake resolver always returns records in the
1210 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1211 bounces) except when queue_only is set, to avoid logging races between
1212 the different processes.
1214 PH/38 Panic-die if .include specifies a non-absolute path.
1216 PH/39 A tweak to the "H" retry rule from its user.
1218 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1219 a label. They prevented compilation on older perls.
1221 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1222 a warning to be raised on newish perls.
1224 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1225 on queue. Changes to match documented behaviour of showing count of
1226 messages matching specified criteria.
1228 PH/40 Changed the default ident timeout from 30s to 5s.
1230 PH/41 Added support for the use of login_cap features, on those BSD systems
1231 that have them, for controlling the resources used by pipe deliveries.
1233 PH/42 The content-scanning code uses fopen() to create files in which to put
1234 message data. Previously it was not paying any attention to the mode of
1235 the files. Exim runs with umask(0) because the rest of the code creates
1236 files with open(), and sets the required mode explicitly. Thus, these
1237 files were ending up world-writeable. This was not a big issue, because,
1238 being within the spool directory, they were not world-accessible. I have
1239 created a function called modefopen, which takes an additional mode
1240 argument. It sets umask(777), creates the file, chmods it to the required
1241 mode, then resets the umask. All the relevant calls to fopen() in the
1242 content scanning code have been changed to use this function.
1244 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1245 to 24 hours. This avoids potential overflow problems when processing G
1246 and H retry rules. I suspect nobody ever tinkers with this value.
1248 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1250 PH/45 When the plaintext authenticator is running as a client, the server's
1251 challenges are checked to ensure they are valid base64 strings. By
1252 default, the authentication attempt is cancelled if an invalid string is
1253 received. Setting client_ignore_invalid_base64 true ignores these errors.
1254 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1255 they are received. Thus, the responses can be made to depend on the
1256 challenges. If an invalid string is ignored, an empty string is placed in
1259 PH/46 Messages that are created by the autoreply transport now contains a
1260 References: header, in accordance with RFCs 2822 and 3834.
1262 PH/47 Added authenticated_sender_force to the smtp transport.
1264 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1266 PH/49 Installed latest patch from the Sieve maintainer.
1268 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1269 was also set, the appendfile transport was unnecessarily scanning a
1270 directory of message files (e.g. for maildir delivery) to find the count
1271 of files (along with the size), even though it did not need this
1272 information. It now does the scan only if it needs to find either the
1273 size of the count of files.
1275 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1277 PH/52 Two bugs concerned with error handling when the smtp transport is
1280 (i) Exim was not creating retry information for temporary errors given
1281 for individual recipients after the DATA command when the smtp transport
1282 was used in LMTP mode. This meant that they could be retried too
1283 frequently, and not timed out correctly.
1285 (ii) Exim was setting the flag that allows error details to be returned
1286 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1287 recipients that were returned after the DATA command.
1289 PH/53 This is related to PH/52, but is more general: for any failing address,
1290 when detailed error information was permitted to be returned to the
1291 sender, but the error was temporary, then after the final timeout, only
1292 "retry timeout exceeded" was returned. Now it returns the full error as
1293 well as "retry timeout exceeded".
1295 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1296 do this, and (what is worse) MTAs that accept it.
1298 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1299 will now be deprecated.
1301 PH/56 New os.c-cygwin from the Cygwin maintainer.
1303 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1304 formats (previously only available in exim formats via -bpr, -bpru,
1305 and -bpra. Now also available in native and exiqgrep formats)
1307 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1308 with very large, slow to parse queues
1310 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1312 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1314 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1315 responses to authentication challenges, though it was showing the
1316 challenges; (ii) I've removed the CR characters from the debug output for
1319 PH/58 Allow for the insertion of a newline as well as a space when a string
1320 is turned into more than one encoded-word during RFC 2047 encoding. The
1321 Sieve code now uses this.
1323 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1324 data_4xx, lost_connection, tls_required.
1326 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1327 message was being sent as an SMTP response.
1329 PH/61 Add -l and -k options to exicyclog.
1331 PH/62 When verifying, if an address was redirected to one new address, so that
1332 verification continued, and the new address failed or deferred after
1333 having set something in $address_data, the value of $address_data was not
1334 passed back to the ACL. This was different to the case when no
1335 redirection occurred. The value is now passed back in both cases.
1337 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1338 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1339 use login_cap.h, so on its own it isn't the distinguishing feature. The
1340 new name refers directly to the setclassresources() function.
1342 PH/65 Added configuration files for NetBSD3.
1344 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1346 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1347 is preferred over IPv4.
1349 PH/68 The bounce_return_message and bounce_return_body options were not being
1350 honoured for bounces generated during the reception of non-SMTP messages.
1351 In particular, this applied to messages rejected by the ACL. This bug has
1352 been fixed. However, if bounce_return_message is true and bounce_return_
1353 body is false, the headers that are returned for a non-SMTP message
1354 include only those that have been read before the error was detected.
1355 (In the case of an ACL rejection, they have all been read.)
1357 PH/69 The HTML version of the specification is now built in a directory called
1358 spec_html instead of spec.html, because the latter looks like a path with
1359 a MIME-type, and this confuses some software.
1361 PH/70 Catch two compiler warnings in sieve.c.
1363 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1364 function verify_get_ident() calls ip_connect() to connect a socket, but
1365 if the "connect()" function timed out, ip_connect() used to close the
1366 socket. However, verify_get_ident() also closes the socket later, and in
1367 between Exim writes to the log, which may get opened at this point. When
1368 the socket was closed in ip_connect(), the log could get the same file
1369 descriptor number as the socket. This naturally causes chaos. The fix is
1370 not to close the socket in ip_connect(); the socket should be closed by
1371 the function that creates it. There was only one place in the code where
1372 this was missing, in the iplookup router, which I don't think anybody now
1373 uses, but I've fixed it anyway.
1375 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1376 well as to direct DNS lookups. Otherwise the handling of names in host
1377 lists is inconsistent and therefore confusing.
1383 PH/01 Two changes to the default runtime configuration:
1385 (1) Move the checks for relay_from_hosts and authenticated clients from
1386 after to before the (commented out) DNS black list checks.
1388 (2) Add control=submission to the relay_from_hosts and authenticated
1389 clients checks, on the grounds that messages accepted by these
1390 statements are most likely to be submissions.
1392 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1394 (1) Generate an error if the third argument for the ${prvs expansion is
1397 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1400 (3) Reset the variables that are obtained from the first argument of
1401 ${prvscheck and used in the second argument before leaving the code,
1402 because their memory is reclaimed, so using them afterwards may do
1405 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1406 one (it's much easier than Tom thought :-).
1408 (5) Because of (4), we can now allow for the use of $prvscheck_result
1409 inside the third argument.
1411 PH/03 For some reason, the default setting of PATH when running a command from
1412 a pipe transport was just "/usr/bin". I have changed it to
1415 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1416 anything to be listed in the output from -bV.
1418 PH/05 When a filter generated an autoreply, the entire To: header line was
1419 quoted in the delivery log line, like this:
1421 => >A.N.Other <ano@some.domain> <original@ddress> ...
1423 This has been changed so that it extracts the operative address. There
1424 may be more than one such address. If so, they are comma-separated, like
1427 => >ano@some.domain,ona@other.domain <original@ddress> ...
1429 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1430 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1431 not being looked up in the rDNS to get a host name, Exim was showing the
1432 IP address twice in Received: lines, even though the IP addresses were
1433 identical. For example:
1435 Received: from [1.2.3.4] (helo=[1.2.3.4])
1437 However, if the real host name was known, it was omitting the HELO data
1438 if it matched the actual IP address. This has been tidied up so that it
1439 doesn't show the same IP address twice.
1441 PH/07 When both +timestamp and +memory debugging was on, the value given by
1442 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1443 called by the debug printing, thereby overwriting the timestamp buffer.
1444 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1446 PH/08 When the original message was included in an autoreply transport, it
1447 always said "this is a copy of the message, including all the headers",
1448 even if body_only or headers_only was set. It now gives an appropriate
1451 PH/09 Applied a patch from the Sieve maintainer which:
1453 o fixes some comments
1454 o adds the (disabled) notify extension core
1455 o adds some debug output for the result of if/elsif tests
1456 o points to the current vacation draft in the documentation
1457 and documents the missing references header update
1461 o fixes a bug in processing the envelope test (when testing
1462 multiple envelope elements, the last element determinted the
1465 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1466 Electronic Mail") by including:
1468 Auto-submitted: auto-generated
1470 in the messages that it generates (bounce messages and others, such as
1471 warnings). In the case of bounce messages for non-SMTP mesages, there was
1472 also a typo: it was using "Auto_submitted" (underscore instead of
1473 hyphen). Since every message generated by Exim is necessarily in response
1474 to another message, thes have all been changed to:
1476 Auto-Submitted: auto-replied
1478 in accordance with these statements in the RFC:
1480 The auto-replied keyword:
1482 - SHOULD be used on messages sent in direct response to another
1483 message by an automatic process,
1485 - MUST NOT be used on manually-generated messages,
1487 - MAY be used on Delivery Status Notifications (DSNs) and Message
1488 Disposition Notifications (MDNs),
1490 - MUST NOT be used on messages generated by automatic or periodic
1491 processes, except for messages which are automatic responses to
1494 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1495 to the default Received: header definition.
1497 PH/12 Added log selector acl_warn_skipped (default on).
1499 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1500 variables because (a) they are in the wrong storage pool and (b) even if
1501 they were copied, it wouldn't work properly because of the caching.
1503 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1504 checking when decoding. Apparently there are clients that generate
1505 overlong encoded strings. Why am I not surprised?
1507 PH/15 If the first argument of "${if match_address" was not empty, but did not
1508 contain an "@" character, Exim crashed. Now it writes a panic log message
1509 and treats the condition as false.
1511 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1513 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1514 extension "envelope-auth". The code is finished and in agreement with
1515 other implementations, but there is no documentation so far and in fact,
1516 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1517 not changing the active code.
1519 Print executed "if" and "elsif" statements when debugging is used. This
1520 helps a great deal to understand what a filter does.
1522 Document more things not specified clearly in RFC3028. I had all this
1523 sorted out, when out of a sudden new issues came to my mind. Oops."
1525 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1528 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1529 canonical form (as documented). However, after a host name lookup from
1530 the IP address, check_host() was doing a simple string comparison with
1531 addresses acquired from the DNS when checking that the found name did
1532 have the original IP as one of its addresses. Since any found IPv6
1533 addresses are likely to be in abbreviated form, the comparison could
1534 fail. Luckily, there already exists a function for doing the comparison
1535 by converting both addresses to binary, so now that is used instead of
1536 the text comparison.
1538 PH/20 There was another similar case to PH/19, when a complete host name was
1539 given in a host list; looking up its IP address could give an abbreviated
1540 form, whereas the current host's name might or might not be abbreviated.
1541 The same fix has been applied.
1547 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1548 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1551 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1552 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1554 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1556 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1557 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1558 Also, for "Auto-Submitted", treat anything other than "no" as
1559 non-personal, in accordance with RFC 3834. (Previously it treated
1560 anything starting "auto-" as non-personal.)
1562 TF/01 The control=submission/name=... option had a problem with syntax
1563 errors if the name included a slash character. The /name= option
1564 now slurps the rest of the string, so it can include any characters
1565 but it must come last in the list of options (after /sender_retain
1568 PH/05 Some modifications to the interface to the fake nameserver for the new
1576 TK/01 Added the "success_on_redirect" address verification option. See
1577 NewStuff for rationale and an example.
1579 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1581 PH/02 Patch to exigrep to allow it to work on syslog lines.
1583 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1584 fread() to skip over the body file's header line, because in Cygwin the
1585 header line is locked and is inaccessible.
1587 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1588 co-exist for some time) to make it clear that it is the Exim ID that is
1589 referenced, not the Message-ID: header line.
1591 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1592 string_format() function, because snprintf() does not exist on all
1595 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1596 ${stat: expansion item.
1598 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1599 protocol synchronization error", to keep the pedants happy.
1601 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1602 well as for IRIX systems, when gcc is being used. See the host.c source
1605 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1607 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1610 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1611 transport and to the smtp transport in LMTP mode.
1613 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1615 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1616 run a filter in a subprocess. This could lead to confusion in subsequent
1617 lookups in the parent process. There should also be a search_tidyup() at
1618 the end of the subprocess.
1620 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1621 only if the host matched helo_try_verify_hosts, which caused the
1622 verification to occur when the EHLO/HELO command was issued. The ACL just
1623 tested the remembered result. Now, if a previous verification attempt has
1624 not happened, "verify = helo" does it there and then.
1626 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1628 TK/03 Fix log output including CR from clamd.
1630 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1631 exist provoked a memory error which could cause a segfault.
1633 PH/15 Installed PCRE 6.2
1635 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1637 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1638 of the problem. Specifically, suggested +O2 rather than +O1 for the
1641 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1643 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1644 record, the port was not correctly logged when the outgoing_port log
1645 selector was set (it logged the transort's default port).
1647 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1648 fallback_hosts, and "hosts" in the smtp transport.
1650 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1651 host errors such as "Connection refused".
1653 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1654 authentication with radiusclient 0.4.9:
1656 - Error returned from rc_read_config was caught wrongly
1657 - Username/password not passed on to radius server due to wrong length.
1659 The presumption is that some radiusclient API changes for 4.51/PH/17
1660 were not taken care of correctly. The code is still untested by me (my
1661 Linux distribution still has 0.3.2 of radiusclient), but it was
1662 contributed by a Radius user.
1664 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1665 expanding the "port" option of the smtp transport.
1667 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1668 while reading a MIME header. Thanks to Tom Hughes for a patch.
1670 PH/24 Include config.h inside local_scan.h so that configuration settings are
1673 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1674 that in an ACL for RCPT (for example), you can examine exactly what was
1677 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1678 commands, but it was not correctly comparing the address with the actual
1679 client host address. Thus, it would show the EHLO address in Received:
1680 header lines when this was not necessary.
1682 PH/27 Added the % operator to ${eval:}.
1684 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1685 it should be ignoring failures (because with -C, for example, it has lost
1686 privilege). It wasn't ignoring creation failures other than "already
1689 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1692 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1693 because an input line was too long, either on its own, or by virtue of
1694 too many continuations, the temporary file was not being removed, and the
1695 return code was incorrect.
1697 PH/31 Missing "BOOL" in function definition in filtertest.c.
1699 PH/32 Applied Sieve patches from the maintainer.
1701 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1703 PH/33 Added "verify = not_blind".
1705 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1706 Local/Makefile (with some defaults set). These are used in built scripts
1707 such as exicyclog, but they have never been used in the exim_install
1708 script (though there are many overriding facilities there). I have
1709 arranged that the exim_install script now takes note of these two
1712 PH/35 Installed configuration files for Dragonfly.
1714 PH/36 When a locally submitted message by a trusted user did not contain a
1715 From: header, and the sender address was obtained from -f or from an SMTP
1716 MAIL command, and the trusted user did not use -F to supply a sender
1717 name, $originator_name was incorrectly used when constructing a From:
1718 header. Furthermore, $originator_name was used for submission mode
1719 messages from external hosts without From: headers in a similar way,
1720 which is clearly wrong.
1722 PH/37 Added control=suppress_local_fixups.
1724 PH/38 When log_selector = +received_sender was set, and the addition of the
1725 sender made the log line's construction buffer exactly full, or one byte
1726 less than full, an overflow happened when the terminating "\n" was
1729 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1730 when the result of a list match is failure because a DNS lookup failed.
