1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.564 2009/07/02 11:38:56 nm4 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TK/01 Added patch by Johannes Berg that expands the main option
10 "spamd_servers" if it starts with a dollar sign.
12 TK/02 Write list of recipients to X-Envelope-Sender header when building
13 the mbox-format spool file for content scanning (suggested by Jakob
16 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
17 (http://www.dcc-servers.net/) support via dccifd. Activated by
18 setting EXPERIMENTAL_DCC=yes in Local/Makefile. Check out
19 experimental_spec.txt for more documentation.
21 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
22 by Mark Daniel Reidel <mr@df.eu>.
24 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
25 When building exim an external PCRE library is now needed -
26 PCRE is a system library on the majority of modern systems.
27 See entry on PCRE_LIBS in EDITME file.
29 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
30 conversation. Added nologin parameter to request.
31 Patch contributed by Kirill Miazine
33 TF/01 Do not log submission mode rewrites if they do not change the address.
35 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
37 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
38 log files in place. Contributed by Roberto Lima
40 NM/04 Bugzilla 667: close socket used by dovecot authenticator
42 TF/03 Bugzilla 615: When checking the local_parts router precondition
43 after a local_part_suffix or local_part_prefix option, Exim now
44 does not use the address's named list lookup cache, since this
45 contains cached lookups for the whole local part.
47 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
48 Robert Millan. Documentation is in experimental-spec.txt
50 TF/04 Bugzilla 668: Fix parallel build (make -j).
52 NM/05 Bugzilla 437: Prevent Maildix aux files being created with mode 000
54 NM/05 Bugzilla 598: Improvement to Dovecot authenticator handling.
55 Patch provided by Jan Srzednicki
57 TF/05 Leading white space used to be stripped from $spam_report which
58 wrecked the formatting. Now it is preserved.
60 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
61 that they are available at delivery time.
63 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
65 TF/08 TLS error reporting now respects the incoming_interface and
66 incoming_port log selectors.
68 TF/09 Produce a more useful error message if an SMTP transport's hosts
69 setting expands to an empty string.
71 NM/06 Bugzilla 744: EXPN did not work under TLS.
72 Patch provided by Phil Pennock
74 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
75 Patch provided by Richard Godbee
77 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
78 acl_smtp_notquit, added index entry.
80 NM/09 Bugzilla 787: Potential buffer overflow in string_format
81 Patch provided by Eugene Bujak
83 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to accept()
84 Patch provided by Maxim Dounin
86 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
87 Patch provided by Phil Pennock
89 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
91 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
92 Patch provided by Brad "anomie" Jorsch
94 NM/14 Bugzilla 614: Added timeout setting to transport filter.
95 Patch provided by Dean Brooks
97 TK/05 Add native DKIM support (does not depend on external libraries).
99 NM/14 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
100 Patch provided by Graeme Fowler
106 TK/01 Add preliminary DKIM support. Currently requires a forked version of
107 ALT-N's libdkim that I have put here:
108 http://duncanthrax.net/exim-experimental/
110 Note to Michael Haardt: I had to rename some vars in sieve.c. They
111 were called 'true' and it seems that C99 defines that as a reserved
112 keyword to be used with 'bool' variable types. That means you could
113 not include C99-style headers which use bools without triggering
114 build errors in sieve.c.
116 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
117 as mailq or other aliases. Changed the --help handling significantly
118 to do whats expected. exim_usage() emits usage/help information.
120 SC/01 Added the -bylocaldomain option to eximstats.
122 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
124 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
126 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
132 PH/01 Another patch from the Sieve maintainer.
134 PH/02 When an IPv6 address is converted to a string for single-key lookup
135 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
136 dots are used instead of colons so that keys in lsearch files need not
137 contain colons. This was done some time before quoting was made available
138 in lsearch files. However, iplsearch files do require colons in IPv6 keys
139 (notated using the quote facility) so as to distinguish them from IPv4
140 keys. This meant that lookups for IP addresses in host lists did not work
141 for iplsearch lookups.
143 This has been fixed by arranging for IPv6 addresses to be expressed with
144 colons if the lookup type is iplsearch. This is not incompatible, because
145 previously such lookups could never work.
147 The situation is now rather anomolous, since one *can* have colons in
148 ordinary lsearch keys. However, making the change in all cases is
149 incompatible and would probably break a number of configurations.
151 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
154 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
155 conversion specification without a maximum field width, thereby enabling
156 a rogue spamd server to cause a buffer overflow. While nobody in their
157 right mind would setup Exim to query an untrusted spamd server, an
158 attacker that gains access to a server running spamd could potentially
159 exploit this vulnerability to run arbitrary code as the Exim user.
161 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
162 $primary_hostname instead of what libspf2 thinks the hosts name is.
164 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
165 a directory entry by the name of the lookup key. Previously, if a
166 symlink pointed to a non-existing file or a file in a directory that
167 Exim lacked permissions to read, a lookup for a key matching that
168 symlink would fail. Now it is enough that a matching directory entry
169 exists, symlink or not. (Bugzilla 503.)
171 PH/03 The body_linecount and body_zerocount variables are now exported in the
174 PH/04 Added the $dnslist_matched variable.
176 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
177 This means they are set thereafter only if the connection becomes
180 PH/06 Added the client_condition to authenticators so that some can be skipped
181 by clients under certain conditions.
183 PH/07 The error message for a badly-placed control=no_multiline_responses left
184 "_responses" off the end of the name.
186 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
188 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
189 (without spaces) instead of just copying the configuration text.
191 PH/10 Added the /noupdate option to the ratelimit ACL condition.
193 PH/11 Added $max_received_linelength.
195 PH/12 Added +ignore_defer and +include_defer to host lists.
197 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
198 way in which PCRE > 7.0 is built.
200 PH/14 Implemented queue_only_load_latch.
202 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
203 MAIL command. The effect was to mangle the value on 64-bit systems.
205 PH/16 Another patch from the Sieve maintainer.
207 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
209 PH/18 If a system quota error occurred while trying to create the file for
210 a maildir delivery, the message "Mailbox is full" was not appended to the
211 bounce if the delivery eventually timed out. Change 4.67/27 below applied
212 only to a quota excession during the actual writing of the file.
214 PH/19 It seems that peer DN values may contain newlines (and other non-printing
215 characters?) which causes problems in log lines. The DN values are now
216 passed through string_printing() before being added to log lines.
218 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
219 and InterBase are left for another time.)
221 PH/21 Added message_body_newlines option.
223 PH/22 Guard against possible overflow in moan_check_errorcopy().
225 PH/23 POSIX allows open() to be a macro; guard against that.
227 PH/24 If the recipient of an error message contained an @ in the local part
228 (suitably quoted, of course), incorrect values were put in $domain and
229 $local_part during the evaluation of errors_copy.
235 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
236 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
239 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
240 issue a MAIL command.
242 PH/02 In an ACL statement such as
244 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
246 if a client was not listed at all, or was listed with a value other than
247 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
248 the condition was not true (as it should be), so access was not denied.
249 The bug was that the ! inversion was incorrectly passed on to the second
250 item. This has been fixed.
252 PH/03 Added additional dnslists conditions == and =& which are different from
253 = and & when the dns lookup returns more than one IP address.
255 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
256 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
258 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
259 FSYNC, which compiles an option called disable_fsync that allows for
260 bypassing fsync(). The documentation is heavily laced with warnings.
262 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
264 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
265 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
266 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
267 including adding "make clean"; (3) Added -fPIC when compiling the test
268 dynamically loaded module, to get rid of a warning.
270 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
271 message fails, move_frozen_messages = true and ignore_bounce_errors_after
272 = 0s. The bug is otherwise harmless.
274 PH/07 There was a bug in the dovecot authenticator such that the value of
275 $auth1 could be overwritten, and so not correctly preserved, after a
276 successful authentication. This usually meant that the value preserved by
277 the server_setid option was incorrect.
279 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
281 PH/09 Installed PCRE release 7.0.
283 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
284 run for batched SMTP input. It is now run at the start of every message
285 in the batch. While fixing this I discovered that the process information
286 (output by running exiwhat) was not always getting set for -bs and -bS
287 input. This is fixed, and it now also says "batched" for BSMTP.
289 PH/11 Added control=no_pipelining.
291 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
292 patch, slightly modified), and move the expansion of helo_data till after
293 the connection is made in the smtp transport (so it can use these
296 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
298 PH/14 Added log_selector = +pid.
300 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
302 PH/16 Add ${if forany and ${if forall.
304 PH/17 Added dsn_from option to vary the From: line in DSNs.
306 PH/18 Flush SMTP output before performing a callout, unless control =
307 no_callout_flush is set.
309 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
310 was true (the default) a successful delivery failed to delete the retry
311 item, thus causing premature timeout of the address. The bug is now
314 PH/20 Added hosts_avoid_pipelining to the smtp transport.
316 PH/21 Long custom messages for fakedefer and fakereject are now split up
317 into multiline reponses in the same way that messages for "deny" and
318 other ACL rejections are.
320 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
321 with slight modification.
323 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
324 draft, changing the syntax and factoring some duplicate code".
326 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
327 for deliveries of the second and subsequent messages over the same SMTP
330 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
331 ${reduce, with only minor "tidies".
333 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
335 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
336 expansion side effects.
338 PH/27 When a message times out after an over-quota error from an Exim-imposed
339 quota, the bounce message says "mailbox is full". This message was not
340 being given when it was a system quota that was exceeded. It now should
343 MH/03 Made $recipients available in local_scan(). local_scan() already has
344 better access to the recipient list through recipients_list[], but
345 $recipients can be useful in postmaster-provided expansion strings.
347 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
348 in the case of a MAIL command with additional options following the
349 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
350 were accidentally chopped off.
352 PH/29 SMTP synchronization checks are implemented when a command is read -
353 there is a check that no more input is waiting when there shouldn't be
354 any. However, for some commands, a delay in an ACL can mean that it is
355 some time before the response is written. In this time, more input might
356 arrive, invalidly. So now there are extra checks after an ACL has run for
357 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
358 pipelining has not been advertised.
360 PH/30 MH's patch to allow iscntrl() characters to be list separators.
362 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
363 returned in the SMTP response when smtp_return_error_details was false.
366 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
367 instead of the C I/O that was originally supplied, because problems were
370 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
371 Exim which did not show up earlier: it was assuming that a call to
372 SSL_CTX_set_info_callback() might give an error value. In fact, there is
373 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
374 was a macro that became an assignment, so it seemed to work. This has
375 changed to a proper function call with a void return, hence the compile
376 error. Exim's code has been fixed.
378 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
381 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
383 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
385 PH/37 If a message is not accepted after it has had an id assigned (e.g.
386 because it turns out to be too big or there is a timeout) there is no
387 "Completed" line in the log. When some messages of this type were
388 selected by exigrep, they were listed as "not completed". Others were
389 picked up by some special patterns. I have improved the selection
390 criteria to be more general.
392 PH/38 The host_find_failed option in the manualroute router can now be set
393 to "ignore", to completely ignore a host whose IP address cannot be
394 found. If all hosts are ignored, the behaviour is controlled by the new
395 host_all_ignored option.
397 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
398 homing or because of multiple MX records with /mx) generated more than
399 one IP address, and the following item turned out to be the local host,
400 all the secondary addresses of the first item were incorrectly removed
401 from the list, along with the local host and any following hosts (which
402 is what is supposed to happen).
404 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
405 whoever called Exim into the -H file. In the case of the daemon it was
406 behaving confusingly. When first started, it used values for whoever
407 started the daemon, but after a SIGHUP it used the Exim user (because it
408 calls itself on a restart). I have changed the code so that it now always
411 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
412 message are rejected with the same error (e.g. no authentication or bad
413 sender address), and a DATA command is nevertheless sent (as can happen
414 with PIPELINING or a stupid MUA), the error message that was given to the
415 RCPT commands is included in the rejection of the DATA command. This is
416 intended to be helpful for MUAs that show only the final error to their
419 PH/42 Another patch from the Sieve maintainer.
421 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
422 Eximstats - Fixed some broken HTML links and added missing column headers
424 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
425 columns for Rejects, Temp Rejects, Ham, and Spam rows.
427 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
429 PH/43 Yet another patch from the Sieve maintainer.
431 PH/44 I found a way to check for a TCP/IP connection going away before sending
432 the response to the final '.' that terminates a message, but only in the
433 case where the client has not sent further data following the '.'
434 (unfortunately, this is allowed). However, in many cases there won't be
435 any further data because there won't be any more messages to send. A call
436 to select() can be used: if it shows that the input is "ready", there is
437 either input waiting, or the socket has been closed. An attempt to read
438 the next input character can distinguish the two cases. Previously, Exim
439 would have sent an OK response which the client would never have see.
440 This could lead to message repetition. This fix should cure that, at
441 least in a lot of common cases.
443 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
444 advertised in response to EHLO.
450 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
451 fixed by 4.65/MH/01 (is this a record?) are fixed:
453 (i) An empty string was always treated as zero by the numeric comparison
454 operators. This behaviour has been restored.
456 (ii) It is documented that the numeric comparison operators always treat
457 their arguments as decimal numbers. This was broken in that numbers
458 starting with 0 were being interpreted as octal.
460 While fixing these problems I realized that there was another issue that
461 hadn't been noticed. Values of message_size_limit (both the global option
462 and the transport option) were treated as octal if they started with 0.
463 The documentation was vague. These values are now always treated as
464 decimal, and I will make that clear in the documentation.
470 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
471 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
474 MH/01 Don't check that the operands of numeric comparison operators are
475 integers when their expansion is in "skipping" mode (fixes bug
476 introduced by 4.64-PH/07).
478 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
479 child addresses, Exim now panics and dies. Previously, because the count
480 is held in a short int, deliveries were likely to be lost. As such a
481 large number of recipients for a single message is ridiculous
482 (performance will be very, very poor), I have chosen to impose a limit
483 rather than extend the field.
489 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
490 leftover -K file (the existence of which was triggered by #402).
491 While we were at it, introduced process PID as part of the -K
492 filename. This should rule out race conditions when creating
495 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
496 processing considerably. Previous code took too long for large mails,
497 triggering a timeout which in turn triggers #401.
499 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
500 in the DK code in transports.c. sendfile() is not really portable,
501 hence the _LINUX specificness.
503 TF/01 In the add_headers option to the mail command in an Exim filter,
504 there was a bug that Exim would claim a syntax error in any
505 header after the first one which had an odd number of characters
508 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
509 callout verification, Exim cached a "reject" for the entire domain. This
510 is correct for most verifications, but it is not correct for a recipient
511 verification with use_sender or use_postmaster set, because in that case
512 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
513 case of MAIL FROM:<> rejection from other early rejections (e.g.
514 rejection of HELO). When verifying a recipient using a non-null MAIL
515 address, the cache is ignored if it shows MAIL FROM:<> rejection.
516 Whatever the result of the callout, the value of the domain cache is
517 left unchanged (for any other kind of callout, getting as far as trying
518 RCPT means that the domain itself is ok).
520 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
523 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
524 manifest itself as EPIPE rather than ECONNECT. When tidying away a
525 session, the daemon ignores ECONNECT errors and logs others; it now
526 ignores EPIPE as well.
528 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
529 (quoted-printable decoding).
531 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
532 later the small subsequent patch to fix an introduced bug.
534 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
536 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
538 PH/08 An error is now given if message_size_limit is specified negative.
540 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
541 to be given (somewhat) arbitrary names.
543 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
546 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
547 miscellaneous code fixes
549 PH/10 Added the log_reject_target ACL modifier to specify where to log
552 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
553 hostname. This is wrong, because it relates to the incoming message (and
554 probably the interface on which it is arriving) and not to the outgoing
555 callout (which could be using a different interface). This has been
556 changed to use the value of the helo_data option from the smtp transport
557 instead - this is what is used when a message is actually being sent. If
558 there is no remote transport (possible with a router that sets up host
559 addresses), $smtp_active_hostname is used.
561 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
562 tweaks were necessary in order to get it to work (see also 21 below):
563 (a) The code assumed that strncpy() returns a negative number on buffer
564 overflow, which isn't the case. Replaced with Exim's string_format()
566 (b) There were several signed/unsigned issues. I just did the minimum
567 hacking in of casts. There is scope for a larger refactoring.
568 (c) The code used strcasecmp() which is not a standard C function.
569 Replaced with Exim's strcmpic() function.
570 (d) The code set only $1; it now sets $auth1 as well.
571 (e) A simple test gave the error "authentication client didn't specify
572 service in request". It would seem that Dovecot has changed its
573 interface. Fortunately there's a specification; I followed it and
574 changed what the client sends and it appears to be working now.
576 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
579 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
580 address A is aliased to B and C, where B exists and C does not. Without
581 -v the output is "A verified" because verification stops after a
582 successful redirection if more than one address is generated. However,
583 with -v the child addresses are also verified. Exim was outputting "A
584 failed to verify" and then showing the successful verification for C,
585 with its parentage. It now outputs "B failed to verify", showing B's
586 parentage before showing the successful verification of C.
588 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
589 look up a TXT record in a specific list after matching in a combined
592 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
593 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
594 they consult the DNS. I had assumed they would set it the way they
595 wanted; and indeed my experiments on Linux seem to show that in some
596 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
597 To be on the safe side, however, I have now made the interface to
598 host_find_byname() similar to host_find_bydns(), with an argument
599 containing the DNS resolver options. The host_find_byname() function now
600 sets these options at its start, just as host_find_bydns() does. The smtp
601 transport options dns_qualify_single and dns_search_parents are passed to
602 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
603 of host_find_byname() use the default settings of RES_DEFNAMES
604 (qualify_single) but not RES_DNSRCH (search_parents).
606 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
607 spool_read_header() do less string testing, by means of a preliminary
608 switch on the second character of optional "-foo" lines. (This is
609 overdue, caused by the large number of possibilities that now exist.
610 Originally there were few.) While I was there, I also converted the
611 str(n)cmp tests so they don't re-test the leading "-" and the first
612 character, in the hope this might squeeze out yet more improvement.
614 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
615 flag allowing group syntax was set by the header_syntax check but not
616 turned off, possible causing trouble later; (2) The flag was not being
617 set at all for the header_verify test, causing "group"-style headers to
618 be rejected. I have now set it in this case, and also caused header_
619 verify to ignore an empty address taken from a group. While doing this, I
620 came across some other cases where the code for allowing group syntax
621 while scanning a header line wasn't quite right (mostly, not resetting
622 the flag correctly in the right place). These bugs could have caused
623 trouble for malformed header lines. I hope it is now all correct.
625 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
626 with the "reply" argument non-NULL. The code, however (which originally
627 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
628 but it didn't always do it. This confused somebody who was copying the
629 code for some other use. I have removed all the tests.
631 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
632 feature that was used to support insecure browsers during the U.S. crypto
633 embargo. It requires special client support, and Exim is probably the
634 only MTA that supported it -- and would never use it because real RSA is
635 always available. This code has been removed, because it had the bad
636 effect of slowing Exim down by computing (never used) parameters for the
637 RSA_EXPORT functionality.
639 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
640 authenticator to fail if there's a tab character in the incoming data
641 (there should never be unless someone is messing about, as it's supposed
642 to be base64-encoded). Also added, on Timo's advice, the "secured" option
643 if the connection is using TLS or if the remote IP is the same as the
644 local IP, and the "valid-client-cert option" if a client certificate has
647 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
648 authenticators. This can be used for authorization after authentication
649 succeeds. (In the case of plaintext, it servers for both authentication
652 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
653 if any retry times were supplied.
655 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
656 connection, where there is no client IP address to check. In this
657 situation, the verify now always succeeds.
659 PH/25 Applied John Jetmore's -Mset patch.
661 PH/26 Added -bem to be like -Mset, but loading a message from a file.
663 PH/27 In a string expansion for a processed (not raw) header when multiple
664 headers of the same name were present, leading whitespace was being
665 removed from all of them, but trailing whitespace was being removed only
666 from the last one. Now trailing whitespace is removed from each header
667 before concatenation. Completely empty headers in a concatenation (as
670 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
671 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
673 PH/29 [Removed. This was a change that I later backed out, and forgot to
674 correct the ChangeLog entry (that I had efficiently created) before
675 committing the later change.]
677 PH/30 Exim was sometimes attempting to deliver messages that had suffered
678 address errors (4xx response to RCPT) over the same connection as other
679 messages routed to the same hosts. Such deliveries are always "forced",
680 so retry times are not inspected. This resulted in far too many retries
681 for the affected addresses. The effect occurred only when there were more
682 hosts than the hosts_max_try setting in the smtp transport when it had
683 the 4xx errors. Those hosts that it had tried were not added to the list
684 of hosts for which the message was waiting, so if all were tried, there
685 was no problem. Two fixes have been applied:
687 (i) If there are any address or message errors in an SMTP delivery, none
688 of the hosts (tried or untried) are now added to the list of hosts
689 for which the message is waiting, so the message should not be a
690 candidate for sending over the same connection that was used for a
691 successful delivery of some other message. This seems entirely
692 reasonable: after all the message is NOT "waiting for some host".
693 This is so "obvious" that I'm not sure why it wasn't done
694 previously. Hope I haven't missed anything, but it can't do any
695 harm, as the worst effect is to miss an optimization.
697 (ii) If, despite (i), such a delivery is accidentally attempted, the
698 routing retry time is respected, so at least it doesn't keep
699 hammering the server.
701 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
702 in ${readsocket because some servers need this prod.
704 PH/32 Added some extra debug output when updating a wait-xxx database.
706 PH/33 The hint "could be header name not terminated by colon", which has been
707 given for certain expansion errors for a long time, was not being given
708 for the ${if def:h_colon_omitted{... case.
710 PH/34 The spec says: "With one important exception, whenever a domain list is
711 being scanned, $domain contains the subject domain." There was at least
712 one case where this was not true.
714 PH/35 The error "getsockname() failed: connection reset by peer" was being
715 written to the panic log as well as the main log, but it isn't really
716 panic-worthy as it just means the connection died rather early on. I have
717 removed the panic log writing for the ECONNRESET error when getsockname()
720 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
721 runs only) independently of the message's sender address. This meant
722 that, if the 4xx error was in fact related to the sender, a different
723 message to the same recipient with a different sender could confuse
724 things. In particualar, this can happen when sending to a greylisting
725 server, but other circumstances could also provoke similar problems.
726 I have changed the default so that the retry time for these errors is now
727 based a combination of the sender and recipient addresses. This change
728 can be overridden by setting address_retry_include_sender=false in the
731 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
732 remote server are returned as part of bounce messages. This was not
733 happening for LMTP over a pipe (the lmtp transport), but now it is the
734 same for both kinds of LMTP.
736 PH/38 Despite being documented as not happening, Exim was rewriting addresses
737 in header lines that were in fact CNAMEs. This is no longer the case.
739 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
740 and queue runs started by the daemon processed all messages. This has
741 been fixed so that -R and -S can now usefully be given with -q<time>.
743 PH/40 Import PCRE release 6.7 (fixes some bugs).
745 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
747 PH/42 Give an error if -q is specified more than once.
749 PH/43 Renamed the variables $interface_address and $interface_port as
750 $received_ip_address and $received_port, to make it clear that these
751 values apply to message reception, and not to the outgoing interface when
752 a message is delivered. (The old names remain recognized, of course.)
754 PH/44 There was no timeout on the connect() call when using a Unix domain
755 socket in the ${readsocket expansion. There now is.
757 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
758 be meaningful with "accept".
760 SC/01 Eximstats V1.43
761 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
763 SC/02 Eximstats V1.44
764 Use a glob alias rather than an array ref in the generated
765 parser. This improves both readability and performance.
767 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
768 Collect SpamAssassin and rejection statistics.
769 Don't display local sender or destination tables unless
770 there is data to show.
771 Added average volumes into the top table text output.
773 SC/04 Eximstats V1.46
774 Collect data on the number of addresses (recipients)
775 as well as the number of messages.
777 SC/05 Eximstats V1.47
778 Added 'Message too big' to the list of mail rejection
779 reasons (thanks to Marco Gaiarin).
781 SC/06 Eximstats V1.48
782 Mainlog lines which have GMT offsets and are too short to
783 have a flag are now skipped.
785 SC/07 Eximstats V1.49 (Alain Williams)
786 Added the -emptyok flag.
788 SC/08 Eximstats V1.50
789 Fixes for obtaining the IP address from reject messages.
791 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
792 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
793 whitesspace changes from 4.64-PH/27
795 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
798 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
799 are found, allow negative numbers in numeric criteria)
801 JJ/06 exipick.20061117.2, added new $message_body_missing variable
803 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
804 to match changes made in 4.64-PH/43
806 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
808 PH/47 Put in an explicit test for a DNS lookup of an address record where the
809 "domain" is actually an IP address, and force a failure. This locks out
810 those revolvers/nameservers that support "A-for-A" lookups, in
811 contravention of the specifications.
813 PH/48 When a host name was looked up from an IP address, and the subsequent
814 forward lookup of the name timed out, the host name was left in
815 $sender_host_name, contrary to the specification.
817 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
818 restricted to single-key lookups, Exim was not diagnosing an error if
819 * or *@ was used with a query-style lookup.
821 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
823 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
824 long ago, but noone interested enough thought of it. Let's just say that
825 the "1.1" means that there are some new functions that weren't there at
826 some point in the past.
828 PH/51 Error processing for expansion failure of helo_data from an smtp
829 transport during callout processing was broken.
831 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
832 tested/used via the -bh/-bhc/-bs options.
834 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
835 bug, fixed in subsequent PCRE releases).
837 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
838 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
840 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
846 SC/01 Use a glob alias rather than an array ref in eximstats generated
847 parser. This improves both readability and performance.
849 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
850 Don't display local sender or destination tables in eximstats unless
851 there is data to show.
852 Added average volumes into the eximstats top table text output.
854 SC/03 Collect data on the number of addresses (recipients) as well
855 as the number of messages in eximstats.
857 TF/01 Correct an error in the documentation for the redirect router. Exim
858 does (usually) call initgroups() when daemonizing.
860 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
861 with consistent privilege compared to when running as a daemon.
863 TF/03 Note in the spec that $authenticated_id is not set for local
864 submissions from trusted users.
866 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
867 Thanks to Dean Brooks <dean@iglou.com> for the patch.
869 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
870 by adding some example configuration directives to the default
871 configuration file. A little bit of work is required to uncomment the
872 directives and define how usernames and passwords are checked, but
873 there is now a framework to start from.
875 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
876 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
877 without this. I don't know how relevant this is to other LDAP libraries.