1732 PH/40 RM_COMMAND is now used in the building process.
1734 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1735 the "build-* directories that it finds.
1737 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1738 address in a domain literal was a prefix of an interface address.
1740 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1741 when verifying a sender address, unless rewrite_headers is false.
1743 PH/44 Wrote a long comment about why errors_to addresses are verified as
1744 recipients, not senders.
1746 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1747 the ratelimit ACL was added.
1749 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1751 PH/46 Added extra information about PostgreSQL errors to the error string.
1753 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1754 avoiding the need to install special zones in a real server. This is
1755 backwards compatible; if it can't find the fake resolver, it drops back.
1756 Thus, both old and new test suites can be run.
1758 TF/02 Added util/ratelimit.pl
1760 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1761 clock is set back in time.
1763 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1764 Candler <B.Candler@pobox.com>.
1766 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1767 OK for addresses that are the result of redirections.
1769 PH/48 A number of further additions for the benefit of the new test suite,
1770 including a fake gethostbyname() that interfaces to the fake DNS resolver
1773 TF/06 The fix for widen_domains has also been applied to qualify_single and
1774 search_parents which are the other dnslookup options that can cause
1777 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1780 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1782 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1783 service" if it encountered an MX record with an empty target hostname.
1784 The message is now "an MX or SRV record indicated no SMTP service".
1786 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1787 if the DNS of the sending site is misconfigured. This is quite a
1788 common situation. This change restores the behaviour of treating a
1789 helo verification defer as a failure.
1791 PH/51 If self=fail was set on a router, the bounce message did not include the
1792 actual error message.
1798 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1800 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1801 command itself ended in error, the underlying message about the transport
1802 filter timeout was being overwritten with the pipe command error. Now the
1803 underlying error message should be appended to the second error message.
1805 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1807 PH/02 Reduce the amount of output that "make" produces by default. Full output
1808 can still be requested.
1810 PH/03 The warning log line about a condition test deferring for a "warn" verb
1811 was being output only once per connection, rather than after each
1812 occurrence (because it was using the same function as for successful
1813 "warn" verbs). This seems wrong, so I have changed it.
1815 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1816 it should not have, which might have caused a crash in the right
1817 circumstances, but probably never did.
1819 PH/04 Installed a modified version of Tony Finch's patch to make submission
1820 mode fix the return path as well as the Sender: header line, and to
1821 add a /name= option so that you can make the user's friendly name appear
1824 TF/03 Added the control = fakedefer ACL modifier.
1826 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1827 Mark Lowes for thorough testing.
1829 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1831 TK/03 Merged latest SRS patch from Miles Wilton.
1833 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1834 with the definition in sysexits.h (which is #included earlier).
1835 Fortunately, Exim does not actually use EX_OK. The code used to try to
1836 preserve the sysexits.h value, by assumimg that macro definitions were
1837 scanned for macro replacements. I have been disabused of this notion,
1838 so now the code just undefines EX_OK before #including unistd.h.
1840 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1841 in the smtp transport. When a block could not be written in a single
1842 write() function, the timeout was being re-applied to each part-write.
1843 This seems wrong - if the receiver was accepting one byte at a time it
1844 would take for ever. The timeout is now adjusted when this happens. It
1845 doesn't have to be particularly precise.
1847 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1848 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1850 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1851 without a domain if the check to <postmaster@domain> fails.
1853 SC/01 Eximstats: added -xls and the ability to specify output files
1854 (patch written by Frank Heydlauf).
1856 SC/02 Eximstats: use FileHandles for outputing results.
1858 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1860 SC/04 Eximstats: fixed display of large numbers with -nvr option
1862 SC/05 Eximstats: fixed merging of reports with empty tables.
1864 SC/06 Eximstats: added the -include_original_destination flag
1866 SC/07 Eximstats: removed tabs and trailing whitespace.
1868 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1870 TK/06 MBOX spool code: Add real "From " MBOX separator line
1871 so the .eml file is really in mbox format (even though
1872 most programs do not really care). Patch from Alex Miller.
1874 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1875 The latter is generated from $received_to and is only set if the
1876 message has one envelope recipient. SA can use these headers,
1877 obviously out-of-the-box. Patch from Alex Miller.
1879 PH/08 The ${def test on a variable was returning false if the variable's
1880 value was "0", contrary to what the specification has always said!
1881 The result should be true unless the variable is empty.
1883 PH/09 The syntax error of a character other than { following "${if
1884 def:variable_name" (after optional whitespace) was not being diagnosed.
1885 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1886 accidental colon was present, for example, could give incorrect results.
1888 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1889 result is used (not including appendfile, where other changes are about
1892 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1893 This involved changing a lot of size variables from int to off_t. It
1894 should work with maildirs and everything.
1896 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1897 spamd dying while we are connected to it.
1899 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1902 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1903 function for BDB 4.3.
1905 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1907 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1908 the sources for PCRE have been re-arranged and more files are now
1911 PH/15 The code I had for printing potentially long long variables in PH/11
1912 above was not the best (it lost precision). The length of off_t variables
1913 is now inspected at build time, and an appropriate printing format (%ld
1914 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1915 to be "long long int" or "long int". This is needed for the internal
1916 formatting function string_vformat().
1918 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1919 the configuration file to be ":syslog", then the script "guesses" where
1920 the logs files are, rather than using the compiled in default. In our
1921 case the guess is not the same as the compiled default, so the script
1922 suddenly stopped working when I started to use syslog. The patch checks
1923 to see if log_file_path is "". If so, it attempts to read it from exim
1924 with no configuration file to get the compiled in version, before it
1925 falls back to the previous guessing code."
1927 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1928 implementing BATV in an Exim configuration. See NewStuff for the gory
1931 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1932 Makefile that are specific to HP-UX.
1934 PH/18 If the "use_postmaster" option was set for a recipient callout together
1935 with the "random" option, the postmaster address was used as the MAIL
1936 FROM address for the random test, but not for the subsequent recipient
1937 test. It is now used for both.
1939 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1940 patch removes a few documentation additions to RFC 3028, because the
1941 latest draft now contains them. It adds the new en;ascii-case comparator
1942 and a new error check for 8bit text in MIME parts. Comparator and
1943 require names are now matched exactly. I enabled the subaddress
1944 extension, but it is not well tested yet (read: it works for me)."
1946 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1947 rework some of the code of TK/09 above to avoid the hardwired use of
1948 "%lld" and "long long". Replaced the call to snprintf() with a call to
1951 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1952 records point to non-existent hosts", "retry timeout exceeded", and
1953 "retry time not reached for any host after a long failure period".
1955 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1956 experimental DomainKeys support:
1958 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1959 (2) On an error such as an illegally used "control", the wrong name for
1960 the control was given.
1962 These problems did NOT occur unless DomainKeys support was compiled.
1964 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1966 PH/24 Added ${if match_ip condition.
1968 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1969 databases so that it will be absolutely obvious if a crash occurs in the
1970 DB library. This is a regular occurrence (often caused by mis-matched
1973 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1974 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1975 picked up on a user's system that detects such things. There doesn't seem
1976 to be a gcc warning option for this - only an attribute that has to be
1977 put on the function's prototype. It seems that in Fedora Core 4 they have
1978 set this on a number of new functions. No doubt there will be more in due
1981 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1982 specify a transport. However, if an address that was verified by such a
1983 router was the subject of a callout, Exim crashed because it tried to
1984 read the rcpt_include_affixes from the non-existent transport. Now it
1985 just assumes that the setting of that option is false. This bug was
1986 introduced by 4.51/PH/31.
1988 PH/28 Changed -d+all to exclude +memory, because that information is very
1989 rarely of interest, but it makes the output a lot bigger. People tend to
1990 do -d+all out of habit.
1992 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1993 code in os-type was giving problems when libc.so lives in lib64, like on
1996 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1997 aren't the modern standard, and it seems that some systems' include files
1998 don't always have them. Exim was already checking for some of the newer
1999 ones like T_AAAA, and defining it itself. I've added checks for all the
2000 record types that Exim uses.
2002 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2003 not automatically generating a new one, as it is supposed to. This
2004 prevented TLS from working. If the file did exist, but contained invalid
2005 data, a new version was generated, as expected. It was only the case of a
2006 non-existent file that was broken.
2008 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2009 with a change in libdomainkeys > 0.64.
2011 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2012 from DNS. If the selector record carries the flag, it now has
2013 precedence over the domain-wide flag.
2015 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2017 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2018 the use of a port name that isn't defined in /etc/services) occurred, the
2019 message was deferred as in a normal delivery, and thus remained on the
2020 spool, instead of being failed because of the mua_wrapper setting. This
2021 is now fixed, and I tidied up some of the mua_wrapper messages at the
2024 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2025 the messages in a hash of arrays rather than using individual hashes.
2026 This is a bit cleaner and results in dramatic memory savings, albeit
2027 at a slight CPU cost.
2029 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2030 as requested by Marc Sherman.
2032 SC/10 Eximstats: added histograms for user specified patterns as requested
2035 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2037 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2038 fopen() in the content-scanning modules that did not already have it.
2044 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2045 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2047 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2049 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2051 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2052 to test Sieve filters that use "vacation".
2054 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2055 that changes the way the GnuTLS parameters are stored in the cache file.
2056 The new format can be generated externally. For backward compatibility,
2057 if the data in the cache doesn't make sense, Exim assumes it has read an
2058 old-format file, and it generates new data and writes a new file. This
2059 means that you can't go back to an older release without removing the
2062 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2063 work if there are any delivery delays because "one_time" forces the
2064 parent to be marked "delivered", so its unseen clone is never tried
2065 again. For this reason, Exim now forbids the simultaneous setting of
2068 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2069 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2070 entry if you want to know the details. The fix, however, neglected to
2071 consider the case when local delivery batching is involved. The test for
2072 "previously delivered" was not happening when checking to see if an
2073 address could be batched with a previous (undelivered) one; under
2074 certain circumstances this could lead to multiple deliveries to the same
2077 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2078 in its include files, and this causes problems building Exim.
2080 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2081 header_syntax) but Exim was just ignoring anything given after a slash.
2082 In particular, this caused confusion with an attempt to use "verify =
2083 reverse_host_lookup/defer_ok". An error is now given when options are
2084 supplied for verify items that do not have them. (Maybe reverse_host_
2085 lookup should have a defer_ok option, but that's a different point.)
2087 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2088 defined by RFC 821) to 2048, because there were problems with some AUTH
2089 commands, and RFC 1869 says the size should be increased for extended
2090 SMTP commands that take arguments.
2092 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2095 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2096 "unknown" error; now it says that the functionality isn't in the binary.
2098 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2099 an address' error message when a string expansion fails (syntax or
2100 whatever). Otherwise the password may appear in the log. Following change
2101 PH/42 below, there is no longer a chance of it appearing in a bounce
2104 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2106 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2107 of its addresses was ever tried. (Bugzilla bug #2.)
2109 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2110 the result incorrectly in the debug output. (It correctly added a newline
2111 to what was transported.)
2113 TF/01 Added $received_time.
2115 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2116 commented out examples of how to interface to a virus scanner and to
2117 SpamAssassin. Also added commented examples of av_scanner and
2118 spamd_address settings.
2120 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2121 and controls are allowed in which ACLs. There were a couple of minor
2122 errors. Some of the entries in the conditions table (which is a table of
2123 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2124 negation of where the condition IS allowed.
2126 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2128 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2129 header file does not have a version number, so I've had to invent a new
2130 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2131 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2132 radiusclient), but it was contributed by a Radius user.
2134 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2135 files or directories, for OpenSSL.
2137 PH/19 When an Exim process that is running as root has to create an Exim log
2138 file, it does so in a subprocess that runs as exim:exim so as to get the
2139 ownership right at creation (otherwise, other Exim processes might see
2140 the file with the wrong ownership). There was no test for failure of this
2141 fork() call, which would lead to the process getting stuck as it waited
2142 for a non-existent subprocess. Forks do occasionally fail when resources
2143 run out. I reviewed all the other calls to fork(); they all seem to check
2146 PH/20 When checking for unexpected SMTP input at connect time (before writing
2147 the banner), Exim was not dealing correctly with a non-positive return
2148 from the read() function. If the client had disconnected by this time,
2149 the result was a log entry for a synchronization error with an empty
2150 string after "input=" when read() returned zero. If read() returned -1
2151 (an event I could not check), uninitialized data bytes were printed.
2152 There were reports of junk text (parts of files, etc) appearing after
2155 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2157 PH/22 Added support for macro redefinition, and (re)definition in between
2158 driver and ACL definitions.
2160 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2161 forgetting to use the resulting value; it was using the unexpanded value.
2163 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2164 hadn't been configured. The fix is from Juergen Kreileder, who
2165 understands it better than I do:
2167 "Here's what I see happening with three configured cyrus_sasl
2168 authenticators configured (plain, login, cram-md5):
2170 On startup auth_cyrus_sasl_init() gets called for each of these.
2171 This means three calls to sasl_listmech() without a specified mech_list.
2172 => SASL tests which mechs of all available mechs actually work
2173 => three warnings about OTP not working
2174 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2176 With the patch, sasl_listmech() also gets called three times. But now
2177 SASL's mech_list option is set to the server_mech specified in the the
2178 authenticator. Or in other words, the answer from sasl_listmech()
2179 gets limited to just the mech you're testing for (which is different
2181 => the return list contains just 'plain' or 'login', 'cram-md5' or
2182 nothing depending on the value of ob->server_mech.
2184 I've just tested the patch: Authentication still works fine,
2185 unavailable mechs specified in the exim configuration are still
2186 caught, and the auth.log warnings about OTP are gone."
2188 PH/25 When debugging is enabled, the contents of the command line are added
2189 to the debugging output, even when log_selector=+arguments is not
2192 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2193 answer is "GNU", and only if the return is "GNU/something" is the answer
2196 PH/27 $acl_verify_message is now set immediately after the failure of a
2197 verification in an ACL, and so is available in subsequent modifiers. In
2198 particular, the message can be preserved by coding like this:
2200 warn !verify = sender
2201 set acl_m0 = $acl_verify_message
2203 Previously, $acl_verify_message was set only while expanding "message"
2204 and "log_message" when a very denied access.
2206 PH/28 Modified OS/os.c-Linux with
2208 -#ifndef OS_LOAD_AVERAGE
2209 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2211 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2212 nomenclature these days.)
2214 PH/29 Installed patch from the Sieve maintainer that adds the options
2215 sieve_useraddress and sieve_subaddress to the redirect router.
2217 PH/30 In these circumstances:
2218 . Two addresses routed to the same list of hosts;
2219 . First host does not offer TLS;
2220 . First host accepts first address;
2221 . First host gives temporary error to second address;
2222 . Second host offers TLS and a TLS session is established;
2223 . Second host accepts second address.
2224 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2225 and peerdn, if requested) that were in fact used only for the second
2228 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2229 attention to any local part prefix or suffix that was matched by the
2230 router that accepted the address. It now behaves in the same way as it
2231 does for delivery: the affixes are removed from the local part unless
2232 rcpt_include_affixes is set on the transport.
2234 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2235 timeout during the DATA phase of an incoming message.
2237 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2238 applied a patch sanctioned by the Sieve maintainer.
2240 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2241 the uid or gid is negative. A case of a negative gid caused this to be
2242 noticed. The fix allows for either to be negative.
2244 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2245 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2248 PH/36 The change PH/12 above was broken. Fixed it.
2250 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2251 the grounds that routing the same address twice would always produce the
2252 same answer. This might have been true once, but it is certainly no
2253 longer true now. Routing a child address may depend on the previous
2254 routing that produced that child. Some complicated redirection strategies
2255 went wrong when messages had multiple recipients, and made Exim's
2256 behaviour dependent on the order in which the addresses were given.