879 PH/02 Add the verb name to the "unknown ACL verb" error.
881 PH/03 Magnus Holmgren's patch for filter_prepend_home.
883 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
885 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
886 directory not expanded when it should be if an expanded home directory
887 was set for the address (which is overridden by the transport).
889 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
892 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
893 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
894 because it is too late at that time, and has no effect.
896 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
897 security issue with \' (bugzilla #107). I could not use the
898 PQescapeStringConn() function, because it needs a PGconn value as one of
901 PH/08 When testing addresses using -bt, indicate those final addresses that
902 are duplicates that would not cause an additional delivery. At least one
903 person was confused, thinking that -bt output corresponded to deliveries.
904 (Suppressing duplicates isn't a good idea as you lose the information
905 about possibly different redirections that led to the duplicates.)
907 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
908 systems where poll() doesn't work, in particular OS X.
910 PH/10 Added more information to debugging output for retry time not reached.
912 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
913 operations in malware.c.
915 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
918 PH/13 If write_rejectlog was set false when logging was sent to syslog with
919 syslog_duplication set false, log lines that would normally be written
920 both the the main log and to the reject log were not written to syslog at
923 PH/14 In the default configuration, change the use of "message" in ACL warn
924 statements to "add_header".
926 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
927 not followed by a command (e.g. "seen endif").
929 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
930 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
933 PH/17 Added extra conditions to the default value of delay_warning_condition
937 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
938 { match{$h_precedence:}{(?i)bulk|list|junk} } \
939 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
942 The Auto-Submitted: and various List- headers are standardised, whereas I
943 don't think Precedence: ever was.
945 PH/18 Refactored debugging code in route_finduser() to show more information,
946 in particular, the error code if getpwnam() issues one.
948 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
949 This is apparently needed in addition to the PH/07 change above to avoid
950 any possible encoding problems.
952 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
953 but not after initializing Perl.
955 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
956 output them only if debugging. By default they are written stderr,
957 apparently, which is not desirable.
959 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
962 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
965 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
967 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
968 authenticated or an ident call has been made. Suppress the default
969 values for $authenticated_id and $authenticated_sender (but permit -oMai
970 and -oMas) when testing with -bh.
972 PH/24 Re-jigged the order of the tests in the default configuration so that the
973 tests for valid domains and recipients precede the DNS black list and CSA
974 tests, on the grounds that those ones are more expensive.
976 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
977 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
978 This bug exists in every version of Exim that I still have, right back to
981 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
982 However, an attempt to turn on case-sensitivity in a regex key by
983 including (?-i) didn't work because the subject string was already
984 lowercased, and the effects were non-intuitive. It turns out that a
985 one-line patch can be used to allow (?-i) to work as expected.
991 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
992 other effects) broke the use of negated acl sub-conditions.
994 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
997 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
998 "Deny" causes Exim to reject the incoming connection with a 554 error.
999 Unfortunately, if there is a major crisis, such as a disk failure,
1000 tcp-wrappers gives "deny", whereas what one would like would be some
1001 kind of temporary error. A kludge has been added to help with this.
1002 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1003 554 error is used if errno is still zero or contains ENOENT (which occurs
1004 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1007 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1009 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1010 errors. Otherwise a message that provokes a temporary error (when other
1011 messages do not) can cause a whole host to time out.
1013 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1014 addresses were routed directly to files or pipes from a redirect router.
1015 File deliveries just didn't batch; pipe deliveries might have suffered
1018 PH/06 A failure to get a lock for a hints database would erroneously always say
1019 "Failed to get write lock", even when it was really a read lock.
1021 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1022 of 0600. This has been changed to use the value of the lockfile_mode
1023 option (which defaults to 0600).
1025 PH/08 Applied small patch from the Sieve maintainer.
1027 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1028 folder from quota calculations, a direct delivery into this folder messed
1029 up the contents of the maildirsize file. This was because the regex was
1030 used only to exclude .Trash (or whatever) when the size of the mailbox
1031 was calculated. There was no check that a delivery was happening into an
1032 excluded directory. This bug has been fixed by ignoring all quota
1033 processing for deliveries into excluded directories.
1035 PH/10 Added the maildirfolder_create_regex option to appendfile.
1041 PH/01 The code for finding all the local interface addresses on a FreeBSD
1042 system running IPv6 was broken. This may well have applied to all BSD
1043 systems, as well as to others that have similar system calls. The broken
1044 code found IPv4 interfaces correctly, but gave incorrect values for the
1045 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1046 that it would not match correctly against @[] and not recognize the IPv6
1049 PH/02 The ipliteral router was not recognizing addresses of the form user@
1050 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1052 PH/03 Added disable_ipv6.
1054 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1055 decoded form, because it is most often used to construct To: headers
1056 lines in autoreplies, and the decoded form may well be syntactically
1057 invalid. However, $reply_address has leading white space removed, and all
1058 newlines turned into spaces so that the autoreply transport does not
1061 PH/05 If group was specified without a user on a router, and no group or user
1062 was specified on a transport, the group from the router was ignored.
1064 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1065 for visible compile-time settings that can be used to change these
1066 numbers, for those that want even more. Backwards compatibility with old
1067 spool files has been maintained. However, going back to a previous Exim
1068 release will lost any variables that are in spool files.
1070 PH/07 Two small changes when running in the test harness: increase delay when
1071 passing a TCP/IP connection to a new process, in case the original
1072 process has to generate a bounce, and remove special handling of
1073 127.0.0.2 (sic), which is no longer necessary.
1075 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1076 be the same on different OS.
1078 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1081 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1082 whether --show-vars was specified or not
1084 JJ/02 exipick: Added support for new ACL variable spool format introduced
1087 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1088 syntactically invalid From: or Reply-to: line, and a filter used this to
1089 generate an autoreply, and therefore failed to obtain an address for the
1090 autoreply, Exim could try to deliver to a non-existent relative file
1091 name, causing unrelated and misleading errors. What now happens is that
1092 it logs this as a hard delivery error, but does not attempt to create a
1095 PH/11 The exinext utility has a -C option for testing purposes, but although
1096 the given file was scanned by exinext itself; it wasn't being passed on
1097 when Exim was called.
1099 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1100 an end-of-file indication when reading a command response.
1102 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1103 compiled. In many other places in Exim, IPv6 addresses are always
1104 recognized, so I have changed this. It also means that IPv4 domain
1105 literals of the form [IPV4:n.n.n.n] are now always recognized.
1107 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1108 used if the router is not running as root, for example, when verifying at
1109 ACL time, or when using -bh. The debugging output from this situation was
1110 non-existent - all you got was a failure to exec. I have made two
1113 (a) Failures to set uid/gid, the current directory, or a process leader
1114 in a subprocess such as that created by queryprogram now generate
1115 suitable debugging ouput when -d is set.
1117 (b) The queryprogram router detects when it is not running as root,
1118 outputs suitable debugging information if -d is set, and then runs
1119 the subprocess without attempting to change uid/gid.
1121 PH/15 Minor change to Makefile for building test_host (undocumented testing
1124 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1125 additional section of a DNS packet that returns MX or SRV records.
1126 Instead, it always explicitly searches for A/AAAA records. This avoids
1127 major problems that occur when a DNS server includes only records of one
1128 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1129 fixed another bug: if SRV records were looked up and the corresponding
1130 address records were *not* found in the additional section, the port
1131 values from the SRV records were lost.
1133 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1134 using the correct key (the original address) when searching the retry
1135 rules in order to find which one to use for generating the retry hint.
1137 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1138 adding the default one. Similarly, if it contains a Reply-To: header, the
1139 errors_reply_to option, if set, is not used.
1141 PH/19 When calculating a retry time, Exim used to measure the "time since
1142 failure" by looking at the "first failed" field in the retry record. Now
1143 it does not use this if it is later than than the arrival time of the
1144 message. Instead it uses the arrival time. This makes for better
1145 behaviour in cases where some deliveries succeed, thus re-setting the
1146 "first failed" field. An example is a quota failure for a huge message
1147 when small messages continue to be delivered. Without this change, the
1148 "time since failure" will always be short, possible causing more frequent
1149 delivery attempts for the huge message than are intended.
1150 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1152 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1153 $1, $2, $3) because the numerical variables can be reset during some
1154 expansion items (e.g. "match"), thereby losing the authentication data.
1156 PH/21 Make -bV show the size of off_t variables so that the test suite can
1157 decide whether to run tests for quotas > 2G.
1159 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1160 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1161 filecount value is greater than 2G or if a quota value is greater than 2G
1162 on a system where the size of off_t is not greater than 4, a panic error
1165 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1166 never match. The debug and -bh output now contains an explicit error
1167 message indicating a malformed IPv4 address or mask.
1169 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1170 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1171 PH/23 above applies.
1173 PH/25 Do not write to syslog when running in the test harness. The only
1174 occasion when this arises is a failure to open the main or panic logs
1175 (for which there is an explicit test).
1177 PH/26 Added the /no_tell option to "control=freeze".
1179 PH/27 If a host name lookup failed very early in a connection, for example, if
1180 the IP address matched host_lookup and the reverse lookup yielded a name
1181 that did not have a forward lookup, an error message of the form "no IP
1182 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1183 could be logged. Now it outputs the IP address instead of "NULL".
1185 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1186 allows the sender and the authenticated sender to be set when
1187 submitting a message from within Exim. Since child_open_exim() is
1188 documented for local_scan(), the new function should be too.
1190 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1191 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1192 results in an empty string is now treated as unset.
1194 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1196 PH/31 Added sender_verify_fail logging option.
1198 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1199 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1200 tidied the source and removed it altogether.
1202 PH/33 When a queue run was abandoned because the load average was too high, a
1203 log line was always written; now it is written only if the queue_run log
1204 selector is set. In addition, the log line for abandonment now contains
1205 information about the queue run such as the pid. This is always present
1206 in "start" and "stop" lines but was omitted from the "abandon" line.
1208 PH/34 Omit spaces between a header name and the colon in the error message that
1209 is given when verify = headers_syntax fails (if there are lots of them,
1210 the message gets confusing).
1212 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1213 names, as there are now some PTR records that contain slashes. This check
1214 is only to protect against broken name servers that fall over on strange
1215 characters, so the fact that it applies to all lookups doesn't matter.
1217 PH/36 Now that the new test suite is complete, we can remove some of the
1218 special code in Exim that was needed for the old test suite. For example,
1219 sorting DNS records because real resolvers return them in an arbitrary
1220 order. The new test suite's fake resolver always returns records in the
1223 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1224 bounces) except when queue_only is set, to avoid logging races between
1225 the different processes.
1227 PH/38 Panic-die if .include specifies a non-absolute path.
1229 PH/39 A tweak to the "H" retry rule from its user.
1231 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1232 a label. They prevented compilation on older perls.
1234 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1235 a warning to be raised on newish perls.
1237 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1238 on queue. Changes to match documented behaviour of showing count of
1239 messages matching specified criteria.
1241 PH/40 Changed the default ident timeout from 30s to 5s.
1243 PH/41 Added support for the use of login_cap features, on those BSD systems
1244 that have them, for controlling the resources used by pipe deliveries.
1246 PH/42 The content-scanning code uses fopen() to create files in which to put
1247 message data. Previously it was not paying any attention to the mode of
1248 the files. Exim runs with umask(0) because the rest of the code creates
1249 files with open(), and sets the required mode explicitly. Thus, these
1250 files were ending up world-writeable. This was not a big issue, because,
1251 being within the spool directory, they were not world-accessible. I have
1252 created a function called modefopen, which takes an additional mode
1253 argument. It sets umask(777), creates the file, chmods it to the required
1254 mode, then resets the umask. All the relevant calls to fopen() in the
1255 content scanning code have been changed to use this function.
1257 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1258 to 24 hours. This avoids potential overflow problems when processing G
1259 and H retry rules. I suspect nobody ever tinkers with this value.
1261 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1263 PH/45 When the plaintext authenticator is running as a client, the server's
1264 challenges are checked to ensure they are valid base64 strings. By
1265 default, the authentication attempt is cancelled if an invalid string is
1266 received. Setting client_ignore_invalid_base64 true ignores these errors.
1267 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1268 they are received. Thus, the responses can be made to depend on the
1269 challenges. If an invalid string is ignored, an empty string is placed in
1272 PH/46 Messages that are created by the autoreply transport now contains a
1273 References: header, in accordance with RFCs 2822 and 3834.
1275 PH/47 Added authenticated_sender_force to the smtp transport.
1277 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1279 PH/49 Installed latest patch from the Sieve maintainer.
1281 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1282 was also set, the appendfile transport was unnecessarily scanning a
1283 directory of message files (e.g. for maildir delivery) to find the count
1284 of files (along with the size), even though it did not need this
1285 information. It now does the scan only if it needs to find either the
1286 size of the count of files.
1288 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1290 PH/52 Two bugs concerned with error handling when the smtp transport is
1293 (i) Exim was not creating retry information for temporary errors given
1294 for individual recipients after the DATA command when the smtp transport
1295 was used in LMTP mode. This meant that they could be retried too
1296 frequently, and not timed out correctly.
1298 (ii) Exim was setting the flag that allows error details to be returned
1299 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1300 recipients that were returned after the DATA command.
1302 PH/53 This is related to PH/52, but is more general: for any failing address,
1303 when detailed error information was permitted to be returned to the
1304 sender, but the error was temporary, then after the final timeout, only
1305 "retry timeout exceeded" was returned. Now it returns the full error as
1306 well as "retry timeout exceeded".
1308 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1309 do this, and (what is worse) MTAs that accept it.
1311 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1312 will now be deprecated.
1314 PH/56 New os.c-cygwin from the Cygwin maintainer.
1316 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1317 formats (previously only available in exim formats via -bpr, -bpru,
1318 and -bpra. Now also available in native and exiqgrep formats)
1320 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1321 with very large, slow to parse queues
1323 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1325 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1327 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1328 responses to authentication challenges, though it was showing the
1329 challenges; (ii) I've removed the CR characters from the debug output for
1332 PH/58 Allow for the insertion of a newline as well as a space when a string
1333 is turned into more than one encoded-word during RFC 2047 encoding. The
1334 Sieve code now uses this.
1336 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1337 data_4xx, lost_connection, tls_required.
1339 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1340 message was being sent as an SMTP response.
1342 PH/61 Add -l and -k options to exicyclog.
1344 PH/62 When verifying, if an address was redirected to one new address, so that
1345 verification continued, and the new address failed or deferred after
1346 having set something in $address_data, the value of $address_data was not
1347 passed back to the ACL. This was different to the case when no
1348 redirection occurred. The value is now passed back in both cases.
1350 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1351 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1352 use login_cap.h, so on its own it isn't the distinguishing feature. The
1353 new name refers directly to the setclassresources() function.
1355 PH/65 Added configuration files for NetBSD3.
1357 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1359 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1360 is preferred over IPv4.
1362 PH/68 The bounce_return_message and bounce_return_body options were not being
1363 honoured for bounces generated during the reception of non-SMTP messages.
1364 In particular, this applied to messages rejected by the ACL. This bug has
1365 been fixed. However, if bounce_return_message is true and bounce_return_
1366 body is false, the headers that are returned for a non-SMTP message
1367 include only those that have been read before the error was detected.
1368 (In the case of an ACL rejection, they have all been read.)
1370 PH/69 The HTML version of the specification is now built in a directory called
1371 spec_html instead of spec.html, because the latter looks like a path with
1372 a MIME-type, and this confuses some software.
1374 PH/70 Catch two compiler warnings in sieve.c.
1376 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1377 function verify_get_ident() calls ip_connect() to connect a socket, but
1378 if the "connect()" function timed out, ip_connect() used to close the
1379 socket. However, verify_get_ident() also closes the socket later, and in
1380 between Exim writes to the log, which may get opened at this point. When
1381 the socket was closed in ip_connect(), the log could get the same file
1382 descriptor number as the socket. This naturally causes chaos. The fix is
1383 not to close the socket in ip_connect(); the socket should be closed by
1384 the function that creates it. There was only one place in the code where
1385 this was missing, in the iplookup router, which I don't think anybody now
1386 uses, but I've fixed it anyway.
1388 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1389 well as to direct DNS lookups. Otherwise the handling of names in host
1390 lists is inconsistent and therefore confusing.
1396 PH/01 Two changes to the default runtime configuration:
1398 (1) Move the checks for relay_from_hosts and authenticated clients from
1399 after to before the (commented out) DNS black list checks.
1401 (2) Add control=submission to the relay_from_hosts and authenticated
1402 clients checks, on the grounds that messages accepted by these
1403 statements are most likely to be submissions.
1405 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1407 (1) Generate an error if the third argument for the ${prvs expansion is
1410 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1413 (3) Reset the variables that are obtained from the first argument of
1414 ${prvscheck and used in the second argument before leaving the code,
1415 because their memory is reclaimed, so using them afterwards may do
1418 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1419 one (it's much easier than Tom thought :-).
1421 (5) Because of (4), we can now allow for the use of $prvscheck_result
1422 inside the third argument.
1424 PH/03 For some reason, the default setting of PATH when running a command from
1425 a pipe transport was just "/usr/bin". I have changed it to
1428 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1429 anything to be listed in the output from -bV.
1431 PH/05 When a filter generated an autoreply, the entire To: header line was
1432 quoted in the delivery log line, like this:
1434 => >A.N.Other <ano@some.domain> <original@ddress> ...
1436 This has been changed so that it extracts the operative address. There
1437 may be more than one such address. If so, they are comma-separated, like
1440 => >ano@some.domain,ona@other.domain <original@ddress> ...
1442 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1443 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1444 not being looked up in the rDNS to get a host name, Exim was showing the
1445 IP address twice in Received: lines, even though the IP addresses were
1446 identical. For example:
1448 Received: from [1.2.3.4] (helo=[1.2.3.4])
1450 However, if the real host name was known, it was omitting the HELO data
1451 if it matched the actual IP address. This has been tidied up so that it
1452 doesn't show the same IP address twice.
1454 PH/07 When both +timestamp and +memory debugging was on, the value given by
1455 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1456 called by the debug printing, thereby overwriting the timestamp buffer.
1457 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1459 PH/08 When the original message was included in an autoreply transport, it
1460 always said "this is a copy of the message, including all the headers",
1461 even if body_only or headers_only was set. It now gives an appropriate
1464 PH/09 Applied a patch from the Sieve maintainer which:
1466 o fixes some comments
1467 o adds the (disabled) notify extension core
1468 o adds some debug output for the result of if/elsif tests
1469 o points to the current vacation draft in the documentation
1470 and documents the missing references header update
1474 o fixes a bug in processing the envelope test (when testing
1475 multiple envelope elements, the last element determinted the
1478 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1479 Electronic Mail") by including:
1481 Auto-submitted: auto-generated
1483 in the messages that it generates (bounce messages and others, such as
1484 warnings). In the case of bounce messages for non-SMTP mesages, there was
1485 also a typo: it was using "Auto_submitted" (underscore instead of
1486 hyphen). Since every message generated by Exim is necessarily in response
1487 to another message, thes have all been changed to:
1489 Auto-Submitted: auto-replied
1491 in accordance with these statements in the RFC:
1493 The auto-replied keyword:
1495 - SHOULD be used on messages sent in direct response to another
1496 message by an automatic process,
1498 - MUST NOT be used on manually-generated messages,
1500 - MAY be used on Delivery Status Notifications (DSNs) and Message
1501 Disposition Notifications (MDNs),
1503 - MUST NOT be used on messages generated by automatic or periodic
1504 processes, except for messages which are automatic responses to
1507 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1508 to the default Received: header definition.
1510 PH/12 Added log selector acl_warn_skipped (default on).
1512 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1513 variables because (a) they are in the wrong storage pool and (b) even if
1514 they were copied, it wouldn't work properly because of the caching.
1516 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1517 checking when decoding. Apparently there are clients that generate
1518 overlong encoded strings. Why am I not surprised?
1520 PH/15 If the first argument of "${if match_address" was not empty, but did not
1521 contain an "@" character, Exim crashed. Now it writes a panic log message
1522 and treats the condition as false.
1524 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1526 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1527 extension "envelope-auth". The code is finished and in agreement with
1528 other implementations, but there is no documentation so far and in fact,
1529 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1530 not changing the active code.
1532 Print executed "if" and "elsif" statements when debugging is used. This
1533 helps a great deal to understand what a filter does.
1535 Document more things not specified clearly in RFC3028. I had all this
1536 sorted out, when out of a sudden new issues came to my mind. Oops."
1538 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1541 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1542 canonical form (as documented). However, after a host name lookup from
1543 the IP address, check_host() was doing a simple string comparison with
1544 addresses acquired from the DNS when checking that the found name did
1545 have the original IP as one of its addresses. Since any found IPv6
1546 addresses are likely to be in abbreviated form, the comparison could
1547 fail. Luckily, there already exists a function for doing the comparison
1548 by converting both addresses to binary, so now that is used instead of
1549 the text comparison.
1551 PH/20 There was another similar case to PH/19, when a complete host name was
1552 given in a host list; looking up its IP address could give an abbreviated
1553 form, whereas the current host's name might or might not be abbreviated.
1554 The same fix has been applied.
1560 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1561 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1564 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1565 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1567 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1569 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1570 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1571 Also, for "Auto-Submitted", treat anything other than "no" as
1572 non-personal, in accordance with RFC 3834. (Previously it treated
1573 anything starting "auto-" as non-personal.)
1575 TF/01 The control=submission/name=... option had a problem with syntax
1576 errors if the name included a slash character. The /name= option
1577 now slurps the rest of the string, so it can include any characters
1578 but it must come last in the list of options (after /sender_retain
1581 PH/05 Some modifications to the interface to the fake nameserver for the new
1589 TK/01 Added the "success_on_redirect" address verification option. See
1590 NewStuff for rationale and an example.
1592 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1594 PH/02 Patch to exigrep to allow it to work on syslog lines.
1596 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1597 fread() to skip over the body file's header line, because in Cygwin the
1598 header line is locked and is inaccessible.
1600 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1601 co-exist for some time) to make it clear that it is the Exim ID that is
1602 referenced, not the Message-ID: header line.
1604 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1605 string_format() function, because snprintf() does not exist on all
1608 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1609 ${stat: expansion item.
1611 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1612 protocol synchronization error", to keep the pedants happy.
1614 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1615 well as for IRIX systems, when gcc is being used. See the host.c source
1618 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1620 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1623 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1624 transport and to the smtp transport in LMTP mode.
1626 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1628 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1629 run a filter in a subprocess. This could lead to confusion in subsequent
1630 lookups in the parent process. There should also be a search_tidyup() at
1631 the end of the subprocess.
1633 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1634 only if the host matched helo_try_verify_hosts, which caused the
1635 verification to occur when the EHLO/HELO command was issued. The ACL just
1636 tested the remembered result. Now, if a previous verification attempt has
1637 not happened, "verify = helo" does it there and then.
1639 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1641 TK/03 Fix log output including CR from clamd.
1643 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1644 exist provoked a memory error which could cause a segfault.
1646 PH/15 Installed PCRE 6.2
1648 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1650 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1651 of the problem. Specifically, suggested +O2 rather than +O1 for the
1654 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1656 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1657 record, the port was not correctly logged when the outgoing_port log
1658 selector was set (it logged the transort's default port).
1660 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1661 fallback_hosts, and "hosts" in the smtp transport.
1663 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1664 host errors such as "Connection refused".
1666 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1667 authentication with radiusclient 0.4.9:
1669 - Error returned from rc_read_config was caught wrongly
1670 - Username/password not passed on to radius server due to wrong length.
1672 The presumption is that some radiusclient API changes for 4.51/PH/17
1673 were not taken care of correctly. The code is still untested by me (my
1674 Linux distribution still has 0.3.2 of radiusclient), but it was
1675 contributed by a Radius user.
1677 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1678 expanding the "port" option of the smtp transport.
1680 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1681 while reading a MIME header. Thanks to Tom Hughes for a patch.
1683 PH/24 Include config.h inside local_scan.h so that configuration settings are
1686 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1687 that in an ACL for RCPT (for example), you can examine exactly what was
1690 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1691 commands, but it was not correctly comparing the address with the actual
1692 client host address. Thus, it would show the EHLO address in Received:
1693 header lines when this was not necessary.
1695 PH/27 Added the % operator to ${eval:}.
1697 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1698 it should be ignoring failures (because with -C, for example, it has lost
1699 privilege). It wasn't ignoring creation failures other than "already
1702 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1705 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1706 because an input line was too long, either on its own, or by virtue of
1707 too many continuations, the temporary file was not being removed, and the
1708 return code was incorrect.
1710 PH/31 Missing "BOOL" in function definition in filtertest.c.
1712 PH/32 Applied Sieve patches from the maintainer.
1714 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1716 PH/33 Added "verify = not_blind".
1718 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1719 Local/Makefile (with some defaults set). These are used in built scripts
1720 such as exicyclog, but they have never been used in the exim_install
1721 script (though there are many overriding facilities there). I have
1722 arranged that the exim_install script now takes note of these two
1725 PH/35 Installed configuration files for Dragonfly.
1727 PH/36 When a locally submitted message by a trusted user did not contain a
1728 From: header, and the sender address was obtained from -f or from an SMTP
1729 MAIL command, and the trusted user did not use -F to supply a sender
1730 name, $originator_name was incorrectly used when constructing a From:
1731 header. Furthermore, $originator_name was used for submission mode
1732 messages from external hosts without From: headers in a similar way,
1733 which is clearly wrong.
1735 PH/37 Added control=suppress_local_fixups.
1737 PH/38 When log_selector = +received_sender was set, and the addition of the
1738 sender made the log line's construction buffer exactly full, or one byte
1739 less than full, an overflow happened when the terminating "\n" was
1742 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1743 when the result of a list match is failure because a DNS lookup failed.
1745 PH/40 RM_COMMAND is now used in the building process.
1747 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1748 the "build-* directories that it finds.
1750 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1751 address in a domain literal was a prefix of an interface address.
1753 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1754 when verifying a sender address, unless rewrite_headers is false.
1756 PH/44 Wrote a long comment about why errors_to addresses are verified as
1757 recipients, not senders.
1759 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1760 the ratelimit ACL was added.
1762 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1764 PH/46 Added extra information about PostgreSQL errors to the error string.
1766 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1767 avoiding the need to install special zones in a real server. This is
1768 backwards compatible; if it can't find the fake resolver, it drops back.
1769 Thus, both old and new test suites can be run.
1771 TF/02 Added util/ratelimit.pl
1773 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1774 clock is set back in time.