2258 I have moved the duplicate checking until after the routing is complete.
2259 Exim scans the addresses that are assigned to local and remote
2260 transports, and removes any duplicates. This means that more work will be
2261 done, as duplicates will always all be routed, but duplicates are
2262 presumably rare, so I don't expect this is of any significance.
2264 For deliveries to pipes, files, and autoreplies, the duplicate checking
2265 still happens during the routing process, since they are not going to be
2268 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2269 It corrects a timeout issue with spamd. This is Ian's comment: "The
2270 background is that sometimes spamd either never reads data from a
2271 connection it has accepted, or it never writes response data. The exiscan
2272 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2273 blindly assumes that writes won't block so it may never time out."
2275 PH/39 Allow G after quota size as well as K and M.
2277 PH/40 The value set for $authenticated_id in an authenticator may not contain
2278 binary zeroes or newlines because the value is written to log lines and
2279 to spool files. There was no check on this. Now the value is run through
2280 the string_printing() function so that such characters are converted to
2281 printable escape sequences.
2283 PH/41 $message_linecount is a new variable that contains the total number of
2284 lines in the message. Compare $body_linecount, which is the count for the
2287 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2288 bounce and delay warning messages, except in certain special cases, which
2291 (a) An SMTP error message from a remote host;
2292 (b) A message specified in a :fail: redirection;
2293 (c) A message specified in a "fail" command in a system filter;
2294 (d) A message specified in a FAIL return from the queryprogram router;
2295 (e) A message specified by the cannot_route_message router option.
2297 In these cases only, Exim does include the error details in bounce and
2298 warning messages. There are also a few cases where bland messages such
2299 as "unrouteable address" or "local delivery error" are given.
2301 PH/43 $value is now also set for the "else" part of a ${run expansion.
2303 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2304 being worked on, but at least Exim now implements the latest version to
2307 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2308 process to complete was treated as a delivery failure, a timeout while
2309 writing the message to the pipe was logged, but erroneously treated as a
2310 successful delivery. Such timeouts include transport filter timeouts. For
2311 consistency with the overall process timeout, these timeouts are now
2312 treated as errors, giving rise to delivery failures by default. However,
2313 there is now a new Boolean option for the pipe transport called
2314 timeout_defer, which, if set TRUE, converts the failures into defers for
2315 both kinds of timeout. A transport filter timeout is now identified in
2318 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2319 systems where "make" and "gmake" are different, calling "gmake" at top
2320 level broke things. I've arranged for the value of $(MAKE) to be passed
2321 from the Makefile to this script so that it can call the same version of
2325 A note about Exim versions 4.44 and 4.50
2326 ----------------------------------------
2328 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2329 changes of various kinds. As a consequence, a big documentation update was
2330 needed. This delayed the release for rather longer than seemed good, especially
2331 in the light of a couple of (minor) security issues. Therefore, the changes
2332 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2333 release. So 4.44 and 4.50 are in effect two different branches that both start
2336 I have left the 4.50 change log unchanged; it contains all the changes since
2337 4.43. The change log for 4.44 is below; many of its items are identical to
2338 those for 4.50. This seems to be the most sensible way to preserve the
2339 historical information.
2345 1. Minor wording change to the doc/README.SIEVE file.
2347 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2348 computation of the current number of files was incorrect.
2350 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2351 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2354 4. Give more explanation in the error message when the command for a transport
2355 filter fails to execute.
2357 5. There are several places where Exim runs a non-Exim command in a
2358 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2359 was being done only for the command run by the queryprogram router. It is
2360 now done for all such subprocesses. The other cases are: ${run, transport
2361 filters, and the commands run by the lmtp and pipe transports.
2363 6. Added CONFIGURE_GROUP build-time option.
2365 7. Some older OS have a limit of 256 on the maximum number of file
2366 descriptors. Exim was using setrlimit() to set 1000 as a large value
2367 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2368 systems. I've change it so that if it can't get 1000, it tries for 256.
2370 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2371 was an oversight, and furthermore, ever since the addition of extra
2372 controls (e.g. 4.43/32), the checks on when to allow different forms of
2373 "control" were broken. There should now be diagnostics for all cases when a
2374 control that does not make sense is encountered.
2376 9. Added the /retain_sender option to "control=submission".
2378 10. $recipients is now available in the predata ACL (oversight).
2380 11. Tidy the search cache before the fork to do a delivery from a message
2381 received from the command line. Otherwise the child will trigger a lookup
2382 failure and thereby defer the delivery if it tries to use (for example) a
2383 cached ldap connection that the parent has called unbind on.
2385 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2386 of $address_data from the recipient verification was clobbered by the
2387 sender verification.
2389 13. The value of address_data from a sender verification is now available in
2390 $sender_address_data in subsequent conditions in the ACL statement.
2392 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2394 15. Added a new option "connect=<time>" to callout options, to set a different
2397 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2398 was its contents. (It was OK if the option was not defined at all.)
2400 17. A "Completed" log line is now written for messages that are removed from
2401 the spool by the -Mrm option.
2403 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2404 information about exactly what failed.
2406 19. Added -dd to debug only the daemon process.
2408 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2409 handles timeouts, both on the server side and network timeouts. Renamed the
2410 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2412 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2413 It is now set to "smtps".
2415 22. $host_address is now set to the target address during the checking of
2416 ignore_target_hosts.
2418 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2419 being passed; this would have caused $sender_host_name to have been used if
2420 matching the list had actually called for a host name (not very likely,
2421 since this list is usually IP addresses). A host name is now passed as
2424 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2425 code that specifies a non-restarting handler (typically sigaction() in
2426 modern systems) in an attempt to fix a rare and obscure crash bug.
2428 25. Narrowed the window for a race in the daemon that could cause it to ignore
2429 SIGCHLD signals. This is not a major problem, because they are used only to
2430 wake it up if nothing else does.
2432 26. A malformed maildirsize file could cause Exim to calculate negative values
2433 for the mailbox size or file count. Odd effects could occur as a result.
2434 The maildirsize information is now recalculated if the size or filecount
2437 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2438 support for a long time. Removed HAVE_SYS_VFS_H.
2440 28. Installed the latest version of exipick from John Jetmore.
2442 29. In an address list, if the pattern was not a regular expression, an empty
2443 subject address (from a bounce message) matched only if the pattern was an
2444 empty string. Non-empty patterns were not even tested. This was the wrong
2445 because it is perfectly reasonable to use an empty address as part of a
2446 database query. An empty address is now tested by patterns that are
2447 lookups. However, all the other forms of pattern expect the subject to
2448 contain a local part and a domain, and therefore, for them, an empty
2449 address still always fails if the pattern is not itself empty.
2451 30. Exim went into a mad DNS loop when attempting to do a callout where the
2452 host was specified on an smtp transport, and looking it up yielded more
2453 than one IP address.
2455 31. Re-factored the code for checking spool and log partition space into a
2456 function that finds that data and another that does the check. The former
2457 is then used to implement four new variables: $spool_space, $log_space,
2458 $spool_inodes, and $log_inodes.
2460 32. The RFC2047 encoding function was originally intended for short strings
2461 such as real names; it was not keeping to the 75-character limit for
2462 encoded words that the RFC imposes. It now respects the limit, and
2463 generates multiple encoded words if necessary. To be on the safe side, I
2464 have increased the buffer size for the ${rfc2047: expansion operator from
2467 33. It is now permitted to omit both strings after an "if" condition; if the
2468 condition is true, the result is "true". As before, when the second string
2469 is omitted, a false condition yields an empty string. This makes it less
2470 cumbersome to write custom ACL and router conditions.
2472 34. Failure to deliver a bounce message always caused it to be frozen, even if
2473 there was an errors_to setting on the router. The errors_to setting is now
2476 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2477 canonical form (fully expanded) before being placed in
2478 $sender_host_address.
2480 36. The table in the code that translates DNS record types into text (T_A to
2481 "A" for instance) was missing entries for NS and CNAME. It is just possible
2482 that this could have caused confusion if both these types were looked up
2483 for the same domain, because the text type is used as part of Exim's
2484 per-process caching. But the chance of anyone hitting this buglet seems
2487 37. The dnsdb lookup has been extended in a number of ways.
2489 (1) There is a new type, "zns", which walks up the domain tree until it
2490 finds some nameserver records. It should be used with care.
2492 (2) There is a new type, "mxh", which is like "mx" except that it returns
2493 just the host names, not the priorities.
2495 (3) It is now possible to give a list of domains (or IP addresses) to be
2496 looked up. The behaviour when one of the lookups defers can be
2497 controlled by a keyword.
2499 (4) It is now possible to specify the separator character for use when
2500 multiple records are returned.
2502 38. The dnslists ACL condition has been extended: it is now possible to supply
2503 a list of IP addresses and/or domains to be looked up in a particular DNS
2506 39. Added log_selector=+queue_time_overall.
2508 40. When running the queue in the test harness, wait just a tad after forking a
2509 delivery process, to get repeatability of debugging output.
2511 41. Include certificate and key file names in error message when GnuTLS fails
2512 to set them up, because the GnuTLS error message doesn't include the name
2513 of the failing file when there is a problem reading it.
2515 42. Allow both -bf and -bF in the same test run.
2517 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2519 44. The "Exiscan patch" is now merged into the mainline Exim source.
2521 45. Sometimes the final signoff response after QUIT could fail to get
2522 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2523 < 0 before doing a fflush(). This bug looks as though it goes back to the
2524 introduction of TLS in release 3.20, but "sometimes" must have been rare
2525 because the tests only now provoked it.
2527 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2528 (this can affect the format of dates).
2530 47. exim_tidydb, when checking for the continued existence of a message for
2531 which it has found a message-specific retry record, was not finding
2532 messages that were in split spool directories. Consequently, it was
2533 deleting retry records that should have stayed in existence.
2535 48. Steve fixed some bugs in eximstats.
2537 49. The SPA authentication driver was not abandoning authentication and moving
2538 on to the next authenticator when an expansion was forced to fail,
2539 contradicting the general specification for all authenticators. Instead it
2540 was generating a temporary error. It now behaves as specified.
2542 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2543 (the order specifies the preference for clients). The order is now AES256,
2544 AES128, 3DES, ARCFOUR128.
2546 51. Small patch to Sieve code - explicitly set From: when generating an
2549 52. Exim crashed if a remote delivery caused a very long error message to be
2550 recorded - for instance if somebody sent an entire SpamAssassin report back
2551 as a large number of 550 error lines. This bug was coincidentally fixed by
2552 increasing the size of one of Exim's internal buffers (big_buffer) that
2553 happened as part of the Exiscan merge. However, to be on the safe side, I
2554 have made the code more robust (and fixed the comments that describe what
2557 53. Now that there can be additional text after "Completed" in log lines (if
2558 the queue_time_overall log selector is set), a one-byte patch to exigrep
2559 was needed to allow it to recognize "Completed" as not the last thing in
2562 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2563 patch that reportedly fixes this has been added. I am not expert enough to
2564 create a test for it. This is what the patch creator wrote:
2566 "I found a little strange behaviour of ldap code when working with
2567 Windows 2003 AD Domain, where users was placed in more than one
2568 Organization Units. When I tried to give exim partial DN, the exit code
2569 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2570 But simultaneously result of request was absolutely normal ldap result,
2571 so I produce this patch..."
2573 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2574 REFERENCE, so I have modified the code to exclude the patch when that macro
2577 55. Some experimental protocols are using DNS PTR records for new purposes. The
2578 keys for these records are domain names, not reversed IP addresses. The
2579 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2580 leaves it alone. Component reversal etc. now happens only for IP addresses.
2583 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2585 57. Double the size of the debug message buffer (to 2048) so that more of very
2586 long debug lines gets shown.
2588 58. The exicyclog utility now does better if the number of log files to keep
2589 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2591 59. Two changes related to the smtp_active_hostname option:
2593 (1) $smtp_active_hostname is now available as a variable.
2594 (2) The default for smtp_banner uses $smtp_active_hostname instead
2595 of $primary_hostname.
2597 60. The host_aton() function is supposed to be passed a string that is known
2598 to be a valid IP address. However, in the case of IPv6 addresses, it was
2599 not checking this. This is a hostage to fortune. Exim now panics and dies
2600 if the condition is not met. A case was found where this could be provoked
2601 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2602 components; fortuitously, this particular loophole had already been fixed
2603 by change 4.50/55 above.
2605 If there are any other similar loopholes, the new check in host_aton()
2606 itself should stop them being exploited. The report I received stated that
2607 data on the command line could provoke the exploit when Exim was running as
2608 exim, but did not say which command line option was involved. All I could
2609 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2610 running as the user.
2613 61. There was a buffer overflow vulnerability in the SPA authentication code
2614 (which came originally from the Samba project). I have added a test to the
2615 spa_base64_to_bits() function which I hope fixes it.
2618 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2619 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2621 63. The daemon start-up calls getloadavg() while still root for those OS that
2622 need the first call to be done as root, but it missed one case: when
2623 deliver_queue_load_max is set with deliver_drop_privilege. This is
2624 necessary for the benefit of the queue runner, because there is no re-exec
2625 when deliver_drop_privilege is set.
2627 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2628 This has been fixed.
2630 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2631 was in use, was not putting the data itself into the right store pool;
2632 consequently, it could be overwritten for a subsequent message in the same
2633 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2636 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2638 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2639 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2640 the function were treating the return as a boolean value, which happened to
2641 work because 0=false and not-0=true, but is not correct code.
2643 68. The host_aton() function was not handling scoped IPv6 addresses (those
2644 with, for example, "%eth0" on the end) correctly.
2646 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2647 negated items (that is, ~something) in unsigned ints. Some compilers
2648 apparently mutter when there is no cast.
2650 70. If an address verification called from an ACL failed, and did not produce a
2651 user-specific message (i.e. there was only a "system" message), nothing was
2652 put in $acl_verify_message. In this situation, it now puts the system
2655 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2656 session; change 4.31/43 added the unwanted input to the log line - except
2657 that it did not do this in the start of session case. It now does.
2659 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2660 This is wrong and can cause the other end to generate a synchronization
2661 error if it is another Exim or anything else that does the synchronization
2662 check. A QUIT command is no longer sent after a timeout.
2664 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2665 during host lookups.
2667 74. The defer_ok option of callout verification was not working if it was used
2668 when verifying addresses in header lines, that is, for this case:
2670 verify = header_sender/callout=defer_ok
2672 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2673 those file descriptors could be used for SMTP connections. If anything
2674 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2675 could be sent to the SMTP client, causing chaos. The daemon now opens
2676 stdin, stdout, and stderr to /dev/null when it puts itself into the
2679 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2680 log by default. The user can override this with suitable Perl magic.
2682 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2683 the log message when discard triggers, was not working for the DATA ACL or
2684 for the non-SMTP ACL.
2686 78. Error message wording change in sieve.c.
2688 79. If smtp_accept_max_per_host was set, the number of connections could be
2689 restricted to fewer than expected, because the daemon was trying to set up
2690 a new connection before checking whether the processes handling previous
2691 connections had finished. The check for completed processes is now done
2692 earlier. On busy systems, this bug wouldn't be noticed because something
2693 else would have woken the daemon, and it would have reaped the completed
2696 80. If a message was submitted locally by a user whose login name contained one
2697 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2698 It caused a spool format error. I have fixed the spool reading code. A
2699 related problem was that the "from" clause in the Received: line became
2700 illegal because of the space(s). It is now covered by ${quote_local_part.