1776 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1777 Candler <B.Candler@pobox.com>.
1779 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1780 OK for addresses that are the result of redirections.
1782 PH/48 A number of further additions for the benefit of the new test suite,
1783 including a fake gethostbyname() that interfaces to the fake DNS resolver
1786 TF/06 The fix for widen_domains has also been applied to qualify_single and
1787 search_parents which are the other dnslookup options that can cause
1790 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1793 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1795 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1796 service" if it encountered an MX record with an empty target hostname.
1797 The message is now "an MX or SRV record indicated no SMTP service".
1799 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1800 if the DNS of the sending site is misconfigured. This is quite a
1801 common situation. This change restores the behaviour of treating a
1802 helo verification defer as a failure.
1804 PH/51 If self=fail was set on a router, the bounce message did not include the
1805 actual error message.
1811 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1813 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1814 command itself ended in error, the underlying message about the transport
1815 filter timeout was being overwritten with the pipe command error. Now the
1816 underlying error message should be appended to the second error message.
1818 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1820 PH/02 Reduce the amount of output that "make" produces by default. Full output
1821 can still be requested.
1823 PH/03 The warning log line about a condition test deferring for a "warn" verb
1824 was being output only once per connection, rather than after each
1825 occurrence (because it was using the same function as for successful
1826 "warn" verbs). This seems wrong, so I have changed it.
1828 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1829 it should not have, which might have caused a crash in the right
1830 circumstances, but probably never did.
1832 PH/04 Installed a modified version of Tony Finch's patch to make submission
1833 mode fix the return path as well as the Sender: header line, and to
1834 add a /name= option so that you can make the user's friendly name appear
1837 TF/03 Added the control = fakedefer ACL modifier.
1839 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1840 Mark Lowes for thorough testing.
1842 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1844 TK/03 Merged latest SRS patch from Miles Wilton.
1846 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1847 with the definition in sysexits.h (which is #included earlier).
1848 Fortunately, Exim does not actually use EX_OK. The code used to try to
1849 preserve the sysexits.h value, by assumimg that macro definitions were
1850 scanned for macro replacements. I have been disabused of this notion,
1851 so now the code just undefines EX_OK before #including unistd.h.
1853 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1854 in the smtp transport. When a block could not be written in a single
1855 write() function, the timeout was being re-applied to each part-write.
1856 This seems wrong - if the receiver was accepting one byte at a time it
1857 would take for ever. The timeout is now adjusted when this happens. It
1858 doesn't have to be particularly precise.
1860 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1861 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1863 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1864 without a domain if the check to <postmaster@domain> fails.
1866 SC/01 Eximstats: added -xls and the ability to specify output files
1867 (patch written by Frank Heydlauf).
1869 SC/02 Eximstats: use FileHandles for outputing results.
1871 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1873 SC/04 Eximstats: fixed display of large numbers with -nvr option
1875 SC/05 Eximstats: fixed merging of reports with empty tables.
1877 SC/06 Eximstats: added the -include_original_destination flag
1879 SC/07 Eximstats: removed tabs and trailing whitespace.
1881 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1883 TK/06 MBOX spool code: Add real "From " MBOX separator line
1884 so the .eml file is really in mbox format (even though
1885 most programs do not really care). Patch from Alex Miller.
1887 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1888 The latter is generated from $received_to and is only set if the
1889 message has one envelope recipient. SA can use these headers,
1890 obviously out-of-the-box. Patch from Alex Miller.
1892 PH/08 The ${def test on a variable was returning false if the variable's
1893 value was "0", contrary to what the specification has always said!
1894 The result should be true unless the variable is empty.
1896 PH/09 The syntax error of a character other than { following "${if
1897 def:variable_name" (after optional whitespace) was not being diagnosed.
1898 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1899 accidental colon was present, for example, could give incorrect results.
1901 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1902 result is used (not including appendfile, where other changes are about
1905 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1906 This involved changing a lot of size variables from int to off_t. It
1907 should work with maildirs and everything.
1909 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1910 spamd dying while we are connected to it.
1912 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1915 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1916 function for BDB 4.3.
1918 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1920 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1921 the sources for PCRE have been re-arranged and more files are now
1924 PH/15 The code I had for printing potentially long long variables in PH/11
1925 above was not the best (it lost precision). The length of off_t variables
1926 is now inspected at build time, and an appropriate printing format (%ld
1927 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1928 to be "long long int" or "long int". This is needed for the internal
1929 formatting function string_vformat().
1931 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1932 the configuration file to be ":syslog", then the script "guesses" where
1933 the logs files are, rather than using the compiled in default. In our
1934 case the guess is not the same as the compiled default, so the script
1935 suddenly stopped working when I started to use syslog. The patch checks
1936 to see if log_file_path is "". If so, it attempts to read it from exim
1937 with no configuration file to get the compiled in version, before it
1938 falls back to the previous guessing code."
1940 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1941 implementing BATV in an Exim configuration. See NewStuff for the gory
1944 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1945 Makefile that are specific to HP-UX.
1947 PH/18 If the "use_postmaster" option was set for a recipient callout together
1948 with the "random" option, the postmaster address was used as the MAIL
1949 FROM address for the random test, but not for the subsequent recipient
1950 test. It is now used for both.
1952 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1953 patch removes a few documentation additions to RFC 3028, because the
1954 latest draft now contains them. It adds the new en;ascii-case comparator
1955 and a new error check for 8bit text in MIME parts. Comparator and
1956 require names are now matched exactly. I enabled the subaddress
1957 extension, but it is not well tested yet (read: it works for me)."
1959 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1960 rework some of the code of TK/09 above to avoid the hardwired use of
1961 "%lld" and "long long". Replaced the call to snprintf() with a call to
1964 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1965 records point to non-existent hosts", "retry timeout exceeded", and
1966 "retry time not reached for any host after a long failure period".
1968 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1969 experimental DomainKeys support:
1971 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1972 (2) On an error such as an illegally used "control", the wrong name for
1973 the control was given.
1975 These problems did NOT occur unless DomainKeys support was compiled.
1977 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1979 PH/24 Added ${if match_ip condition.
1981 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1982 databases so that it will be absolutely obvious if a crash occurs in the
1983 DB library. This is a regular occurrence (often caused by mis-matched
1986 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1987 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1988 picked up on a user's system that detects such things. There doesn't seem
1989 to be a gcc warning option for this - only an attribute that has to be
1990 put on the function's prototype. It seems that in Fedora Core 4 they have
1991 set this on a number of new functions. No doubt there will be more in due
1994 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1995 specify a transport. However, if an address that was verified by such a
1996 router was the subject of a callout, Exim crashed because it tried to
1997 read the rcpt_include_affixes from the non-existent transport. Now it
1998 just assumes that the setting of that option is false. This bug was
1999 introduced by 4.51/PH/31.
2001 PH/28 Changed -d+all to exclude +memory, because that information is very
2002 rarely of interest, but it makes the output a lot bigger. People tend to
2003 do -d+all out of habit.
2005 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2006 code in os-type was giving problems when libc.so lives in lib64, like on
2009 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2010 aren't the modern standard, and it seems that some systems' include files
2011 don't always have them. Exim was already checking for some of the newer
2012 ones like T_AAAA, and defining it itself. I've added checks for all the
2013 record types that Exim uses.
2015 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2016 not automatically generating a new one, as it is supposed to. This
2017 prevented TLS from working. If the file did exist, but contained invalid
2018 data, a new version was generated, as expected. It was only the case of a
2019 non-existent file that was broken.
2021 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2022 with a change in libdomainkeys > 0.64.
2024 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2025 from DNS. If the selector record carries the flag, it now has
2026 precedence over the domain-wide flag.
2028 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2030 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2031 the use of a port name that isn't defined in /etc/services) occurred, the
2032 message was deferred as in a normal delivery, and thus remained on the
2033 spool, instead of being failed because of the mua_wrapper setting. This
2034 is now fixed, and I tidied up some of the mua_wrapper messages at the
2037 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2038 the messages in a hash of arrays rather than using individual hashes.
2039 This is a bit cleaner and results in dramatic memory savings, albeit
2040 at a slight CPU cost.
2042 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2043 as requested by Marc Sherman.
2045 SC/10 Eximstats: added histograms for user specified patterns as requested
2048 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2050 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2051 fopen() in the content-scanning modules that did not already have it.
2057 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2058 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2060 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2062 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2064 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2065 to test Sieve filters that use "vacation".
2067 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2068 that changes the way the GnuTLS parameters are stored in the cache file.
2069 The new format can be generated externally. For backward compatibility,
2070 if the data in the cache doesn't make sense, Exim assumes it has read an
2071 old-format file, and it generates new data and writes a new file. This
2072 means that you can't go back to an older release without removing the
2075 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2076 work if there are any delivery delays because "one_time" forces the
2077 parent to be marked "delivered", so its unseen clone is never tried
2078 again. For this reason, Exim now forbids the simultaneous setting of
2081 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2082 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2083 entry if you want to know the details. The fix, however, neglected to
2084 consider the case when local delivery batching is involved. The test for
2085 "previously delivered" was not happening when checking to see if an
2086 address could be batched with a previous (undelivered) one; under
2087 certain circumstances this could lead to multiple deliveries to the same
2090 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2091 in its include files, and this causes problems building Exim.
2093 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2094 header_syntax) but Exim was just ignoring anything given after a slash.
2095 In particular, this caused confusion with an attempt to use "verify =
2096 reverse_host_lookup/defer_ok". An error is now given when options are
2097 supplied for verify items that do not have them. (Maybe reverse_host_
2098 lookup should have a defer_ok option, but that's a different point.)
2100 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2101 defined by RFC 821) to 2048, because there were problems with some AUTH
2102 commands, and RFC 1869 says the size should be increased for extended
2103 SMTP commands that take arguments.
2105 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2108 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2109 "unknown" error; now it says that the functionality isn't in the binary.
2111 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2112 an address' error message when a string expansion fails (syntax or
2113 whatever). Otherwise the password may appear in the log. Following change
2114 PH/42 below, there is no longer a chance of it appearing in a bounce
2117 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2119 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2120 of its addresses was ever tried. (Bugzilla bug #2.)
2122 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2123 the result incorrectly in the debug output. (It correctly added a newline
2124 to what was transported.)
2126 TF/01 Added $received_time.
2128 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2129 commented out examples of how to interface to a virus scanner and to
2130 SpamAssassin. Also added commented examples of av_scanner and
2131 spamd_address settings.
2133 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2134 and controls are allowed in which ACLs. There were a couple of minor
2135 errors. Some of the entries in the conditions table (which is a table of
2136 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2137 negation of where the condition IS allowed.
2139 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2141 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2142 header file does not have a version number, so I've had to invent a new
2143 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2144 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2145 radiusclient), but it was contributed by a Radius user.
2147 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2148 files or directories, for OpenSSL.
2150 PH/19 When an Exim process that is running as root has to create an Exim log
2151 file, it does so in a subprocess that runs as exim:exim so as to get the
2152 ownership right at creation (otherwise, other Exim processes might see
2153 the file with the wrong ownership). There was no test for failure of this
2154 fork() call, which would lead to the process getting stuck as it waited
2155 for a non-existent subprocess. Forks do occasionally fail when resources
2156 run out. I reviewed all the other calls to fork(); they all seem to check
2159 PH/20 When checking for unexpected SMTP input at connect time (before writing
2160 the banner), Exim was not dealing correctly with a non-positive return
2161 from the read() function. If the client had disconnected by this time,
2162 the result was a log entry for a synchronization error with an empty
2163 string after "input=" when read() returned zero. If read() returned -1
2164 (an event I could not check), uninitialized data bytes were printed.
2165 There were reports of junk text (parts of files, etc) appearing after
2168 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2170 PH/22 Added support for macro redefinition, and (re)definition in between
2171 driver and ACL definitions.
2173 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2174 forgetting to use the resulting value; it was using the unexpanded value.
2176 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2177 hadn't been configured. The fix is from Juergen Kreileder, who
2178 understands it better than I do:
2180 "Here's what I see happening with three configured cyrus_sasl
2181 authenticators configured (plain, login, cram-md5):
2183 On startup auth_cyrus_sasl_init() gets called for each of these.
2184 This means three calls to sasl_listmech() without a specified mech_list.
2185 => SASL tests which mechs of all available mechs actually work
2186 => three warnings about OTP not working
2187 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2189 With the patch, sasl_listmech() also gets called three times. But now
2190 SASL's mech_list option is set to the server_mech specified in the the
2191 authenticator. Or in other words, the answer from sasl_listmech()
2192 gets limited to just the mech you're testing for (which is different
2194 => the return list contains just 'plain' or 'login', 'cram-md5' or
2195 nothing depending on the value of ob->server_mech.
2197 I've just tested the patch: Authentication still works fine,
2198 unavailable mechs specified in the exim configuration are still
2199 caught, and the auth.log warnings about OTP are gone."
2201 PH/25 When debugging is enabled, the contents of the command line are added
2202 to the debugging output, even when log_selector=+arguments is not
2205 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2206 answer is "GNU", and only if the return is "GNU/something" is the answer
2209 PH/27 $acl_verify_message is now set immediately after the failure of a
2210 verification in an ACL, and so is available in subsequent modifiers. In
2211 particular, the message can be preserved by coding like this:
2213 warn !verify = sender
2214 set acl_m0 = $acl_verify_message
2216 Previously, $acl_verify_message was set only while expanding "message"
2217 and "log_message" when a very denied access.
2219 PH/28 Modified OS/os.c-Linux with
2221 -#ifndef OS_LOAD_AVERAGE
2222 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2224 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2225 nomenclature these days.)
2227 PH/29 Installed patch from the Sieve maintainer that adds the options
2228 sieve_useraddress and sieve_subaddress to the redirect router.
2230 PH/30 In these circumstances:
2231 . Two addresses routed to the same list of hosts;
2232 . First host does not offer TLS;
2233 . First host accepts first address;
2234 . First host gives temporary error to second address;
2235 . Second host offers TLS and a TLS session is established;
2236 . Second host accepts second address.
2237 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2238 and peerdn, if requested) that were in fact used only for the second
2241 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2242 attention to any local part prefix or suffix that was matched by the
2243 router that accepted the address. It now behaves in the same way as it
2244 does for delivery: the affixes are removed from the local part unless
2245 rcpt_include_affixes is set on the transport.
2247 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2248 timeout during the DATA phase of an incoming message.
2250 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2251 applied a patch sanctioned by the Sieve maintainer.
2253 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2254 the uid or gid is negative. A case of a negative gid caused this to be
2255 noticed. The fix allows for either to be negative.
2257 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2258 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2261 PH/36 The change PH/12 above was broken. Fixed it.
2263 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2264 the grounds that routing the same address twice would always produce the
2265 same answer. This might have been true once, but it is certainly no
2266 longer true now. Routing a child address may depend on the previous
2267 routing that produced that child. Some complicated redirection strategies
2268 went wrong when messages had multiple recipients, and made Exim's
2269 behaviour dependent on the order in which the addresses were given.
2271 I have moved the duplicate checking until after the routing is complete.
2272 Exim scans the addresses that are assigned to local and remote
2273 transports, and removes any duplicates. This means that more work will be
2274 done, as duplicates will always all be routed, but duplicates are
2275 presumably rare, so I don't expect this is of any significance.
2277 For deliveries to pipes, files, and autoreplies, the duplicate checking
2278 still happens during the routing process, since they are not going to be
2281 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2282 It corrects a timeout issue with spamd. This is Ian's comment: "The
2283 background is that sometimes spamd either never reads data from a
2284 connection it has accepted, or it never writes response data. The exiscan
2285 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2286 blindly assumes that writes won't block so it may never time out."
2288 PH/39 Allow G after quota size as well as K and M.
2290 PH/40 The value set for $authenticated_id in an authenticator may not contain
2291 binary zeroes or newlines because the value is written to log lines and
2292 to spool files. There was no check on this. Now the value is run through
2293 the string_printing() function so that such characters are converted to
2294 printable escape sequences.
2296 PH/41 $message_linecount is a new variable that contains the total number of
2297 lines in the message. Compare $body_linecount, which is the count for the
2300 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2301 bounce and delay warning messages, except in certain special cases, which
2304 (a) An SMTP error message from a remote host;
2305 (b) A message specified in a :fail: redirection;
2306 (c) A message specified in a "fail" command in a system filter;
2307 (d) A message specified in a FAIL return from the queryprogram router;
2308 (e) A message specified by the cannot_route_message router option.
2310 In these cases only, Exim does include the error details in bounce and
2311 warning messages. There are also a few cases where bland messages such
2312 as "unrouteable address" or "local delivery error" are given.
2314 PH/43 $value is now also set for the "else" part of a ${run expansion.
2316 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2317 being worked on, but at least Exim now implements the latest version to
2320 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2321 process to complete was treated as a delivery failure, a timeout while
2322 writing the message to the pipe was logged, but erroneously treated as a
2323 successful delivery. Such timeouts include transport filter timeouts. For
2324 consistency with the overall process timeout, these timeouts are now
2325 treated as errors, giving rise to delivery failures by default. However,
2326 there is now a new Boolean option for the pipe transport called
2327 timeout_defer, which, if set TRUE, converts the failures into defers for
2328 both kinds of timeout. A transport filter timeout is now identified in
2331 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2332 systems where "make" and "gmake" are different, calling "gmake" at top
2333 level broke things. I've arranged for the value of $(MAKE) to be passed
2334 from the Makefile to this script so that it can call the same version of
2338 A note about Exim versions 4.44 and 4.50
2339 ----------------------------------------
2341 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2342 changes of various kinds. As a consequence, a big documentation update was
2343 needed. This delayed the release for rather longer than seemed good, especially
2344 in the light of a couple of (minor) security issues. Therefore, the changes
2345 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2346 release. So 4.44 and 4.50 are in effect two different branches that both start
2349 I have left the 4.50 change log unchanged; it contains all the changes since
2350 4.43. The change log for 4.44 is below; many of its items are identical to
2351 those for 4.50. This seems to be the most sensible way to preserve the
2352 historical information.
2358 1. Minor wording change to the doc/README.SIEVE file.
2360 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2361 computation of the current number of files was incorrect.
2363 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2364 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2367 4. Give more explanation in the error message when the command for a transport
2368 filter fails to execute.
2370 5. There are several places where Exim runs a non-Exim command in a
2371 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2372 was being done only for the command run by the queryprogram router. It is
2373 now done for all such subprocesses. The other cases are: ${run, transport
2374 filters, and the commands run by the lmtp and pipe transports.
2376 6. Added CONFIGURE_GROUP build-time option.
2378 7. Some older OS have a limit of 256 on the maximum number of file
2379 descriptors. Exim was using setrlimit() to set 1000 as a large value
2380 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2381 systems. I've change it so that if it can't get 1000, it tries for 256.
2383 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2384 was an oversight, and furthermore, ever since the addition of extra
2385 controls (e.g. 4.43/32), the checks on when to allow different forms of
2386 "control" were broken. There should now be diagnostics for all cases when a
2387 control that does not make sense is encountered.
2389 9. Added the /retain_sender option to "control=submission".
2391 10. $recipients is now available in the predata ACL (oversight).
2393 11. Tidy the search cache before the fork to do a delivery from a message
2394 received from the command line. Otherwise the child will trigger a lookup
2395 failure and thereby defer the delivery if it tries to use (for example) a
2396 cached ldap connection that the parent has called unbind on.
2398 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2399 of $address_data from the recipient verification was clobbered by the
2400 sender verification.
2402 13. The value of address_data from a sender verification is now available in
2403 $sender_address_data in subsequent conditions in the ACL statement.
2405 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2407 15. Added a new option "connect=<time>" to callout options, to set a different
2410 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2411 was its contents. (It was OK if the option was not defined at all.)
2413 17. A "Completed" log line is now written for messages that are removed from
2414 the spool by the -Mrm option.
2416 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2417 information about exactly what failed.
2419 19. Added -dd to debug only the daemon process.
2421 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2422 handles timeouts, both on the server side and network timeouts. Renamed the
2423 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2425 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2426 It is now set to "smtps".
2428 22. $host_address is now set to the target address during the checking of
2429 ignore_target_hosts.
2431 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2432 being passed; this would have caused $sender_host_name to have been used if
2433 matching the list had actually called for a host name (not very likely,
2434 since this list is usually IP addresses). A host name is now passed as
2437 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2438 code that specifies a non-restarting handler (typically sigaction() in
2439 modern systems) in an attempt to fix a rare and obscure crash bug.
2441 25. Narrowed the window for a race in the daemon that could cause it to ignore
2442 SIGCHLD signals. This is not a major problem, because they are used only to
2443 wake it up if nothing else does.
2445 26. A malformed maildirsize file could cause Exim to calculate negative values
2446 for the mailbox size or file count. Odd effects could occur as a result.
2447 The maildirsize information is now recalculated if the size or filecount
2450 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2451 support for a long time. Removed HAVE_SYS_VFS_H.
2453 28. Installed the latest version of exipick from John Jetmore.
2455 29. In an address list, if the pattern was not a regular expression, an empty
2456 subject address (from a bounce message) matched only if the pattern was an
2457 empty string. Non-empty patterns were not even tested. This was the wrong
2458 because it is perfectly reasonable to use an empty address as part of a
2459 database query. An empty address is now tested by patterns that are
2460 lookups. However, all the other forms of pattern expect the subject to
2461 contain a local part and a domain, and therefore, for them, an empty
2462 address still always fails if the pattern is not itself empty.
2464 30. Exim went into a mad DNS loop when attempting to do a callout where the
2465 host was specified on an smtp transport, and looking it up yielded more
2466 than one IP address.
2468 31. Re-factored the code for checking spool and log partition space into a
2469 function that finds that data and another that does the check. The former
2470 is then used to implement four new variables: $spool_space, $log_space,
2471 $spool_inodes, and $log_inodes.
2473 32. The RFC2047 encoding function was originally intended for short strings
2474 such as real names; it was not keeping to the 75-character limit for
2475 encoded words that the RFC imposes. It now respects the limit, and
2476 generates multiple encoded words if necessary. To be on the safe side, I
2477 have increased the buffer size for the ${rfc2047: expansion operator from
2480 33. It is now permitted to omit both strings after an "if" condition; if the
2481 condition is true, the result is "true". As before, when the second string
2482 is omitted, a false condition yields an empty string. This makes it less
2483 cumbersome to write custom ACL and router conditions.
2485 34. Failure to deliver a bounce message always caused it to be frozen, even if
2486 there was an errors_to setting on the router. The errors_to setting is now
2489 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2490 canonical form (fully expanded) before being placed in
2491 $sender_host_address.
2493 36. The table in the code that translates DNS record types into text (T_A to
2494 "A" for instance) was missing entries for NS and CNAME. It is just possible
2495 that this could have caused confusion if both these types were looked up
2496 for the same domain, because the text type is used as part of Exim's
2497 per-process caching. But the chance of anyone hitting this buglet seems
2500 37. The dnsdb lookup has been extended in a number of ways.
2502 (1) There is a new type, "zns", which walks up the domain tree until it
2503 finds some nameserver records. It should be used with care.
2505 (2) There is a new type, "mxh", which is like "mx" except that it returns
2506 just the host names, not the priorities.
2508 (3) It is now possible to give a list of domains (or IP addresses) to be
2509 looked up. The behaviour when one of the lookups defers can be
2510 controlled by a keyword.
2512 (4) It is now possible to specify the separator character for use when
2513 multiple records are returned.
2515 38. The dnslists ACL condition has been extended: it is now possible to supply
2516 a list of IP addresses and/or domains to be looked up in a particular DNS
2519 39. Added log_selector=+queue_time_overall.
2521 40. When running the queue in the test harness, wait just a tad after forking a
2522 delivery process, to get repeatability of debugging output.
2524 41. Include certificate and key file names in error message when GnuTLS fails
2525 to set them up, because the GnuTLS error message doesn't include the name
2526 of the failing file when there is a problem reading it.
2528 42. Allow both -bf and -bF in the same test run.
2530 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2532 44. The "Exiscan patch" is now merged into the mainline Exim source.
2534 45. Sometimes the final signoff response after QUIT could fail to get
2535 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2536 < 0 before doing a fflush(). This bug looks as though it goes back to the
2537 introduction of TLS in release 3.20, but "sometimes" must have been rare
2538 because the tests only now provoked it.
2540 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2541 (this can affect the format of dates).
2543 47. exim_tidydb, when checking for the continued existence of a message for
2544 which it has found a message-specific retry record, was not finding
2545 messages that were in split spool directories. Consequently, it was
2546 deleting retry records that should have stayed in existence.
2548 48. Steve fixed some bugs in eximstats.
2550 49. The SPA authentication driver was not abandoning authentication and moving
2551 on to the next authenticator when an expansion was forced to fail,
2552 contradicting the general specification for all authenticators. Instead it
2553 was generating a temporary error. It now behaves as specified.
2555 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2556 (the order specifies the preference for clients). The order is now AES256,
2557 AES128, 3DES, ARCFOUR128.
2559 51. Small patch to Sieve code - explicitly set From: when generating an
2562 52. Exim crashed if a remote delivery caused a very long error message to be
2563 recorded - for instance if somebody sent an entire SpamAssassin report back
2564 as a large number of 550 error lines. This bug was coincidentally fixed by
2565 increasing the size of one of Exim's internal buffers (big_buffer) that
2566 happened as part of the Exiscan merge. However, to be on the safe side, I
2567 have made the code more robust (and fixed the comments that describe what
2570 53. Now that there can be additional text after "Completed" in log lines (if
2571 the queue_time_overall log selector is set), a one-byte patch to exigrep
2572 was needed to allow it to recognize "Completed" as not the last thing in
2575 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2576 patch that reportedly fixes this has been added. I am not expert enough to
2577 create a test for it. This is what the patch creator wrote:
2579 "I found a little strange behaviour of ldap code when working with
2580 Windows 2003 AD Domain, where users was placed in more than one
2581 Organization Units. When I tried to give exim partial DN, the exit code
2582 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2583 But simultaneously result of request was absolutely normal ldap result,
2584 so I produce this patch..."
2586 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2587 REFERENCE, so I have modified the code to exclude the patch when that macro
2590 55. Some experimental protocols are using DNS PTR records for new purposes. The
2591 keys for these records are domain names, not reversed IP addresses. The
2592 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2593 leaves it alone. Component reversal etc. now happens only for IP addresses.
2596 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2598 57. Double the size of the debug message buffer (to 2048) so that more of very
2599 long debug lines gets shown.