2702 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2705 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2707 83. Patch from Sieve maintainer to fix unterminated string problem in
2708 "vacation" handling.
2710 84. Some minor changes to the Linux configuration files to help with other
2711 OS variants using glibc.
2713 85. One more patch for Sieve to update vacation handling to latest spec.
2716 ----------------------------------------------------
2717 See the note above about the 4.44 and 4.50 releases.
2718 ----------------------------------------------------
2724 1. Change 4.43/35 introduced a bug that caused file counts to be
2725 incorrectly computed when quota_filecount was set in an appendfile
2728 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2729 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2732 3. Give more explanation in the error message when the command for a transport
2733 filter fails to execute.
2735 4. There are several places where Exim runs a non-Exim command in a
2736 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2737 was being done only for the command run by the queryprogram router. It is
2738 now done for all such subprocesses. The other cases are: ${run, transport
2739 filters, and the commands run by the lmtp and pipe transports.
2741 5. Some older OS have a limit of 256 on the maximum number of file
2742 descriptors. Exim was using setrlimit() to set 1000 as a large value
2743 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2744 systems. I've change it so that if it can't get 1000, it tries for 256.
2746 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2747 was an oversight, and furthermore, ever since the addition of extra
2748 controls (e.g. 4.43/32), the checks on when to allow different forms of
2749 "control" were broken. There should now be diagnostics for all cases when a
2750 control that does not make sense is encountered.
2752 7. $recipients is now available in the predata ACL (oversight).
2754 8. Tidy the search cache before the fork to do a delivery from a message
2755 received from the command line. Otherwise the child will trigger a lookup
2756 failure and thereby defer the delivery if it tries to use (for example) a
2757 cached ldap connection that the parent has called unbind on.
2759 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2760 of $address_data from the recipient verification was clobbered by the
2761 sender verification.
2763 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2764 was its contents. (It was OK if the option was not defined at all.)
2766 11. A "Completed" log line is now written for messages that are removed from
2767 the spool by the -Mrm option.
2769 12. $host_address is now set to the target address during the checking of
2770 ignore_target_hosts.
2772 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2773 being passed; this would have caused $sender_host_name to have been used if
2774 matching the list had actually called for a host name (not very likely,
2775 since this list is usually IP addresses). A host name is now passed as
2778 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2779 code that specifies a non-restarting handler (typically sigaction() in
2780 modern systems) in an attempt to fix a rare and obscure crash bug.
2782 15. Narrowed the window for a race in the daemon that could cause it to ignore
2783 SIGCHLD signals. This is not a major problem, because they are used only to
2784 wake it up if nothing else does.
2786 16. A malformed maildirsize file could cause Exim to calculate negative values
2787 for the mailbox size or file count. Odd effects could occur as a result.
2788 The maildirsize information is now recalculated if the size or filecount
2791 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2792 support for a long time. Removed HAVE_SYS_VFS_H.
2794 18. Updated exipick to current release from John Jetmore.
2796 19. Allow an empty sender to be matched against a lookup in an address list.
2797 Previously the only cases considered were a regular expression, or an
2800 20. Exim went into a mad DNS lookup loop when doing a callout where the
2801 host was specified on the transport, if the DNS lookup yielded more than
2804 21. The RFC2047 encoding function was originally intended for short strings
2805 such as real names; it was not keeping to the 75-character limit for
2806 encoded words that the RFC imposes. It now respects the limit, and
2807 generates multiple encoded words if necessary. To be on the safe side, I
2808 have increased the buffer size for the ${rfc2047: expansion operator from
2811 22. Failure to deliver a bounce message always caused it to be frozen, even if
2812 there was an errors_to setting on the router. The errors_to setting is now
2815 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2816 canonical form (fully expanded) before being placed in
2817 $sender_host_address.
2819 24. Updated eximstats to version 1.33
2821 25. Include certificate and key file names in error message when GnuTLS fails
2822 to set them up, because the GnuTLS error message doesn't include the name
2823 of the failing file when there is a problem reading it.
2825 26. Expand error message when OpenSSL has problems setting up cert/key files.
2828 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2829 (this can affect the format of dates).
2831 28. exim_tidydb, when checking for the continued existence of a message for
2832 which it has found a message-specific retry record, was not finding
2833 messages that were in split spool directories. Consequently, it was
2834 deleting retry records that should have stayed in existence.
2836 29. eximstats updated to version 1.35
2837 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2838 1.35 - bugfix such that pie charts by volume are generated correctly
2840 30. The SPA authentication driver was not abandoning authentication and moving
2841 on to the next authenticator when an expansion was forced to fail,
2842 contradicting the general specification for all authenticators. Instead it
2843 was generating a temporary error. It now behaves as specified.
2845 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2846 (the order specifies the preference for clients). The order is now AES256,
2847 AES128, 3DES, ARCFOUR128.
2849 31. Small patch to Sieve code - explicitly set From: when generating an
2852 32. Exim crashed if a remote delivery caused a very long error message to be
2853 recorded - for instance if somebody sent an entire SpamAssassin report back
2854 as a large number of 550 error lines. This bug was coincidentally fixed by
2855 increasing the size of one of Exim's internal buffers (big_buffer) that
2856 happened as part of the Exiscan merge. However, to be on the safe side, I
2857 have made the code more robust (and fixed the comments that describe what
2860 33. Some experimental protocols are using DNS PTR records for new purposes. The
2861 keys for these records are domain names, not reversed IP addresses. The
2862 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2863 leaves it alone. Component reversal etc. now happens only for IP addresses.
2866 34. The host_aton() function is supposed to be passed a string that is known
2867 to be a valid IP address. However, in the case of IPv6 addresses, it was
2868 not checking this. This is a hostage to fortune. Exim now panics and dies
2869 if the condition is not met. A case was found where this could be provoked
2870 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2871 components; fortuitously, this particular loophole had already been fixed
2872 by change 4.50/55 or 4.44/33 above.
2874 If there are any other similar loopholes, the new check in host_aton()
2875 itself should stop them being exploited. The report I received stated that
2876 data on the command line could provoke the exploit when Exim was running as
2877 exim, but did not say which command line option was involved. All I could
2878 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2879 running as the user.
2882 35. There was a buffer overflow vulnerability in the SPA authentication code
2883 (which came originally from the Samba project). I have added a test to the
2884 spa_base64_to_bits() function which I hope fixes it.
2887 36. The daemon start-up calls getloadavg() while still root for those OS that
2888 need the first call to be done as root, but it missed one case: when
2889 deliver_queue_load_max is set with deliver_drop_privilege. This is
2890 necessary for the benefit of the queue runner, because there is no re-exec
2891 when deliver_drop_privilege is set.
2893 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2894 was in use, was not putting the data itself into the right store pool;
2895 consequently, it could be overwritten for a subsequent message in the same
2896 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2899 38. Sometimes the final signoff response after QUIT could fail to get
2900 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2901 < 0 before doing a fflush(). This bug looks as though it goes back to the
2902 introduction of TLS in release 3.20, but "sometimes" must have been rare
2903 because the tests only now provoked it.
2909 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2910 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2911 Now it returns an integer. A number of calls were still expecting a T/F
2912 return. Fortuitously, in all cases, the tests worked in OK situations,
2913 which is the norm. However, things would have gone wrong on any write
2914 failures on the smtp file descriptor. This function is used when sending
2915 messages over SMTP and also when doing verify callouts.
2917 2. When Exim is called to do synchronous delivery of a locally submitted
2918 message (the -odf or -odi options), it no longer closes stderr before doing
2921 3. Implemented the mua_wrapper option.
2923 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2925 5. Implemented the functions header_remove(), header_testname(),
2926 header_add_at_position(), and receive_remove_recipient(), and exported them
2929 6. If an ACL "warn" statement specified the addition of headers, Exim already
2930 inserted X-ACL-Warn: at the start if there was no header name. However, it
2931 was not making this test for the second and subsequent header lines if
2932 there were newlines in the string. This meant that an invalid header could
2933 be inserted if Exim was badly configured.
2935 7. Allow an ACL "warn" statement to add header lines at the start or after all
2936 the Received: headers, as well as at the end.
2938 8. Added the rcpt_4xx retry error code.
2940 9. Added postmaster_mailfrom=xxx to callout verification option.
2942 10. Added mailfrom=xxxx to the callout verification option, for verify=
2945 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2946 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2948 12. Inserted some casts to stop certain compilers warning when using pointer
2949 differences as field lengths or precisions in printf-type calls (mostly
2950 affecting debugging statements).
2952 13. Added optional readline() support for -be (dynamically loaded).
2954 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2955 same clock tick as a message's arrival, so that its received time was the
2956 same as the "first fail" time on the retry record, and that message
2957 remained on the queue past the ultimate address timeout, every queue runner
2958 would try a delivery (because it was past the ultimate address timeout) but
2959 after another failure, the ultimate address timeout, which should have then
2960 bounced the address, did not kick in. This was a "< instead of <=" error;
2961 in most cases the first failure would have been in the next clock tick
2962 after the received time, and all would be well.
2964 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2965 being recognized when the domain list was tested by the match_domain
2966 condition in an expansion string.
2968 16. Added the ${str2b64: operator.
2970 17. Exim was always calling setrlimit() to set a large limit for the number of
2971 processes, without checking whether the existing limit was already
2972 adequate. (It did check for the limit on file descriptors.) Furthermore,
2973 errors from getrlimit() and setrlimit() were being ignored. Now they are
2974 logged to the main and panic logs, but Exim does carry on, to try to do its
2975 job under whatever limits there are.
2977 18. Imported PCRE 5.0.
2979 19. Trivial typo in log message " temporarily refused connection" (the leading
2982 20. If the log selector return_path_on_delivery was set and an address was
2983 redirected to /dev/null, the delivery process crashed because it assumed
2984 that a return path would always be set for a "successful" delivery. In this
2985 case, the whole delivery is bypassed as an optimization, and therefore no
2988 21. Internal re-arrangement: the function for sending a challenge and reading
2989 a response while authentication was assuming a zero-terminated challenge
2990 string. It's now changed to take a pointer and a length, to allow for
2991 binary data in such strings.
2993 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2995 23. Exim was not respecting finduser_retries when seeking the login of the
2996 uid under which it was called; it was always trying 10 times. (The default
2997 setting of finduser_retries is zero.) Also, it was sleeping after the final
2998 failure, which is pointless.
3000 24. Implemented tls_on_connect_ports.
3002 25. Implemented acl_smtp_predata.
3004 26. If the domain in control=submission is set empty, Exim assumes that the
3005 authenticated id is a complete email address when it generates From: or
3006 Sender: header lines.
3008 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3009 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3010 chown and chgrp in /bin and hostname in /usr/bin.
3012 28. Exim was keeping the "process log" file open after each use, just as it
3013 does for the main log. This opens the possibility of it remaining open for
3014 long periods when the USR1 signal hits a daemon. Occasional processlog
3015 errors were reported, that could have been caused by this. Anyway, it seems
3016 much more sensible not to leave this file open at all, so that is what now
3019 29. The long-running daemon process does not normally write to the log once it
3020 has entered its main loop, and it closes the log before doing so. This is
3021 so that log files can straightforwardly be renamed and moved. However,
3022 there are a couple of unusual error situations where the daemon does write
3023 log entries, and I had neglected to close the log afterwards.
3025 30. The text of an SMTP error response that was received during a remote
3026 delivery was being truncated at 512 bytes. This is too short for some of
3027 the long messages that one sometimes sees. I've increased the limit to
3030 31. It is now possible to make retry rules that apply only when a message has a
3031 specific sender, in particular, an empty sender.
3033 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3034 it possible to be selective about when SMTP synchronization is enforced.
3036 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3038 32. Implemented hosts_connection_nolog.
3040 33. Added an ACL for QUIT.
3042 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3045 35. Added mailbox_size and mailbox_filecount to appendfile.
3047 36. Added control = no_multiline_responses to ACLs.
3049 37. There was a bug in the logic of the code that waits for the clock to tick
3050 in the case where the clock went backwards by a substantial amount such
3051 that the microsecond fraction of "now" was more than the microsecond
3052 fraction of "then" (but the whole seconds number was less).
3054 38. Added support for the libradius Radius client library this is found on
3055 FreeBSD (previously only the radiusclient library was supported).
3061 1. When certain lookups returned multiple values in the form name=value, the
3062 quoting of the values was not always being done properly. Specifically:
3063 (a) If the value started with a double quote, but contained no whitespace,
3065 (b) If the value contained whitespace other than a space character (i.e.
3066 tabs or newlines or carriage returns) it was not quoted.
3067 This fix has been applied to the mysql and pgsql lookups by writing a
3068 separate quoting function and calling it from the lookup code. The fix
3069 should probably also be applied to nisplus, ibase and oracle lookups, but
3070 since I cannot test any of those, I have not disturbed their existing code.
3072 2. A hit in the callout cache for a specific address caused a log line with no
3073 reason for rejecting RCPT. Now it says "Previous (cached) callout
3074 verification failure".
3076 3. There was an off-by-one bug in the queryprogram router. An over-long
3077 return line was truncated at 256 instead of 255 characters, thereby
3078 overflowing its buffer with the terminating zero. As well as fixing this, I
3079 have increased the buffer size to 1024 (and made a note to document this).
3081 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3082 when Exim is waiting for an SMTP response from a remote server, Exim
3083 restarts its select() call on the socket, thereby resetting its timeout.
3084 This is not a problem when such interrupts are rare. Somebody set up a cron
3085 job to run exiwhat every 2 minutes, which is less than the normal select()
3086 timeout (5 or 10 minutes). This meant that the select() timeout never
3087 kicked in because it was always reset. I have fixed this by comparing the
3088 time when an interrupt arrives with the time at the start of the first call
3089 to select(). If more time than the timeout has elapsed, the interrupt is
3090 treated as a timeout.
3092 5. Some internal re-factoring in preparation for the addition of Sieve
3093 extensions (by MH). In particular, the "personal" test is moved to a
3094 separate function, and given an option for scanning Cc: and Bcc: (which is
3095 not set for Exim filters).
3097 6. When Exim created an email address using the login of the caller as the
3098 local part (e.g. when creating a From: or Sender: header line), it was not
3099 quoting the local part when it contained special characters such as @.
3101 7. Installed new OpenBSD configuration files.
3103 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3104 try to make them clearer.
3106 9. Callout options, other than the timeout value, were being ignored when
3107 verifying sender addresses in header lines. For example, when using
3109 verify = header_sender/callout=no_cache
3111 the cache was (incorrectly) being used.
3113 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3114 only the Cygwin environment.
3116 11. When return_path_on_delivery was set as a log selector, if different remote
3117 addresses in the same message used different return paths and parallel
3118 remote delivery occurred, the wrong values would sometimes be logged.
3119 (Whenever a remote delivery process finished, the return path value from
3120 the most recently started remote delivery process was logged.)
3122 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3123 lines when AUTH and/or TLS are in use. This is the "received protocol"
3124 field. Exim used to use "asmtp" for authenticated SMTP, without any
3125 indication (in the protocol name) for TLS use. Now it follows the RFC and
3126 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3127 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3128 names appear in log lines as well as in Received: header lines.
3130 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3131 extensions, and comparison tests, and to fix some bugs.
3133 14. Changes to the "personal" filter test:
3135 (1) The test was buggy in that it was just doing the equivalent of
3136 "contains" tests on header lines. For example, if a user's address was
3137 anne@some.where, the "personal" test would incorrectly be true for
3139 To: susanne@some.where
3141 This test is now done by extracting each address from the header in turn,
3142 and checking the entire address. Other tests that are part of "personal"
3143 are now done using regular expressions (for example, to check local parts
3144 of addresses in From: header lines).