2601 58. The exicyclog utility now does better if the number of log files to keep
2602 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2604 59. Two changes related to the smtp_active_hostname option:
2606 (1) $smtp_active_hostname is now available as a variable.
2607 (2) The default for smtp_banner uses $smtp_active_hostname instead
2608 of $primary_hostname.
2610 60. The host_aton() function is supposed to be passed a string that is known
2611 to be a valid IP address. However, in the case of IPv6 addresses, it was
2612 not checking this. This is a hostage to fortune. Exim now panics and dies
2613 if the condition is not met. A case was found where this could be provoked
2614 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2615 components; fortuitously, this particular loophole had already been fixed
2616 by change 4.50/55 above.
2618 If there are any other similar loopholes, the new check in host_aton()
2619 itself should stop them being exploited. The report I received stated that
2620 data on the command line could provoke the exploit when Exim was running as
2621 exim, but did not say which command line option was involved. All I could
2622 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2623 running as the user.
2626 61. There was a buffer overflow vulnerability in the SPA authentication code
2627 (which came originally from the Samba project). I have added a test to the
2628 spa_base64_to_bits() function which I hope fixes it.
2631 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2632 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2634 63. The daemon start-up calls getloadavg() while still root for those OS that
2635 need the first call to be done as root, but it missed one case: when
2636 deliver_queue_load_max is set with deliver_drop_privilege. This is
2637 necessary for the benefit of the queue runner, because there is no re-exec
2638 when deliver_drop_privilege is set.
2640 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2641 This has been fixed.
2643 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2644 was in use, was not putting the data itself into the right store pool;
2645 consequently, it could be overwritten for a subsequent message in the same
2646 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2649 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2651 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2652 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2653 the function were treating the return as a boolean value, which happened to
2654 work because 0=false and not-0=true, but is not correct code.
2656 68. The host_aton() function was not handling scoped IPv6 addresses (those
2657 with, for example, "%eth0" on the end) correctly.
2659 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2660 negated items (that is, ~something) in unsigned ints. Some compilers
2661 apparently mutter when there is no cast.
2663 70. If an address verification called from an ACL failed, and did not produce a
2664 user-specific message (i.e. there was only a "system" message), nothing was
2665 put in $acl_verify_message. In this situation, it now puts the system
2668 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2669 session; change 4.31/43 added the unwanted input to the log line - except
2670 that it did not do this in the start of session case. It now does.
2672 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2673 This is wrong and can cause the other end to generate a synchronization
2674 error if it is another Exim or anything else that does the synchronization
2675 check. A QUIT command is no longer sent after a timeout.
2677 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2678 during host lookups.
2680 74. The defer_ok option of callout verification was not working if it was used
2681 when verifying addresses in header lines, that is, for this case:
2683 verify = header_sender/callout=defer_ok
2685 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2686 those file descriptors could be used for SMTP connections. If anything
2687 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2688 could be sent to the SMTP client, causing chaos. The daemon now opens
2689 stdin, stdout, and stderr to /dev/null when it puts itself into the
2692 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2693 log by default. The user can override this with suitable Perl magic.
2695 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2696 the log message when discard triggers, was not working for the DATA ACL or
2697 for the non-SMTP ACL.
2699 78. Error message wording change in sieve.c.
2701 79. If smtp_accept_max_per_host was set, the number of connections could be
2702 restricted to fewer than expected, because the daemon was trying to set up
2703 a new connection before checking whether the processes handling previous
2704 connections had finished. The check for completed processes is now done
2705 earlier. On busy systems, this bug wouldn't be noticed because something
2706 else would have woken the daemon, and it would have reaped the completed
2709 80. If a message was submitted locally by a user whose login name contained one
2710 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2711 It caused a spool format error. I have fixed the spool reading code. A
2712 related problem was that the "from" clause in the Received: line became
2713 illegal because of the space(s). It is now covered by ${quote_local_part.
2715 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2718 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2720 83. Patch from Sieve maintainer to fix unterminated string problem in
2721 "vacation" handling.
2723 84. Some minor changes to the Linux configuration files to help with other
2724 OS variants using glibc.
2726 85. One more patch for Sieve to update vacation handling to latest spec.
2729 ----------------------------------------------------
2730 See the note above about the 4.44 and 4.50 releases.
2731 ----------------------------------------------------
2737 1. Change 4.43/35 introduced a bug that caused file counts to be
2738 incorrectly computed when quota_filecount was set in an appendfile
2741 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2742 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2745 3. Give more explanation in the error message when the command for a transport
2746 filter fails to execute.
2748 4. There are several places where Exim runs a non-Exim command in a
2749 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2750 was being done only for the command run by the queryprogram router. It is
2751 now done for all such subprocesses. The other cases are: ${run, transport
2752 filters, and the commands run by the lmtp and pipe transports.
2754 5. Some older OS have a limit of 256 on the maximum number of file
2755 descriptors. Exim was using setrlimit() to set 1000 as a large value
2756 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2757 systems. I've change it so that if it can't get 1000, it tries for 256.
2759 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2760 was an oversight, and furthermore, ever since the addition of extra
2761 controls (e.g. 4.43/32), the checks on when to allow different forms of
2762 "control" were broken. There should now be diagnostics for all cases when a
2763 control that does not make sense is encountered.
2765 7. $recipients is now available in the predata ACL (oversight).
2767 8. Tidy the search cache before the fork to do a delivery from a message
2768 received from the command line. Otherwise the child will trigger a lookup
2769 failure and thereby defer the delivery if it tries to use (for example) a
2770 cached ldap connection that the parent has called unbind on.
2772 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2773 of $address_data from the recipient verification was clobbered by the
2774 sender verification.
2776 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2777 was its contents. (It was OK if the option was not defined at all.)
2779 11. A "Completed" log line is now written for messages that are removed from
2780 the spool by the -Mrm option.
2782 12. $host_address is now set to the target address during the checking of
2783 ignore_target_hosts.
2785 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2786 being passed; this would have caused $sender_host_name to have been used if
2787 matching the list had actually called for a host name (not very likely,
2788 since this list is usually IP addresses). A host name is now passed as
2791 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2792 code that specifies a non-restarting handler (typically sigaction() in
2793 modern systems) in an attempt to fix a rare and obscure crash bug.
2795 15. Narrowed the window for a race in the daemon that could cause it to ignore
2796 SIGCHLD signals. This is not a major problem, because they are used only to
2797 wake it up if nothing else does.
2799 16. A malformed maildirsize file could cause Exim to calculate negative values
2800 for the mailbox size or file count. Odd effects could occur as a result.
2801 The maildirsize information is now recalculated if the size or filecount
2804 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2805 support for a long time. Removed HAVE_SYS_VFS_H.
2807 18. Updated exipick to current release from John Jetmore.
2809 19. Allow an empty sender to be matched against a lookup in an address list.
2810 Previously the only cases considered were a regular expression, or an
2813 20. Exim went into a mad DNS lookup loop when doing a callout where the
2814 host was specified on the transport, if the DNS lookup yielded more than
2817 21. The RFC2047 encoding function was originally intended for short strings
2818 such as real names; it was not keeping to the 75-character limit for
2819 encoded words that the RFC imposes. It now respects the limit, and
2820 generates multiple encoded words if necessary. To be on the safe side, I
2821 have increased the buffer size for the ${rfc2047: expansion operator from
2824 22. Failure to deliver a bounce message always caused it to be frozen, even if
2825 there was an errors_to setting on the router. The errors_to setting is now
2828 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2829 canonical form (fully expanded) before being placed in
2830 $sender_host_address.
2832 24. Updated eximstats to version 1.33
2834 25. Include certificate and key file names in error message when GnuTLS fails
2835 to set them up, because the GnuTLS error message doesn't include the name
2836 of the failing file when there is a problem reading it.
2838 26. Expand error message when OpenSSL has problems setting up cert/key files.
2841 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2842 (this can affect the format of dates).
2844 28. exim_tidydb, when checking for the continued existence of a message for
2845 which it has found a message-specific retry record, was not finding
2846 messages that were in split spool directories. Consequently, it was
2847 deleting retry records that should have stayed in existence.
2849 29. eximstats updated to version 1.35
2850 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2851 1.35 - bugfix such that pie charts by volume are generated correctly
2853 30. The SPA authentication driver was not abandoning authentication and moving
2854 on to the next authenticator when an expansion was forced to fail,
2855 contradicting the general specification for all authenticators. Instead it
2856 was generating a temporary error. It now behaves as specified.
2858 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2859 (the order specifies the preference for clients). The order is now AES256,
2860 AES128, 3DES, ARCFOUR128.
2862 31. Small patch to Sieve code - explicitly set From: when generating an
2865 32. Exim crashed if a remote delivery caused a very long error message to be
2866 recorded - for instance if somebody sent an entire SpamAssassin report back
2867 as a large number of 550 error lines. This bug was coincidentally fixed by
2868 increasing the size of one of Exim's internal buffers (big_buffer) that
2869 happened as part of the Exiscan merge. However, to be on the safe side, I
2870 have made the code more robust (and fixed the comments that describe what
2873 33. Some experimental protocols are using DNS PTR records for new purposes. The
2874 keys for these records are domain names, not reversed IP addresses. The
2875 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2876 leaves it alone. Component reversal etc. now happens only for IP addresses.
2879 34. The host_aton() function is supposed to be passed a string that is known
2880 to be a valid IP address. However, in the case of IPv6 addresses, it was
2881 not checking this. This is a hostage to fortune. Exim now panics and dies
2882 if the condition is not met. A case was found where this could be provoked
2883 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2884 components; fortuitously, this particular loophole had already been fixed
2885 by change 4.50/55 or 4.44/33 above.
2887 If there are any other similar loopholes, the new check in host_aton()
2888 itself should stop them being exploited. The report I received stated that
2889 data on the command line could provoke the exploit when Exim was running as
2890 exim, but did not say which command line option was involved. All I could
2891 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2892 running as the user.
2895 35. There was a buffer overflow vulnerability in the SPA authentication code
2896 (which came originally from the Samba project). I have added a test to the
2897 spa_base64_to_bits() function which I hope fixes it.
2900 36. The daemon start-up calls getloadavg() while still root for those OS that
2901 need the first call to be done as root, but it missed one case: when
2902 deliver_queue_load_max is set with deliver_drop_privilege. This is
2903 necessary for the benefit of the queue runner, because there is no re-exec
2904 when deliver_drop_privilege is set.
2906 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2907 was in use, was not putting the data itself into the right store pool;
2908 consequently, it could be overwritten for a subsequent message in the same
2909 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2912 38. Sometimes the final signoff response after QUIT could fail to get
2913 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2914 < 0 before doing a fflush(). This bug looks as though it goes back to the
2915 introduction of TLS in release 3.20, but "sometimes" must have been rare
2916 because the tests only now provoked it.
2922 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2923 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2924 Now it returns an integer. A number of calls were still expecting a T/F
2925 return. Fortuitously, in all cases, the tests worked in OK situations,
2926 which is the norm. However, things would have gone wrong on any write
2927 failures on the smtp file descriptor. This function is used when sending
2928 messages over SMTP and also when doing verify callouts.
2930 2. When Exim is called to do synchronous delivery of a locally submitted
2931 message (the -odf or -odi options), it no longer closes stderr before doing
2934 3. Implemented the mua_wrapper option.
2936 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2938 5. Implemented the functions header_remove(), header_testname(),
2939 header_add_at_position(), and receive_remove_recipient(), and exported them
2942 6. If an ACL "warn" statement specified the addition of headers, Exim already
2943 inserted X-ACL-Warn: at the start if there was no header name. However, it
2944 was not making this test for the second and subsequent header lines if
2945 there were newlines in the string. This meant that an invalid header could
2946 be inserted if Exim was badly configured.
2948 7. Allow an ACL "warn" statement to add header lines at the start or after all
2949 the Received: headers, as well as at the end.
2951 8. Added the rcpt_4xx retry error code.
2953 9. Added postmaster_mailfrom=xxx to callout verification option.
2955 10. Added mailfrom=xxxx to the callout verification option, for verify=
2958 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2959 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2961 12. Inserted some casts to stop certain compilers warning when using pointer
2962 differences as field lengths or precisions in printf-type calls (mostly
2963 affecting debugging statements).
2965 13. Added optional readline() support for -be (dynamically loaded).
2967 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2968 same clock tick as a message's arrival, so that its received time was the
2969 same as the "first fail" time on the retry record, and that message
2970 remained on the queue past the ultimate address timeout, every queue runner
2971 would try a delivery (because it was past the ultimate address timeout) but
2972 after another failure, the ultimate address timeout, which should have then
2973 bounced the address, did not kick in. This was a "< instead of <=" error;
2974 in most cases the first failure would have been in the next clock tick
2975 after the received time, and all would be well.
2977 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2978 being recognized when the domain list was tested by the match_domain
2979 condition in an expansion string.
2981 16. Added the ${str2b64: operator.
2983 17. Exim was always calling setrlimit() to set a large limit for the number of
2984 processes, without checking whether the existing limit was already
2985 adequate. (It did check for the limit on file descriptors.) Furthermore,
2986 errors from getrlimit() and setrlimit() were being ignored. Now they are
2987 logged to the main and panic logs, but Exim does carry on, to try to do its
2988 job under whatever limits there are.
2990 18. Imported PCRE 5.0.