3146 (2) The list of non-personal local parts in From: addresses has been
3147 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3148 taken from the Sieve specification recommendations.
3150 (3) If the message contains any header line starting with "List-" it is
3151 treated as non-personal.
3153 (4) The test for "circular" in the Subject: header line has been removed
3154 because it now seems ill-conceived.
3156 15. Minor typos in src/EDITME comments corrected.
3158 16. Installed latest exipick from John Jetmore.
3160 17. If headers_add on a router specified a text string that was too long for
3161 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3162 of string_sprintf() is now avoided.
3164 18. $message_body_size was not set (it was always zero) when running the DATA
3165 ACL and the local_scan() function.
3167 19. For the "mail" command in an Exim filter, no default was being set for
3168 the once_repeat time, causing a random time value to be used if "once" was
3169 specified. (If the value happened to be <= 0, no repeat happened.) The
3170 default is now 0s, meaning "never repeat". The "vacation" command was OK
3171 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3172 (I found it when inspecting the code).
3174 20. There is now an overall timeout for performing a callout verification. It
3175 defaults to 4 times the callout timeout, which applies to individual SMTP
3176 commands during the callout. The overall timeout applies when there is more
3177 than one host that can be tried. The timeout is checked before trying the
3178 next host. This prevents very long delays if there are a large number of
3179 hosts and all are timing out (e.g. when the network connections are timing
3180 out). The value of the overall timeout can be changed by specifying an
3181 additional sub-option for "callout", called "maxwait". For example:
3183 verify = sender/callout=5s,maxwait=20s
3185 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3186 to the end before writing, but this should make it even safer).
3188 22. Exim was forgetting that it had advertised PIPELINING for the second and
3189 subsequent messages on an SMTP connection. It was also not resetting its
3190 memory on STARTTLS and an internal HELO.
3192 23. When Exim logs an SMTP synchronization error within a session, it now
3193 records whether PIPELINING has been advertised or not.
3195 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3196 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3197 rather than long int.
3199 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3201 26. Added the never_mail option to autoreply.
3207 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3208 crash if the getsockname() call failed; this can happen if a connection is
3209 closed very soon after it is established. The problem was simply in the
3210 order in which certain operations were done, causing Exim to try to write
3211 to the SMTP stream before it had set up the file descriptor. The bug has
3212 been fixed by making things happen in the correct order.
3218 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3219 before the connection was closed, thus losing the rejection response.
3221 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3222 some early Solaris releases, but causes trouble in current releases where
3223 socklen_t is defined.
3225 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3228 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3231 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3232 that releasing the top of it at the end releases what was used for sub-
3233 expansions (unless the block got too big). However, discard this block if
3234 the first thing is a variable or header, so that we can use its block when
3235 it is dynamic (useful for very large $message_headers, for example).
3237 6. Lookups now cache *every* query, not just the most recent. A new, separate
3238 store pool is used for this. It can be recovered when all lookup caches are
3239 flushed. Lookups now release memory at the end of their result strings.
3240 This has involved some general refactoring of the lookup sources.
3242 7. Some code has been added to the store_xxx() functions to reduce the amount
3243 of flapping under certain conditions.
3245 8. log_incoming_interface used to affect only the <= reception log lines. Now
3246 it causes the local interface and port to be added to several more SMTP log
3247 lines, for example "SMTP connection from", and rejection lines.
3249 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3251 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3253 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3254 could be overwritten at the end of the current message (or the start of a
3255 new message if it was set in a HELO ACL). The value is now preserved for
3256 the duration of the SMTP connection.
3258 12. If a transport had a headers_rewrite setting, and a matching header line
3259 contained an unqualified address, that address was qualified, even if it
3260 did not match any rewriting rules. The underlying bug was that the values
3261 of the flags that permit the existence of unqualified sender and recipient
3262 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3263 non-local messages, and by -bnq for local messages) were not being
3264 preserved with the message after it was received.
3266 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3267 "next input=" as part of the text comprising the host identity instead of
3268 the correct text. The code was using the same buffer for two different
3269 strings. However, depending on which order the printing function evaluated
3270 its arguments, the bug did not always show up. Under Linux, for example, my
3271 test suite worked just fine.
3273 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3274 doesn't work with some older versions of Perl. It has been changed to "my",
3275 which in any case is probably the better facility to use.
3277 15. A really picky compiler found some instances of statements for creating
3278 error messages that either had too many or two few arguments for the format
3281 16. The size of the buffer for calls to the DNS resolver has been increased
3282 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3283 for addresses that have a lot of PTR records. This alleviates a problem; it
3284 does not fully solve it.
3286 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3287 buffer now truncates the list and logs the incident, which is the same
3288 action as happens when Exim is looking up a host name and its aliases.
3289 Previously in this situation something unpredictable would happen;
3290 sometimes it was "internal error: store_reset failed".
3292 18. If a server dropped the connection unexpectedly when an Exim client was
3293 using GnuTLS and trying to read a response, the client delivery process
3294 crashed while trying to generate an error log message.
3296 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3297 string, for example:
3299 warn message = H1: something\nH2: something
3301 the text was added as a single header line from Exim's point of view
3302 though it ended up OK in the delivered message. However, searching for the
3303 second and subsequent header lines using $h_h2: did not work. This has been
3304 fixed. Similarly, if a system filter added multiple headers in this way,
3305 the routers could not see them.
3307 20. Expanded the error message when iplsearch is called with an invalid key to
3308 suggest using net-iplsearch in a host list.
3310 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3311 ACLs are no longer actually imposed (and a message to that effect is
3314 22. If a "gecos" field in a passwd entry contained escaped characters, in
3315 particular, if it contained a \" sequence, Exim got it wrong when building
3316 a From: or a Sender: header from that name. A second bug also caused
3317 incorrect handling when an unquoted " was present following a character
3318 that needed quoting.
3320 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3321 was not being matched caselessly.
3323 24. Arranged for all hyphens in the exim.8 source to be escaped with
3326 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3327 properly. Recipient callout cache records were still being keyed to include
3328 the sender, even when use_sender was set false. This led to far more
3329 callouts that were necessary. The sender is no longer included in the key
3330 when use_sender is false.
3332 26. Added "control = submission" modifier to ACLs.
3334 27. Added the ${base62d: operator to decode base 62 numbers.
3336 28. dnsdb lookups can now access SRV records.
3338 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3339 the configuration file.
3341 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3342 (-v) mode. This makes the output for a verbose queue run more intelligible.
3344 31. Added a use_postmaster feature to recipient callouts.
3346 32. Added the $body_zerocount variable, containing the number of binary zero
3347 bytes in the message body.
3349 33. The time of last modification of the "new" subdirectory is now used as the
3350 "mailbox time last read" when there is a quota error for a maildir
3353 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3355 35. Added +ignore_unknown as a special item in host lists.
3357 36. Code for decoding IPv6 addresses in host lists is now included, even if
3358 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3359 address was recognized as an IP address, but was then not correctly decoded
3360 into binary, causing unexpected and incorrect effects when compared with
3367 1. Very minor rewording of debugging text in manualroute to say "list of
3368 hosts" instead of "hostlist".
3370 2. If verify=header_syntax was set, and a header line with an unqualified
3371 address (no domain) and a large number of spaces between the end of the
3372 name and the colon was received, the reception process suffered a buffer
3373 overflow, and (when I tested it) crashed. This was caused by some obsolete
3374 code that should have been removed. The fix is to remove it!
3376 3. When running in the test harness, delay a bit after writing a bounce
3377 message to get a bit more predictability in the log output.
3379 4. Added a call to search_tidyup() just before forking a reception process. In
3380 theory, someone could use a lookup in the expansion of smtp_accept_max_
3381 per_host which, without the tidyup, could leave open a database connection.
3383 5. Added the variables $recipient_data and $sender_data which get set from a
3384 lookup success in an ACL "recipients" or "senders" condition, or a router
3385 "senders" option, similar to $domain_data and $local_part_data.
3387 6. Moved the writing of debug_print from before to after the "senders" test
3390 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3391 problems for message scanning, either using a data ACL, or using
3392 local_scan() because the Received: header was not generated till after they
3393 were called (in order to set the time as the time of reception completion).
3394 I have revised the way this works. The header is now generated after the
3395 body is received, but before the ACL or local_scan() are called. After they
3396 are run, the timestamp in the header is updated.
3402 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3403 before starting a queue runner without re-exec. This happened only when
3404 deliver_drop_privilege was set or when the Exim user was set to root. The
3405 effect of the bug was that timeouts during subsequent deliveries caused
3406 crashes instead of being properly handled. The handler is now left at its
3407 default (and expected) setting.
3409 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3410 message, again when deliver_drop_privilege is set or Exim is run as root.
3411 The bug described in (1) was not present in this case, but the tidying up
3412 of the other signals was missing. I have made the two cases consistent.
3414 3. The ignore_target_hosts setting on a manualroute router was being ignored
3415 for hosts that were looked up using the /MX notation.
3417 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3420 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3421 operated on the sender address. After changing the $sender_address to <>
3422 for the sender address verify, Exim was re-instated it as the original
3423 (before rewriting) address, but remembering that it had rewritten it, so it
3424 wasn't rewriting it again. This bug also had the effect of breaking the
3425 sender address verification caching when the sender address was rewritten.
3427 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3428 This has been changed so that if the ip literal address matches
3429 ignore_target_hosts, the router declines.
3431 7. Added expansion conditions match_domain, match_address, and match_local_
3432 part (NOT match_host).
3434 8. The placeholder for the Received: header didn't have a length field set.
3436 9. Added code to Exim itself and to exim_lock to test for a specific race
3437 condition that could lead to file corruption when using MBX delivery. The
3438 issue is with the lockfile that is created in /tmp. If this file is removed
3439 after a process has opened it but before that process has acquired a lock,
3440 there is the potential for a second process to recreate the file and also
3441 acquire a lock. This could lead to two Exim processes writing to the file
3442 at the same time. The added code performs the same test as UW imapd; it
3443 checks after acquiring the lock that its file descriptor still refers to
3444 the same named file.
3446 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3447 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3450 11. Added the smtp_active_hostname option. If used, this will typically be made
3451 to depend on the incoming interface address. Because $interface_address is
3452 not set up until the daemon has forked a reception process, error responses
3453 that can happen earlier (such as "too many connections") no longer contain
3456 12. If an expansion in a condition on a "warn" statement fails because a lookup
3457 defers, the "warn" statement is abandoned, and the next ACL statement is
3458 processed. Previously this caused the whole ACL to be aborted.
3460 13. Added the iplsearch lookup type.
3462 14. Added ident_timeout as a log selector.
3464 15. Added tls_certificate_verified as a log selector.
3466 16. Added a global option tls_require_ciphers (compare the smtp transport
3467 option of the same name). This controls incoming TLS connections.
3469 17. I finally figured out how to make tls_require_ciphers do a similar thing
3470 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3471 before starting the TLS session.
3473 18. Tabs are now shown as \t in -bP output.
3475 19. If the log selector return_path_on_delivery was set, Exim crashed when
3476 bouncing a message because it had too many Received: header lines.
3478 20. If two routers both had headers_remove settings, and the first one included
3479 a superfluous trailing colon, the final name in the first list and the
3480 first name in the second list were incorrectly joined into one item (with a
3481 colon in the middle).
3487 1. Added -C and -D options to the exinext utility, mainly to make it easier
3488 to include in the automated testing, but these could be helpful when
3489 multiple configurations are in use.
3491 2. The exinext utility was not formatting the output nicely when there was
3492 an alternate port involved in the retry record key, nor when there was a
3493 message id as well (for retries that were specific to a specific message
3494 and a specific host). It was also confused by IPv6 addresses, because of
3495 the additional colons they contain. I have fixed the IPv4 problem, and
3496 patched it up to do a reasonable job for IPv6.
3498 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3499 delivery, the log line now contains "pipelined" if PIPELINING was used.
3501 4. An SMTP transport process used to panic and die if the bind() call to set
3502 an explicit outgoing interface failed. This has been changed; it is now
3503 treated in the same way as a connect() failure.
3505 5. A reference to $sender_host_name in the part of a conditional expansion
3506 that was being skipped was still causing a DNS lookup. This no longer
3509 6. The def: expansion condition was not recognizing references to header lines
3510 that used bh_ and bheader_.
3512 7. Added the _cache feature to named lists.
3514 8. The code for checking quota_filecount in the appendfile transport was
3515 allowing one more file than it should have been.
3517 9. For compatibility with Sendmail, the command line option
3525 and sets the incoming protocol and host name (for trusted callers). The
3526 host name and its colon can be omitted when only the protocol is to be set.
3527 Note the Exim already has two private options, -pd and -ps, that refer to
3528 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3529 "s", but I don't think that's a major issue.
3531 10. A number of refactoring changes to the code, none of which should affect
3534 (a) The number of logging options was getting close to filling up the
3535 32-bit word that was used as a bit map. I have split them into two classes:
3536 those that are passed in the argument to log_write(), and those that are
3537 only ever tested independently outside of that function. These are now in
3538 separate 32-bit words, so there is plenty of room for expansion again.
3539 There is no change in the user interface or the logging behaviour.
3541 (b) When building, for example, log lines, the code previously used a
3542 macro that called string_cat() twice, in order to add two strings. This is
3543 not really sufficiently general. Furthermore, there was one instance where
3544 it was actually wrong because one of the argument was used twice, and in
3545 one call a function was used. (As it happened, calling the function twice
3546 did not affect the overall behaviour.) The macro has been replaced by a
3547 function that can join an arbitrary number of extra strings onto a growing
3550 (c) The code for expansion conditions now uses a table and a binary chop
3551 instead of a serial search (which was left over from when there were very
3552 few conditions). Also, it now recognizes conditions like "pam" even when
3553 the relevant support is not compiled in: a suitably worded error message is
3554 given if an attempt is made to use such a condition.
3556 11. Added ${time_interval:xxxxx}.
3558 12. A bug was causing one of the ddress fields not to be passed back correctly
3559 from remote delivery subprocesses. The field in question was not being
3560 subsequently used, so this caused to problems in practice.
3562 13. Added new log selectors queue_time and deliver_time.
3564 14. Might have fixed a bug in maildirsizefile handling that threw up
3565 "unexpected character" debug warnings, and recalculated the data
3566 unnecessarily. In any case, I expanded the warning message to give more
3569 15. Added the message "Restricted characters in address" to the statements in
3570 the default ACL that block characters like @ and % in local parts.
3572 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3573 Three changes have been made:
3575 (a) There was a serious bug; a negative response to MAIL caused the whole
3576 recipient domain to be cached as invalid, thereby blocking all messages
3577 to all local parts at the same domain, from all senders. This bug has
3578 been fixed. The domain is no longer cached after a negative response to
3579 MAIL if the sender used is not empty.
3581 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3584 (c) A new callout option, "use_sender" has been added for people who want
3585 the modified behaviour.
3591 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3594 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3595 indeed breaks things for older releases.
3597 3. Added additional logging to the case where there is a problem reading data
3598 from a filter that is running in a subprocess using a pipe, in order to
3599 try to track down a specific problem.
3601 4. Testing facility fudge: when running in the test harness and attempting
3602 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3603 getting "No route to host". Convert this to a timeout.
3605 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3608 6. Some OS don't have socklen_t but use size_t instead. This affects the
3609 fifth argument of getsockopt() amongst other things. This is now
3610 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3611 can be set for individual OS. I have set it for SunOS5, OSF1, and
3612 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3613 some earlier ones do not.
3615 7. Change 4.30/15 was not doing the test caselessly.
3617 8. The standard form for an IPv6 address literal was being rejected by address
3618 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3619 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3620 this, as well as the form without the "IPv6" on the front (but only when
3621 address literals are enabled, of course).
3623 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3625 10. Exim crashed if a message with an empty sender address specified by -f
3626 encountered a router with an errors_to setting. This could be provoked only
3627 by a command such as
3631 where an empty string was supplied; "<>" did not hit this bug.
3633 11. Installed PCRE release 4.5.
3635 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3636 remained set. It is now erased.
3638 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3639 times from message ids (which are base 36 rather than the normal 62).
3641 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3642 were being counted as actual protocol errors, and logged if the log
3643 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3644 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3645 and DATA following a set of rejected RCPTs do not count as protocol errors.
3646 In other words, Exim assumes they were pipelined, though this may not
3647 actually be the case. Of course, in all cases the client gets an
3648 appropriate error code.
3650 15. If a lookup fails in an ACL condition, a message about the failure may
3651 be available; it is used if testing the ACL cannot continue, because most
3652 such messages specify what the cause of the deferral is. However, some
3653 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3654 that caused an old message to be retained and used if a later statement
3655 caused a defer, replacing the real cause of the deferral.
3657 16. If an IP address had so many PTR records that the DNS lookup buffer
3658 was not large enough to hold them, Exim could crash while trying to process
3659 the truncated data. It now detects and logs this case.
3661 17. Further to 4.21/58, another change has been made: if (and only if) the
3662 first line of a message (the first header line) ends with CRLF, a bare LF
3663 in a subsequent header line has a space inserted after it, so as not to
3664 terminate the header.
3666 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3667 unnecessarily, used atoi() instead of strtol(), and didn't check the
3668 termination when getting file sizes from file names by regex.
3670 19. Completely re-implemented the support for maildirsize files, in the light
3671 of a number of problems with the previous contributed implementation
3672 (4.30/29). In particular:
3674 . If the quota is zero, the maildirsize file is maintained, but no quota is
3677 . If the maildir directory does not exist, it is created before any attempt
3678 to write a maildirsize file.
3680 . The quota value in the file is just a cache; if the quota is changed in
3681 the transport, the new value overrides.
3683 . A regular expression is available for excluding directories from the
3686 20. The autoreply transport checks the characters in options that define the
3687 message's headers; it allows continued headers, but it was checking with
3688 isspace() after an embedded newline instead of explicitly looking for a
3691 21. If all the "regular" hosts to which an address was routed had passed their
3692 expiry times, and had not reached their retry times, the address was
3693 bounced, even if fallback hosts were defined. Now Exim should go on to try
3696 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3697 equivalent code in the SMTP transport. Some hosts send humungous responses
3698 to HELO/EHLO, more than 1024 it seems.
3700 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3701 gives compiler warnings in some environments. I've now done it "properly",
3704 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3705 (because of problems with the built-in one) was declared to return uschar *
3706 instead of char *, causing compiler failure.
3708 25. Fixed a file descriptor leak when processing alias/forward files.
3710 26. Fixed a minor format string issue in dbfn.c.
3712 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3714 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3715 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3718 29. When a sender address is verified, it is cached, to save repeating the test
3719 when there is more than one recipient in a message. However, when the
3720 verification involves a callout, it is possible for different callout
3721 options to be set for different recipients. It is too complicated to keep
3722 track of this in the cache, so now Exim always runs a verification when a
3723 callout is required, relying on the callout cache for the optimization.
3724 The overhead is duplication of the address routing, but this should not be
3727 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3728 to be verified with callout=postmaster, and the main callout worked but the
3729 postmaster check failed, the verification correctly failed. However, if a
3730 subsequent RCPT command asked for sender verification *without* the
3731 postmaster check, incorrect caching caused this verification also to fail,
3734 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3735 it was not caching the DNS options (qualify_single, search_parents) that
3736 were used when the lookup failed. A subsequent lookup with different
3737 options therefore always gave the same answer, though there were cases
3738 where it should not have. (Example: a "domains = !$mx_any" option on a
3739 dnslookup router: the "domains" option is always processed without any
3740 widening, but the router might have qualify_single set.) Now Exim uses the
3741 cached value only when the same options are set.
3743 32. Added John Jetmore's "exipick" utility to the distribution.
3745 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3746 than a timeout (e.g. a certificate is required, and is not provided), an
3747 Exim server now closes the connection immediately. Previously it waited for
3748 the client to close - but if the client is SSL, it seems that they each
3749 wait for each other, leading to a delay before one of them times out.
3751 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3752 maintained 0.8.x compatibility because I don't think many are using it, and
3753 it is clearly obsolete.
3755 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3758 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3759 client certificate was expired. A simple patch fixes this, though I don't
3760 understand the full logic of why the verify callback is called multiple
3763 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3764 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3765 which causes problems with some clients (such as the Certicom SSL Plus
3766 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3767 disables the coutermeasure allowing Eudora to connect."
3769 38. Exim was not checking that a write() to a log file succeeded. This could
3770 lead to Bad Things if a log got too big, in particular if it hit a file
3771 size limit. Exim now panics and dies if it cannot write to a log file, just
3772 as it does if it cannot open a log file.
3774 39. Modified OS/Makefile-Linux so that it now contains
3776 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3778 The two -D definitions ensure that Exim is compiled with large file
3779 support, which makes it possible to handle log files that are bigger than
3782 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3783 instance) a domain was checked against a named list that involved a lookup,
3784 causing $domain_data to be set, then another domain was checked against the
3785 same list, then the first domain was re-checked, the value of $domain_data
3786 after the final check could be wrong. In particular, if the second check
3787 failed, it could be set empty. This bug probably also applied to
3790 41. The strip_trailing_dot option was not being applied to the address given
3791 with the -f command-line option.
3793 42. The code for reading a message's header from the spool was incrementing
3794 $received_count, but never initializing it. This meant that the value was
3795 incorrect (doubled) while delivering a message in the same process in which
3796 it was received. In the most common configuration of Exim, this never
3797 happens - a fresh exec is done - but it can happen when
3798 deliver_drop_privilege is set.
3800 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3801 it now includes up to 150 characters of the unexpected data in the log
3804 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3805 and building data strings. The size of both of these buffers was 10 000
3806 bytes - far larger than anybody would *ever* want, thought I. Needless to
3807 say, somebody hit the limit. I have increased the maximum line length to
3808 20 000 and the maximum data length of concatenated lines to 100 000. I have
3809 also fixed two bugs, because there was no checking on these buffers. Tsk,
3810 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3811 buffer is too small.
3813 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3814 lsearch lookups. Now it does.
3816 46. When parsing a route_list item in a manualroute router, a fixed-length
3817 buffer was used for the list of hosts. I made this 1024 bytes long,
3818 thinking that nobody would ever have a list of hosts that long. Wrong.
3819 Somebody had a whole pile of complicated expansion conditions, and the
3820 string was silently truncated, leading to an expansion error. It turns out
3821 that it is easier to change to an unlimited length (owing to other changes
3822 that have happened since this code was originally written) than to build
3823 structure for giving a limitation error. The length of the item that
3824 expands into the list of hosts is now unlimited.
3826 47. The lsearch lookup could not handle data where the length of text line was
3827 more than 4095 characters. Such lines were truncated, leading to shortened
3828 data being returned. It should now handle lines of any length.
3830 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3831 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3834 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3835 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3836 no longer happens in comment lines. A list of replacements is now placed
3837 at the head of all of the source files, except those whose only change is
3838 to replace PERL_COMMAND in the very first #! line.
3840 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3841 messages on the queue, with a bottom-up merge sort, using code contributed
3842 by Michael Haardt. This should make operations like -bp somewhat faster on
3843 large queues. It won't affect queue runners, except when queue_run_in_order
3846 51. Installed eximstats 1.31 in the distribution.
3848 52. Added support for SRV lookups to the dnslookup router.
3850 53. If an ACL referred to $message_body or $message_body_end, the value was not
3851 reset for any messages that followed in the same SMTP session.
3853 54. The store-handling optimization for building very long strings was not
3854 differentiating between the different store pools. I don't think this
3855 actually made any difference in practice, but I've tidied it.
3857 55. While running the routers to verify a sender address, $sender_address
3858 was still set to the sender address. This is wrong, because when routing to
3859 send a bounce to the sender, it would be empty. Therefore, I have changed
3860 it so that, while verifying a sender address, $sender_address is set to <>.
3861 (There is no change to what happens when verifying a recipient address.)
3863 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3864 target A or AAAA records (if not already returned) without resetting the
3865 qualify_single or search_parents options of the DNS resolver. These are
3866 inappropriate in this case because the targets of MX and SRV records must
3867 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3868 target that, when qualified, matched something in the local domain. These
3869 two options are now turned off when doing these lookups.
3871 57. It seems that at least some releases of Reiserfs (which does not have the
3872 concept of a fixed number of inodes) returns zero and not -1 for the
3873 number of available inodes. This interacted badly with check_spool_inodes,
3874 which assumed that -1 was the "no such thing" setting. What I have done is
3875 to check that the total number of inodes is greater than zero before doing
3876 the test of how many are available.
3878 58. When a "warn" ACL statement has a log_message modifier, the message is
3879 remembered, and not repeated. This is to avoid a lot of repetition when a
3880 message has many recipients that cause the same warning to be written.
3881 Howewer, Exim was preserving the list of already written lines for an
3882 entire SMTP session, which doesn't seem right. The memory is now reset if a
3883 new message is started.
3885 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3886 debugging output unless log_rewrite was also set.
3888 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3889 of (int)(handle) when we know that handle contains (void *)(-1).
3891 61. The Exim daemon panic-logs an error return when it closes the incoming
3892 connection. However "connection reset by peer" seems to be common, and
3893 isn't really an error worthy of noting specially, so that particular error
3896 62. When Exim is trying to find all the local interfaces, it used to panic and
3897 die if the ioctl to get the interface flags failed. However, it seems that
3898 on at least one OS (Solaris 9) it is possible to have an interface that is
3899 included in the list of interfaces, but for which you get a failure error
3900 for this call. This happens when the interface is not "plumbed" into a
3901 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3902 failure of the "get flags" call assumes that the interface is down.
3904 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3905 makes life easier for people who are doing arithmetic on fields extracted
3906 from dates, where you often get leading zeros that should not be
3907 interpreted as octal.
3909 64. Added qualify_domain to the redirect router, to override the global
3912 65. If a pathologically long header line contained very many addresses (the
3913 report of this problem mentioned 10 000) and each of them was rewritten,
3914 Exim could use up a very large amount of memory. (It kept on making new
3915 copies of the header line as it rewrote, and never released the old ones.)
3916 At the expense of a bit more processing, the header rewriting function has
3917 been changed so that it no longer eats memory in this way.
3919 66. The generation of the Received: header has been moved from the time that a
3920 message starts to be received, to the time that it finishes. The timestamp
3921 in the Received: header should now be very close to that of the <= log
3922 line. There are two side-effects of this change:
3924 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3925 logged header lines no longer include the local Received: line, because
3926 it has not yet been created. The same applies to a copy of the message
3927 that is returned to a non-SMTP sender when a message is rejected.
3929 (b) When a filter file is tested using -bf, no additional Received: header
3930 is added to the test message. After some thought, I decided that this
3933 This change does not affect the value of $received_for. It is still set
3934 after address rewriting, but before local_scan() is called.
3936 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3938 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3939 gave an unhelpful panic error message, and a defer error. I have managed to
3940 change this behaviour so that it now rejects any supplied certificate,
3941 which seems right, as the list of acceptable certificates is empty.
3943 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3944 gave an unhelpful defer error. I have not managed to make this reject any
3945 supplied certificates, but the error message it gives is "no certificate
3946 supplied", which is not helpful.
3948 70. exigrep's output now also includes lines that are not associated with any
3949 message, but which match the given pattern. Implemented by a patch from
3950 Martin Sluka, which also tidied up the Perl a bit.
3952 71. Recipient callout verification, like sender verification, was using <> in
3953 the MAIL FROM command. This isn't really the right thing, since the actual
3954 sender may affect whether the remote host accepts the recipient or not. I
3955 have changed it to use the actual sender in the callout; this means that
3956 the cache record is now keyed on a recipient/sender pair, not just the
3957 recipient address. There doesn't seem to be a real danger of callout loops,
3958 since a callout by the remote host to check the sender would use <>.
3959 [SEE ABOVE: changed after hitting problems.]
3961 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3962 temporary errors. However, in the case of such a code being given after
3963 the end of a data transmission (i.e. after ".") Exim was failing to write
3964 a retry record for the message. (Yes, there was some broken host that was
3965 actually sending 8xx at this point.)
3967 73. An unknown lookup type in a host list could cause Exim to panic-die when
3968 the list was checked. (An example that provoked this was putting <; in the
3969 middle of a list instead of at the start.) If this happened during a DATA
3970 ACL check, a -D file could be left lying around. This kind of configuration
3971 error no longer causes Exim to die; instead it causes a defer errror. The
3972 incident is still logged to the main and panic logs.
3974 74. Buglet left over from Exim 3 conversion. The message "too many messages
3975 in one connection" was written to the rejectlog but not the mainlog, except
3976 when address rewriting (yes!) was being logged.
3978 75. Added write_rejectlog option.
3980 76. When a system filter was run not as root (that is, when system_filter_user
3981 was set), the values of the $n variables were not being returned to the
3982 main process; thus, they were not subsequently available in the $sn
3985 77. Added +return_path_on_delivery log selector.
3987 78. A connection timeout was being treated differently from recipients deferred
3988 when testing hosts_max_try with a message that was older than the host's
3989 retry timeout. (The host should not be counted, thus allowing all hosts to
3990 be tried at least once before bouncing.) This may have been the cause of an
3991 occasionally reported bug whereby a message would remain on the queue
3992 longer than the retry timeout, but would be bounced if a delivery was
3993 forced. I say "may" because I never totally pinned down the problem;
3994 setting up timeout/retry tests is difficult. See also the next item.
3996 79. The ultimate address timeout was not being applied to errors that involved
3997 a combination of host plus message (for example, a timeout on a MAIL
3998 command). When an address resolved to a number of possible hosts, and they
3999 were not all tried for each delivery (e.g. because of hosts_max_try), a
4000 message could remain on the queue longer than the retry timeout.
4002 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4005 81. Fixed an obscure SMTP outgoing bug which required at least the following
4006 conditions: (a) there was another message waiting for the same server;
4007 (b) the server returned 5xx to all RCPT commands in the first message so
4008 that the message was not completed; (c) the server dropped the connection
4009 or gave a negative response to the RSET that Exim sends to abort the
4010 transaction. The observed case was a dropped connection after DATA that had
4011 been sent in pipelining mode. That is, the server had advertised PIPELINING
4012 but was not implementing it correctly. The effect of the bug was incorrect
4013 behaviour, such as trying another host, and this could lead to a crash.
4019 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4020 and daemon.c were passed as pointers to ints; they should have been
4021 pointers to socklen_t variables (which are typically unsigned ints).
4023 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4026 3. Fixed a really odd bug that affected only the testing scheme; patching a
4027 certain fixed string in the binary changed the value of another string that
4028 happened to be identical to the end of the original first string.
4030 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4031 name", it returns that address as the IP address. On some operating
4032 systems (e.g. Solaris), it also passes back the IP address string as the
4033 "host name". However, on others (e.g. Linux), it passes back an empty
4034 string. Exim wasn't checking for this, and was changing the host name to an
4035 empty string, assuming it had been canonicized.