2992 19. Trivial typo in log message " temporarily refused connection" (the leading
2995 20. If the log selector return_path_on_delivery was set and an address was
2996 redirected to /dev/null, the delivery process crashed because it assumed
2997 that a return path would always be set for a "successful" delivery. In this
2998 case, the whole delivery is bypassed as an optimization, and therefore no
3001 21. Internal re-arrangement: the function for sending a challenge and reading
3002 a response while authentication was assuming a zero-terminated challenge
3003 string. It's now changed to take a pointer and a length, to allow for
3004 binary data in such strings.
3006 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3008 23. Exim was not respecting finduser_retries when seeking the login of the
3009 uid under which it was called; it was always trying 10 times. (The default
3010 setting of finduser_retries is zero.) Also, it was sleeping after the final
3011 failure, which is pointless.
3013 24. Implemented tls_on_connect_ports.
3015 25. Implemented acl_smtp_predata.
3017 26. If the domain in control=submission is set empty, Exim assumes that the
3018 authenticated id is a complete email address when it generates From: or
3019 Sender: header lines.
3021 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3022 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3023 chown and chgrp in /bin and hostname in /usr/bin.
3025 28. Exim was keeping the "process log" file open after each use, just as it
3026 does for the main log. This opens the possibility of it remaining open for
3027 long periods when the USR1 signal hits a daemon. Occasional processlog
3028 errors were reported, that could have been caused by this. Anyway, it seems
3029 much more sensible not to leave this file open at all, so that is what now
3032 29. The long-running daemon process does not normally write to the log once it
3033 has entered its main loop, and it closes the log before doing so. This is
3034 so that log files can straightforwardly be renamed and moved. However,
3035 there are a couple of unusual error situations where the daemon does write
3036 log entries, and I had neglected to close the log afterwards.
3038 30. The text of an SMTP error response that was received during a remote
3039 delivery was being truncated at 512 bytes. This is too short for some of
3040 the long messages that one sometimes sees. I've increased the limit to
3043 31. It is now possible to make retry rules that apply only when a message has a
3044 specific sender, in particular, an empty sender.
3046 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3047 it possible to be selective about when SMTP synchronization is enforced.
3049 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3051 32. Implemented hosts_connection_nolog.
3053 33. Added an ACL for QUIT.
3055 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3058 35. Added mailbox_size and mailbox_filecount to appendfile.
3060 36. Added control = no_multiline_responses to ACLs.
3062 37. There was a bug in the logic of the code that waits for the clock to tick
3063 in the case where the clock went backwards by a substantial amount such
3064 that the microsecond fraction of "now" was more than the microsecond
3065 fraction of "then" (but the whole seconds number was less).
3067 38. Added support for the libradius Radius client library this is found on
3068 FreeBSD (previously only the radiusclient library was supported).
3074 1. When certain lookups returned multiple values in the form name=value, the
3075 quoting of the values was not always being done properly. Specifically:
3076 (a) If the value started with a double quote, but contained no whitespace,
3078 (b) If the value contained whitespace other than a space character (i.e.
3079 tabs or newlines or carriage returns) it was not quoted.
3080 This fix has been applied to the mysql and pgsql lookups by writing a
3081 separate quoting function and calling it from the lookup code. The fix
3082 should probably also be applied to nisplus, ibase and oracle lookups, but
3083 since I cannot test any of those, I have not disturbed their existing code.
3085 2. A hit in the callout cache for a specific address caused a log line with no
3086 reason for rejecting RCPT. Now it says "Previous (cached) callout
3087 verification failure".
3089 3. There was an off-by-one bug in the queryprogram router. An over-long
3090 return line was truncated at 256 instead of 255 characters, thereby
3091 overflowing its buffer with the terminating zero. As well as fixing this, I
3092 have increased the buffer size to 1024 (and made a note to document this).
3094 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3095 when Exim is waiting for an SMTP response from a remote server, Exim
3096 restarts its select() call on the socket, thereby resetting its timeout.
3097 This is not a problem when such interrupts are rare. Somebody set up a cron
3098 job to run exiwhat every 2 minutes, which is less than the normal select()
3099 timeout (5 or 10 minutes). This meant that the select() timeout never
3100 kicked in because it was always reset. I have fixed this by comparing the
3101 time when an interrupt arrives with the time at the start of the first call
3102 to select(). If more time than the timeout has elapsed, the interrupt is
3103 treated as a timeout.
3105 5. Some internal re-factoring in preparation for the addition of Sieve
3106 extensions (by MH). In particular, the "personal" test is moved to a
3107 separate function, and given an option for scanning Cc: and Bcc: (which is
3108 not set for Exim filters).
3110 6. When Exim created an email address using the login of the caller as the
3111 local part (e.g. when creating a From: or Sender: header line), it was not
3112 quoting the local part when it contained special characters such as @.
3114 7. Installed new OpenBSD configuration files.
3116 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3117 try to make them clearer.
3119 9. Callout options, other than the timeout value, were being ignored when
3120 verifying sender addresses in header lines. For example, when using
3122 verify = header_sender/callout=no_cache
3124 the cache was (incorrectly) being used.
3126 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3127 only the Cygwin environment.
3129 11. When return_path_on_delivery was set as a log selector, if different remote
3130 addresses in the same message used different return paths and parallel
3131 remote delivery occurred, the wrong values would sometimes be logged.
3132 (Whenever a remote delivery process finished, the return path value from
3133 the most recently started remote delivery process was logged.)
3135 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3136 lines when AUTH and/or TLS are in use. This is the "received protocol"
3137 field. Exim used to use "asmtp" for authenticated SMTP, without any
3138 indication (in the protocol name) for TLS use. Now it follows the RFC and
3139 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3140 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3141 names appear in log lines as well as in Received: header lines.
3143 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3144 extensions, and comparison tests, and to fix some bugs.
3146 14. Changes to the "personal" filter test:
3148 (1) The test was buggy in that it was just doing the equivalent of
3149 "contains" tests on header lines. For example, if a user's address was
3150 anne@some.where, the "personal" test would incorrectly be true for
3152 To: susanne@some.where
3154 This test is now done by extracting each address from the header in turn,
3155 and checking the entire address. Other tests that are part of "personal"
3156 are now done using regular expressions (for example, to check local parts
3157 of addresses in From: header lines).
3159 (2) The list of non-personal local parts in From: addresses has been
3160 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3161 taken from the Sieve specification recommendations.
3163 (3) If the message contains any header line starting with "List-" it is
3164 treated as non-personal.
3166 (4) The test for "circular" in the Subject: header line has been removed
3167 because it now seems ill-conceived.
3169 15. Minor typos in src/EDITME comments corrected.
3171 16. Installed latest exipick from John Jetmore.
3173 17. If headers_add on a router specified a text string that was too long for
3174 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3175 of string_sprintf() is now avoided.
3177 18. $message_body_size was not set (it was always zero) when running the DATA
3178 ACL and the local_scan() function.
3180 19. For the "mail" command in an Exim filter, no default was being set for
3181 the once_repeat time, causing a random time value to be used if "once" was
3182 specified. (If the value happened to be <= 0, no repeat happened.) The
3183 default is now 0s, meaning "never repeat". The "vacation" command was OK
3184 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3185 (I found it when inspecting the code).
3187 20. There is now an overall timeout for performing a callout verification. It
3188 defaults to 4 times the callout timeout, which applies to individual SMTP
3189 commands during the callout. The overall timeout applies when there is more
3190 than one host that can be tried. The timeout is checked before trying the
3191 next host. This prevents very long delays if there are a large number of
3192 hosts and all are timing out (e.g. when the network connections are timing
3193 out). The value of the overall timeout can be changed by specifying an
3194 additional sub-option for "callout", called "maxwait". For example:
3196 verify = sender/callout=5s,maxwait=20s
3198 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3199 to the end before writing, but this should make it even safer).
3201 22. Exim was forgetting that it had advertised PIPELINING for the second and
3202 subsequent messages on an SMTP connection. It was also not resetting its
3203 memory on STARTTLS and an internal HELO.
3205 23. When Exim logs an SMTP synchronization error within a session, it now
3206 records whether PIPELINING has been advertised or not.
3208 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3209 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3210 rather than long int.
3212 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3214 26. Added the never_mail option to autoreply.
3220 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3221 crash if the getsockname() call failed; this can happen if a connection is
3222 closed very soon after it is established. The problem was simply in the
3223 order in which certain operations were done, causing Exim to try to write
3224 to the SMTP stream before it had set up the file descriptor. The bug has
3225 been fixed by making things happen in the correct order.
3231 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3232 before the connection was closed, thus losing the rejection response.
3234 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3235 some early Solaris releases, but causes trouble in current releases where
3236 socklen_t is defined.
3238 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3241 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3244 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3245 that releasing the top of it at the end releases what was used for sub-
3246 expansions (unless the block got too big). However, discard this block if
3247 the first thing is a variable or header, so that we can use its block when
3248 it is dynamic (useful for very large $message_headers, for example).
3250 6. Lookups now cache *every* query, not just the most recent. A new, separate
3251 store pool is used for this. It can be recovered when all lookup caches are
3252 flushed. Lookups now release memory at the end of their result strings.
3253 This has involved some general refactoring of the lookup sources.
3255 7. Some code has been added to the store_xxx() functions to reduce the amount
3256 of flapping under certain conditions.
3258 8. log_incoming_interface used to affect only the <= reception log lines. Now
3259 it causes the local interface and port to be added to several more SMTP log
3260 lines, for example "SMTP connection from", and rejection lines.
3262 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3264 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3266 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3267 could be overwritten at the end of the current message (or the start of a
3268 new message if it was set in a HELO ACL). The value is now preserved for
3269 the duration of the SMTP connection.
3271 12. If a transport had a headers_rewrite setting, and a matching header line
3272 contained an unqualified address, that address was qualified, even if it
3273 did not match any rewriting rules. The underlying bug was that the values
3274 of the flags that permit the existence of unqualified sender and recipient
3275 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3276 non-local messages, and by -bnq for local messages) were not being
3277 preserved with the message after it was received.
3279 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3280 "next input=" as part of the text comprising the host identity instead of
3281 the correct text. The code was using the same buffer for two different
3282 strings. However, depending on which order the printing function evaluated
3283 its arguments, the bug did not always show up. Under Linux, for example, my
3284 test suite worked just fine.
3286 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3287 doesn't work with some older versions of Perl. It has been changed to "my",
3288 which in any case is probably the better facility to use.
3290 15. A really picky compiler found some instances of statements for creating
3291 error messages that either had too many or two few arguments for the format
3294 16. The size of the buffer for calls to the DNS resolver has been increased
3295 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3296 for addresses that have a lot of PTR records. This alleviates a problem; it
3297 does not fully solve it.
3299 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3300 buffer now truncates the list and logs the incident, which is the same
3301 action as happens when Exim is looking up a host name and its aliases.
3302 Previously in this situation something unpredictable would happen;
3303 sometimes it was "internal error: store_reset failed".
3305 18. If a server dropped the connection unexpectedly when an Exim client was
3306 using GnuTLS and trying to read a response, the client delivery process
3307 crashed while trying to generate an error log message.
3309 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3310 string, for example:
3312 warn message = H1: something\nH2: something
3314 the text was added as a single header line from Exim's point of view
3315 though it ended up OK in the delivered message. However, searching for the
3316 second and subsequent header lines using $h_h2: did not work. This has been
3317 fixed. Similarly, if a system filter added multiple headers in this way,
3318 the routers could not see them.
3320 20. Expanded the error message when iplsearch is called with an invalid key to
3321 suggest using net-iplsearch in a host list.
3323 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3324 ACLs are no longer actually imposed (and a message to that effect is
3327 22. If a "gecos" field in a passwd entry contained escaped characters, in
3328 particular, if it contained a \" sequence, Exim got it wrong when building
3329 a From: or a Sender: header from that name. A second bug also caused
3330 incorrect handling when an unquoted " was present following a character
3331 that needed quoting.
3333 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3334 was not being matched caselessly.
3336 24. Arranged for all hyphens in the exim.8 source to be escaped with
3339 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3340 properly. Recipient callout cache records were still being keyed to include
3341 the sender, even when use_sender was set false. This led to far more
3342 callouts that were necessary. The sender is no longer included in the key
3343 when use_sender is false.
3345 26. Added "control = submission" modifier to ACLs.
3347 27. Added the ${base62d: operator to decode base 62 numbers.
3349 28. dnsdb lookups can now access SRV records.
3351 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3352 the configuration file.
3354 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3355 (-v) mode. This makes the output for a verbose queue run more intelligible.
3357 31. Added a use_postmaster feature to recipient callouts.
3359 32. Added the $body_zerocount variable, containing the number of binary zero
3360 bytes in the message body.
3362 33. The time of last modification of the "new" subdirectory is now used as the
3363 "mailbox time last read" when there is a quota error for a maildir
3366 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3368 35. Added +ignore_unknown as a special item in host lists.
3370 36. Code for decoding IPv6 addresses in host lists is now included, even if
3371 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3372 address was recognized as an IP address, but was then not correctly decoded
3373 into binary, causing unexpected and incorrect effects when compared with
3380 1. Very minor rewording of debugging text in manualroute to say "list of
3381 hosts" instead of "hostlist".
3383 2. If verify=header_syntax was set, and a header line with an unqualified
3384 address (no domain) and a large number of spaces between the end of the
3385 name and the colon was received, the reception process suffered a buffer
3386 overflow, and (when I tested it) crashed. This