4037 5. Although rare, it is permitted to have more than one PTR record for a given
4038 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4039 all the names associated with an address, because they do in Solaris.
4040 However, it seems that they do not in Linux for data that comes from the
4041 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4042 I found this out when I moved to a new Linux workstation and tried to run
4043 the Exim test suite.
4045 To get round this problem I have changed the code so that it now does its
4046 own call to the DNS to look up PTR records when searching for a host name.
4047 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4048 addresses that are only in /etc/hosts are still found.
4050 This behaviour is, however, controlled by an option called host_lookup_
4051 order, which defaults to "bydns:byaddr". If people want to use the other
4052 order, or indeed, just use one or the other means of lookup, they can
4053 specify it in this variable.
4055 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4056 some operating systems, this comes back from gethostbyaddr() as an empty
4057 string, and this is what Exim used to test for. However, it seems that in
4058 other systems, "." is yielded. Exim now tests for this case too.
4060 7. The values of check_spool_space and check_log_space are now held internally
4061 as a number of kilobytes instead of an absolute number of bytes. If a
4062 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4063 kilobyte. This means that much larger values can be stored.
4065 8. Exim monitor: an attempt to get the action menu when not actually pointing
4066 at a message produces an empty menu entitled "No message selected". This
4067 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4068 no entries in it ("Shell widget menu has zero width and/or height"). So I
4069 have added a single, blank menu entry in this case.
4071 9. Added ${quote_local_part.
4073 10. MIME decoding is now applied to the contents of Subject: header lines when
4076 11. Now that a reference to $sender_host_address automatically causes a reverse
4077 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4078 host lookup before query-style lookups in lists that might use this
4079 variable. This has therefore been abolished, and the "net-" prefix is no
4080 longer necessary for query-style lookups.
4082 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4083 have been a typo for LFLAGS, so it has been changed.
4085 13. The install script calls Exim with "-C /dev/null" in order to find the
4086 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4087 to be output. Howeve, since Exim outputs its version number before the
4088 error, it didn't break the script. It just looked ugly. I fixed this by
4089 always allowing "-C /dev/null" if the caller is root.
4091 14. Ignore overlarge ACL variable number when reading spool file - insurance
4092 against a later release with more variables having written the file.
4094 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4095 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4096 this, as well as the form without the "IPv6" on the front.
4098 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4099 OS/Makefile-Darwin file.
4101 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4102 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4104 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4105 variable. However, some OS use ints for inodes. I've added cast to long int
4106 to get rid of the compiler warning.
4108 19. I had forgotten to lock out "/../" in configuration file names when
4109 ALT_CONFIG_PREFIX was set.
4111 20. Routers used for verification do not need to specify transports. However,
4112 if such a router generated a host list, and callout was configured, Exim
4113 crashed, because it could not find a port number from the (non-existent)
4114 transport. It now assumes port 25 in this circumstance.
4116 21. Added the -t option to exigrep.
4118 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4119 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4120 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4121 them is set, LOOKUP_LSEARCH is forced.
4123 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4125 24. Added sender and host information to the "rejected by local_scan()" log
4126 line; previously there was no indication of these.
4128 25. Added .include_if_exists.
4130 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4131 Linux. It turns out that not all file systems support this. Apparently some
4132 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4133 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4134 not supported on the file descriptor, is now ignored when Exim is trying to
4135 sync a directory. This applies only to Linux.
4137 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4139 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4140 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4143 29. Added support for maildirsize files from supplied patch (modified a bit).
4145 30. The use of :fail: followed by an empty string could lead Exim to respond to
4146 sender verification failures with (e.g.):
4148 550 Verification failed for <xxx>
4149 550 Sender verify failed
4151 where the first response line was missing the '-' that indicates it is not
4152 the final line of the response.
4154 31. The loop for finding the name of the user that called Exim had a hardwired
4155 limit of 10; it now uses the value of finduser_retries, which is used for
4156 all other user lookups.
4158 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4161 33. Exim was neglecting to zero errno before one call of strtol() when
4162 expanding a string and expecting an integer value. On some systems this
4163 resulted in spurious "integer overflow" errors. Also, it was casting the
4164 result into an int without checking.
4166 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4167 did not work. The code looks as if it has *never* worked, though it appears
4168 to have been documented since at least releast 1.62. I have made it work.
4170 35. The "timeout_DNS" error in retry rules, also documented since at least
4171 1.62, also never worked. As it isn't clear exactly what this means, and
4172 clearly it isn't a major issue, I have abolished the feature by treating it
4173 as "timeout", and writing a warning to the main and panic logs.
4175 36. The display of retry rules for -brt wasn't always showing the error code
4178 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4179 timeout_connect_A, timeout_connect_MX.
4181 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4182 to the empty sender.
4184 39. The daemon was not analysing the content of -oX till after it had closed
4185 stderr and disconnected from the controlling terminal. This meant that any
4186 syntax errors were only noted on the panic log, and the return code from
4187 the command was 0. By re-arranging the code a little, I've made the
4188 decoding happen first, so such errors now appear on stderr, and the return
4189 code is 1. However, the actual setting up of the sockets still happens in
4190 the disconnected process, so errors there are still only recorded on the
4193 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4194 connections (as happens on some IP stacks) was logged at start up time as
4195 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4196 it from "IPv6 and IPv4", which means that two separate sockets are being
4199 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4200 says whether AF_INET or AF_INET6 was passed as an argument.
4202 42. Exiwhat output was messed up when time zones were included in log
4205 43. Exiwhat now gives more information about the daemon's listening ports,
4206 and whether -tls-on-connect was used.
4208 44. The "port" option of the smtp transport is now expanded.
4210 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4211 silently ignored. Now an error message is written to the main and panic
4214 46. There's a new ACL modifier called "logwrite" which writes to a log file
4215 as soon as it is encountered.
4217 47. Added $local_user_uid and $local_user_gid at routing time.
4219 48. Exim crashed when trying to verify a sender address that was being
4222 49. Exim was recognizing only a space character after ".include". It now also
4223 recognizes a tab character.
4225 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4226 extracting the relevant information from the specification. The man page no
4227 longer contains scrambled data for the -d option, and I've added a section
4228 at the front about calling Exim under different names.
4230 51. Added "extra_headers" argument to the "mail" command in filter files.
4232 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4235 53. Installed eximstats 1.29.
4237 54. Added transport_filter_timeout as a generic transport option.
4239 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4240 Cc: header lines. This was required by RFC 822, but it not required by RFC
4243 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4244 incoming messages that did not have them. Now it does so only if the
4245 message originates locally, that is, if there is no associated remote host
4246 address. When Resent- header lines are present, this applies to the Resent-
4247 lines rather than the non-Resent- lines.
4249 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4250 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4252 58. Messages for configuration errors now include the name of the main
4253 configuration file - useful now that there may be more than one file in a
4254 list (.included file names were always shown).
4256 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4257 for those rare installations that do not start the daemon as root or run it
4258 setuid root. I've cut out the call to initgroups() if the daemon is not
4261 60. The Exim user and group can now be bound into the binary as text strings
4262 that are looked up at the start of Exim's processing.
4264 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4266 62. Added $mailstore_basename variable.
4268 63. Installed patch to sieve.c from Michael Haardt.
4270 64. When Exim failed to open the panic log after failing to open the main log,
4271 the original message it was trying to log was written to stderr and debug
4272 output, but if they were not available (the usual case in production), it
4273 was lost. Now it is written to syslog before the two lines that record the
4274 failures to open the logs.
4276 65. Users' Exim filters run in subprocesses under the user's uid. It is
4277 possible for a "deliver" command or an alias in a "personal" command to
4278 provoke an address rewrite. If logging of address rewriting is configured,
4279 this fails because the process is not running as root or exim. There may be
4280 a better way of dealing with this, but for the moment (because 4.30 needs
4281 to be released), I have disabled address rewrite logging when running a
4282 filter in a non-root, non-exim process.
4288 1. The buildconfig auxiliary program wasn't quoting the value set for
4289 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4290 not defined. This bug was masked in 4.22 by the effect that was fixed in
4293 2. Some messages that were rejected after a message id was allocated were
4294 shown as "incomplete" by exigrep. It no longer does this for messages that
4295 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4297 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4298 have allow_domain_literals set, the ID did not get logged in the <= line.
4299 Domain literals are now always recognized in Message-ID: header lines.
4301 4. The first argument for a ${extract expansion item is the key name or field
4302 number. Leading and trailing spaces in this item were not being ignored,
4303 causing some misleading effects.
4305 5. When deliver_drop_privilege was set, single queue runner processes started
4306 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4307 same command in the process it spins off) were not dropping privilege.
4309 6. When the daemon running as "exim" started a queue runner, it always
4310 re-executed Exim in the spun-off process. This is a waste of effort when
4311 deliver_drop_privilege is set. The new process now just calls the
4312 queue-runner function directly.
4318 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4321 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4322 ignored. Though the use of -oP was forcing the writing of a pid file, it
4323 was always written to the default place.
4325 3. If the message "no IP address found for host xxxx" is generated during
4326 incoming verification, it is now followed by identification of the incoming
4327 connection (so you can more easily find what provoked it).
4329 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4331 5. Added some features to "harden" Exim a bit more against certain attacks:
4333 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4334 be put in Local/Makefile. This is like the never_users runtime option,
4335 but it cannot be overridden. The default setting is "root".
4337 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4338 prefix string with which any file named in a -C command line option
4341 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4342 is retained for -C and -D only if the caller of Exim is root. Without
4343 it, the exim user may also use -C and -D and retain privilege.
4345 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4346 command line option is disabled.
4348 6. Macro names set by the -D option must start with an upper case letter, just
4349 like macro names defined in the configuration file.
4351 7. Added "dereference=" facility to LDAP.
4353 8. Two instances of the typo "uknown" in the source files are fixed.
4355 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4356 the Configure-Makefile script screwed up while processing it.
4358 10. Incorporated PCRE 4.4.
4360 11. The SMTP synchronization check was not operating right at the start of an
4361 SMTP session. For example, it could not catch a HELO sent before the client
4362 waited for the greeting. There is now a check for outstanding input at the
4363 point when the greeting is written. Because of the duplex, asynchronous
4364 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4365 way, but not yet received, when the check is performed.
4367 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4368 on TCP/IP sockets, because this apparently causes some broken clients to
4371 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4372 unchanged) from the Cygwin maintainer.
4374 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4375 was supported instead of testing for mbx. Effectively a typo.
4377 15. The spa authenticator server code was not checking that the input it
4378 received was valid base64.
4380 16. The debug output line for the "set" modifier in ACLs was not showing the
4381 name of the variable that was being set.
4383 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4385 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4386 lookup on its own. Something else was needed to trigger the lookup. For
4387 example, a match in host_lookup or the need for a host name in a host list.
4388 Now, if $sender_host_name is referenced and the host name has not yet been
4389 looked up, a lookup is performed. If the lookup fails, the variable remains
4390 empty, and $host_lookup_failed is set to "1".
4392 19. Added "eqi" as a case-independent comparison operator.
4394 20. The saslauthd authentication condition could segfault if neither service
4395 nor realm was specified.
4397 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4398 error message that was logged was misleading, and incoming SMTP
4399 connections were dropped. The message is now more accurate, and temporary
4400 errors are given to SMTP connections.
4402 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4403 (except RSET) with a 421 error, until QUIT is received. However, it was
4404 failing to send a response to QUIT.
4406 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4407 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4408 ACL is now run after the helo_try_verify_hosts code.
4410 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4411 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4412 case-independent, so other case variants are also recognized). Apparently
4413 some systems use these upper case variants.
4415 25. If more than two messages were waiting for the same host, and a transport
4416 filter was specified for the transport, Exim sent two messages over the
4417 same TCP/IP connection, and then failed with "socket operation on non-
4418 socket" when it tried to send the third.
4420 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4422 27. The extern definition of crypt16() in expand.c was not being excluded when
4423 the OS had its own crypt16() function.
4425 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4426 as a preferred synonym for return_size_limit, both as an option and as an
4429 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4431 30. Changed the default configuration ACL to relax the local part checking rule
4432 for addresses that are not in any local domains. For these addresses,
4433 slashes and pipe symbols are allowed within local parts, but the sequence
4434 /../ is explicitly forbidden.
4436 31. SPA server authentication was not clearing the challenge buffer before
4439 32. log_message in a "warn" ACL statement was writing to the reject log as
4440 well as to the main log, which contradicts the documentation and doesn't
4441 seem right (because no rejection is happening). So I have stopped it.
4443 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4444 However, I am unable to do any testing of this.
4446 34. Fixed an infelicity in the appendfile transport. When checking directories
4447 for a mailbox, to see if any needed to be created, it was accidentally
4448 using path names with one or more superfluous leading slashes; tracing
4449 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4451 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4452 contents are added to the log line that is written for every discarded
4453 recipient. (Previously a log_message setting was ignored.)
4455 36. The ${quote: operator now quotes the string if it is empty.
4457 37. The install script runs exim in order to find its version number. If for
4458 some reason other than non-existence or emptiness, which it checks, it
4459 could not run './exim', it was installing it with an empty version number,
4460 i.e. as "exim-". This error state is now caught, and the installation is
4463 38. An argument was missing from the function that creates an error message
4464 when Exim fails to connect to the socket for saslauthd authentication.
4465 This could cause Exim to crash, or give a corrupted message.
4467 39. Added isip, isip4, and isip6 to ${if conditions.
4469 40. The ACL variables $acl_xx are now saved with the message, and can be
4470 accessed later in routers, transports, and filters.
4472 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4473 strings in the file are not string-expanded.
4475 42. If a MAIL command specified a SIZE value that was too large to fit into an
4476 int variable, the check against message_size_limit failed. Such values are
4477 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4478 day this will have to be increased, but I don't think I want to be around
4479 when emails are that large.
4486 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4487 iconv() is not standard in FreeBSD.
4489 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4490 IPv6 enabled. The observed symptom was a segmentation fault on return from
4491 the function os_common_find_running_interfaces() in src/os.c.
4493 3. In the check_special_case() function in daemon.c I had used "errno" as an
4494 argument name, which causes warnings on some systems. This was basically a
4495 typo, since it was named "eno" in the comments!
4497 4. The code that waits for the clock to tick (at a resolution of some fraction
4498 of a second) so as to ensure message-id uniqueness was always waiting for
4499 at least one whole tick, when it could have waited for less. [This is
4500 almost certainly not relevant at current processor speeds, where it is
4501 unlikely to ever wait at all. But we try to future-proof.]
4503 5. The function that sleeps for a time interval that includes fractions of a
4504 second contained a race. It did not block SIGALRM between setting the
4505 timer, and suspending (a couple of lines later). If the interval was short
4506 and the sigsuspend() was delayed until after it had expired, the suspension
4507 never ended. On busy systems this could lead to processes getting stuck for
4510 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4511 process, before it forks any delivery processes. The open lookup caching
4512 mechanism meant that the open file or database connection was passed into
4513 the delivery process. The problem was that delivery processes always tidy
4514 up cached lookup data. This could cause a problem for the next delivery
4515 process started by the queue runner, because the external queue runner
4516 process does not know about the closure. So the next delivery process
4517 still has data in the lookup cache. In the case of a file lookup, there was
4518 no problem because closing a file descriptor in a subprocess doesn't affect
4519 the parent. However, if the lookup was caching a connection to a database,
4520 the connection was closed, and the second delivery process was likely to
4521 see errors such as "PGSQL: query failed: server closed the connection
4522 unexpectedly". The problem has been fixed by closing all cached lookups
4523 in a queue runner before running a delivery process.
4525 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4526 seem to have the "const" qualifier which it has on other OS. I've
4529 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4530 *of the same type* (client or server) with the same public name that an
4531 error should be diagnosed.
4533 9. When Exim looked up a host name for an IP address, but failed to find the
4534 original IP address when looking up the host name (a safety check), it
4535 output the message "<ip address> does not match any IP for NULL", which was
4536 confusing, to say the least. The bug was that the host name should have
4537 appeared instead of "NULL".
4539 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4540 user that is built into the binary, and the -C or -D options is used, root
4541 privilege is dropped before the configuration file is read. In addition,
4542 logging is switched to stderr instead of the normal log files. If the
4543 configuration then re-defines the Exim user, the unprivileged environment
4544 is probably not what is expected, so Exim logs a panic warning message (but
4547 However, if deliver_drop_privilege is set, the unprivileged state may well
4548 be exactly what is intended, so the warning has been cut out in that case,
4549 and Exim is allowed to try to write to its normal log files.
4555 1. smtp_return_error_details was not giving details for temporary sender
4556 or receiver verification errors.
4558 2. Diagnose a configuration error if two authenticators have the same public
4561 3. Exim used not to create the message log file for a message until the first
4562 delivery attempt. This could be confusing when incoming messages were held
4563 for policy or load reasons. The message log file is now created at the time
4564 the message is received, and an initial "Received" line is written to it.
4566 4. The automatically generated man page for command line options had a minor
4567 bug that caused no ill effects; however, a more serious problem was that
4568 the procedure for building the man page automatically didn't always
4569 operate. Consequently, release 4.20 contains an out-of-date version. This
4570 shouldn't happen again.
4572 5. When building Exim with embedded Perl support, the script that builds the
4573 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4574 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4576 6. The freeze_tell option was not being used for messages that were frozen on
4577 arrival, either by an ACL or by local_scan().
4579 7. Added the smtp_incomplete_transaction log selector.
4581 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4582 was accepting AUTH without a new EHLO.
4584 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4585 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4588 10. Logging of TCP/IP connections (when configured) now happens in the main
4589 daemon process instead of the child process, so that the TCP/IP connection
4590 count is more accurate (but it can never be perfect).
4592 11. The use of "drop" in a nested ACL was not being handled correctly in the
4593 outer ACL. Now, if condition failure induced by the nested "drop" causes
4594 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4595 or "require"), the connection is dropped.
4597 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4598 that yield "discard" can now be used with an "accept" or a "discard" verb,
4599 but an error is generated for any others (because I can't see a useful way
4600 to define what should happen).
4602 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4603 are now processed in the same way as lines in the Exim configuration file.
4604 In particular, continuation lines are supported.
4606 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4608 15. Added -ti meaning -t -i.
4610 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4611 domains, and warn by logging if others are found.
4613 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4614 returned from GIFCONF when Exim is trying to find the list of interfaces on
4615 a host. The code in os.c has been modified to copy each ifreq to an aligned
4616 structure in all cases.
4618 Also, in some cases, the returned ifreq's were being copied to a 'struct
4619 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4620 means the last couple of bytes of an IPv6 address could be chopped if the
4621 ifreq contained only a normal sockaddr (14 bytes storage).
4623 18. Named domain lists were not supported in the hosts_treat_as_local option.
4624 An entry such as +xxxx was not recognized, and was treated as a literal
4627 19. Ensure that header lines added by a DATA ACL are included in the reject log
4628 if the ACL subsequently rejects the message.
4630 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4631 MD5 (which is deprecated).
4633 21. When testing a filter file using -bf, Exim was writing a message when it
4634 took the sender from a "From " line in the message, but it was not doing so
4635 when it took $return_path from a Return-Path: header line. It now does.
4637 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4638 with a valid header line field name (a series of printing characters
4639 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4641 23. Changed "disc" in the source to "disk" to conform to the documentation and
4642 the book and for uniformity.
4644 24. Ignore Sendmail's -Ooption=value command line item.
4646 25. When execve() failed while trying to run a command in a pipe transport,
4647 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4648 could be confused with a return value of 69 from the command itself. This
4649 has been changed to 127, the value the shell returns if it is asked to run
4650 a non-existent command. The wording for the related log line suggests a
4651 non-existent command as the problem.
4653 26. If received_header_text expands to an empty string, do not add a Received:
4654 header line to the message. (Well, it adds a token one on the spool, but
4655 marks it "old" so that it doesn't get used or transmitted.)
4657 27. Installed eximstats 1.28 (addition of -nt option).
4659 28. There was no check for failure on the call to getsockname() in the daemon
4660 code. This can fail if there is a shortage of resources on the system, with
4661 ENOMEM, for example. A temporary error is now given on failure.
4663 29. Contrary to the C standard, it seems that in some environments, the
4664 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4665 program. Exim now does this explicitly; it affects the formatting of
4666 timestamps using strftime().
4668 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4669 complaints. I've now initialized all the variables, to avoid this.
4671 32. Header lines added by a system filter were not being "seen" during
4672 transport-time rewrites.
4674 33. The info_callback() function passed to OpenSSL is set up with type void
4675 (*)(SSL *, int, int), as described somewhere. However, when calling the
4676 function (actually a macro) that sets it up, the type void(*)() is
4677 expected. I've put in a cast to prevent warnings from picky compilers.
4679 34. If a DNS black list lookup found a CNAME record, but there were no A
4680 records associated with the domain it pointed at, Exim crashed.
4682 35. If a DNS black list lookup returned more than one A record, Exim ignored
4683 all but the first. It now scans all returned addresses if a particular IP
4684 value is being sought. In this situation, the contents of the
4685 $dnslist_value variable are a list of all the addresses, separated by a
4688 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4689 to accept a host name and all its aliases if the forward lookup for any of
4690 them yielded the IP address of the incoming connection. Now it accepts only
4691 those names whose forward lookup yields the correct IP address. Any other
4692 names are discarded. This closes a loophole whereby a rogue DNS
4693 administrator could create reverse DNS records to break through a
4694 wildcarded host restriction in an ACL.
4696 37. If a user filter or a system filter that ran in a subprocess used any of
4697 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4698 the wrong values were passed to the pipe command ($thisaddress had the
4699 value of $0, $0 had the value of $1, etc). This bug was introduced by
4700 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4702 38. Improved the line breaking for long SMTP error messages from ACLs.
4703 Previously, if there was no break point between 40 and 75 characters, Exim
4704 left the rest of the message alone. Two changes have been made: (a) I've
4705 reduced the minimum length to 35 characters; (b) if it can't find a break
4706 point between 35 and 75 characters, it looks ahead and uses the first one
4707 that it finds. This may give the occasional overlong line, but at least the
4708 remaining text gets split now.
4710 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4711 file descriptors might be low, and that setting 1000 would always raise it.
4712 It turns out that in some environments, the limit is already over 1000 and
4713 that lowering it causes trouble. So now Exim takes care not to decrease it.
4715 40. When delivering a message, the value of $return_path is set to $sender_
4716 address at the start of routing (routers may change the value). By an
4717 oversight, this default was not being set up when an address was tested by
4718 -bt or -bv, which affected the outcome if any router or filter referred to
4721 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4722 log only when "message" or "log_message" are set. However, if one of the
4723 conditions was an address verification, or a call to a nested ACL, the
4724 messages generated by the underlying test were being passed through. This
4725 no longer happens. The underlying message is available in $acl_verify_
4726 message for both "message" and "log_message" expansions, so it can be
4727 passed through if needed.
4729 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4730 new expansion $bh_ to give the encoded byte string without charset
4731 translation. Translation happens only if iconv() is available; HAVE_ICONV
4732 indicates this at build time. HEADERS_CHARSET gives the charset to
4733 translate to; headers_charset can change it in the configuration, and
4734 "headers charset" can change it in an individual filter file.
4736 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4737 that creates RFC 2047 encoded "words" labels them as that charset instead
4738 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4739 expansion operator; (ii) when Exim creates a From: line for a local
4740 message; (iii) when a header line is rewritten to include a "phrase" part.
4742 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4743 buggy, causing it to skip the first lines of messages whose message ID
4744 ended in 'D'. This would not have bitten before Exim release 4.14, because
4745 message IDs were unlikely to end in 'D' before then. The effect was to have
4746 incorrect size information for certain domains.
4748 45. #include "config.h" was missing at the start of the crypt16.c module. This
4749 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4752 46. If there was a timeout during a "random" callout check, Exim treated it as
4753 a failure of the random address, and carried on sending RSET and the real
4754 address. If the delay was just some slowness somewhere, the response to the
4755 original RCPT would be taken as a response to RSET and so on, causing
4756 mayhem of various kinds.
4758 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4759 when I implemented it. It didn't allow for the fact that some option values
4760 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4761 the right test for positive values.
4763 48. Domain names in DNS records are case-independent. Exim always looks them up
4764 in lower case. Some resolvers return domain names in exactly the case they
4765 appear in the zone file, that is, they may contain uppercase letters. Not
4766 all resolvers do this - some return always lower case. Exim was treating a
4767 change of case by a resolver as a change of domain, similar to a widening
4768 of a domain abbreviation. This triggered its re-routing code and so it was
4769 trying to route what was effectively the same domain again. This normally
4770 caused routing to fail (because the router wouldn't handle the domain
4771 twice). Now Exim checks for this case specially, and just changes the
4772 casing of the domain that it ultimately uses when it transmits the message
4775 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4778 50. If a filter generated a file delivery with a non-absolute name (possible if
4779 no home directory exists for the router), the forbid_file option was not
4782 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4783 the existing equality matching.
4785 52. Exim was using ints instead of ino_t variables in some places where it was
4786 dealing with inode numbers.
4788 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4789 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4790 called TMPDIR, and if it finds it is different, it changes its value.
4792 54. The smtp_printf() function is now made available to local_scan() so
4793 additional output lines can be written before returning. There is also an
4794 smtp_fflush() function to enable the detection of a dropped connection.
4795 The variables smtp_input and smtp_batched_input are exported to
4798 55. Changed the default runtime configuration: the message "Unknown user"
4799 has been removed from the ACL, and instead placed on the localuser router,
4800 using the cannot_route_message feature. This means that any verification
4801 failures that generate their own messages won't get overridden. Similarly,
4802 the "Unrouteable address" message that was in the ACL for unverifiable
4803 relay addresses has also been removed.
4805 56. Added hosts_avoid_esmtp to the smtp transport.
4807 57. The exicyclog script was not checking for the esoteric option
4808 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4809 will work only if exicyclog is run under the appropriate euid.
4811 58. Following a discussion on the list, the rules by which Exim recognises line
4812 endings on incoming messages have been changed. The -dropcr and drop_cr
4813 options are now no-ops, retained only for backwards compatibility. The
4814 following line terminators are recognized: LF CRLF CR. However, special
4815 processing applies to CR:
4817 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4818 nor a local message in the state where . is a terminator.
4820 (ii) If a bare CR is encountered in a header line, an extra space is added
4821 after the line terminator so as not to end the header. The reasoning
4822 behind this is that bare CRs in header lines are most likely either
4823 to be mistakes, or people trying to play silly games.
4825 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4826 was being incorrectly given as 18 bytes larger than it should have been.
4827 This is a VOB (very old bug).
4829 60. This may never have affected anything current, but just in case it has:
4830 When the local host is found other than at the start of a list of hosts,
4831 the local host, those with the same MX, and any that follow, are discarded.
4832 When the list in question was part of a longer list of hosts, the following
4833 hosts (not currently being processed) were also being discarded. This no
4834 longer happens. I'm not sure if this situation could ever has previously
4837 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4840 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4841 header. This is something that is recommended in a new Internet Draft, and
4842 is something that is documented as being done by Sendmail. There are two
4843 possible values. For messages generated by the autoreply transport, Exim
4846 Auto-Submitted: auto-replied
4848 whereas for all other generated messages (e.g. bounces) it adds
4850 Auto-Submitted: auto-generated
4852 63. The "personal" condition in filters now includes a test for the
4853 Auto-Submitted: header. If it contains the string "auto-" the message it
4854 not considered personal.
4856 64. Added rcpt_include_affixes as a generic transport option.
4858 65. Added queue_only_override (default true).
4860 66. Added the syslog_duplication option.
4862 67. If what should have been the first header line of a message consisted of
4863 a space followed by a colon, Exim was mis-interpreting it as a header line.
4864 It isn't of course - it is syntactically invalid and should therefore be
4865 treated as the start of the message body. The misbehaviour could have
4866 caused a number of strange effects, including loss of data in subsequent
4867 header lines, and spool format errors.
4869 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4870 client host had authenticated. This control can now be exercised by an ACL
4871 for more flexibility.
4873 69. By default, callouts do not happen when testing with -bh. There is now a
4874 variant, -bhc, which does actually run the callout code, including
4875 consulting and updating the callout cache.
4877 70. Added support for saslauthd authentication, courtesy of Alexander
4880 71. If statvfs() failed on the spool or log directories while checking their
4881 size for availability, Exim confusingly gave the error "space shortage".
4882 Furthermore, in debugging mode it crashed with a floating point exception.
4883 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4884 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4885 serious problem, Exim now writes to the main and panic logs when this
4886 happens, with details of the failure. It then refuses to accept the
4887 incoming message, giving the message "spool directory problem" or "log
4888 directory problem" with a 421 code for SMTP messages.
4890 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4891 0, 1, and 2 exist, because some OS complain for execs without them (see
4892 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4893 descriptors. However, the code omitted to check that the open succeeded,
4894 causing mysterious errors if for some reason the permissions on /dev/null
4895 got screwed. Now Exim writes a message to the main and panic logs, and
4896 bombs out if it can't open /dev/null.
4898 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4899 interact so that it is all more flexible. It is supposed to remain
4900 backwards compatible. Also added extra_local_interfaces.
4902 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4903 to bomb out with an assertion failure - to the client this appears as a
4904 connection drop. This problem occurs in the part of the code that was taken
4905 from the Samba project. Fortunately, the assertion is in a very simple
4906 function, so I have fixed this by reproducing the function inline in the
4907 one place where it is called, and arranging for authentication to fail
4908 instead of killing the process with assert().
4910 75. The SPA client code was not working when the server requested OEM rather
4911 than Unicode encoding.
4913 76. Added code to make require_files with a specific uid setting more usable in
4914 the case where statting the file as root fails - usually a non-root-mounted
4915 NFS file system. When this happens and the failure is EACCES, Exim now
4916 forks a subprocess and does the per-uid checking as the relevant uid.
4918 77. Added process_log_path.
4920 78. If log_file_path was not explicitly set, a setting of check_log_space or
4921 check_log_inodes was ignored.
4923 79. If a space check for the spool or log partitions fails, the incident is now
4924 logged. Of course, in the latter case the data may get lost...
4926 80. Added the %p formatting code to string_format() so that it can be used to
4927 print addresses in debug_print(). Adjusted all the address printing in the
4928 debugging in store.c to use %p rather than %d.
4930 81. There was a concern that a line of code in smtp_in.c could overflow a
4931 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4932 initially expressed, the concern was not well-founded, because trailing
4933 spaces are removed early. However, if the trailing spaces were followed by
4934 a NULL, they did not get removed, so the overflow was possible. Two fixes
4937 (a) I re-wrote the offending code in a cleaner fashion.
4938 (b) If an incoming SMTP command contains a NULL character, it is rejected
4941 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4942 runs initgroups(), so that if the Exim user is in any additional groups,
4943 they will be used during message reception.
4949 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob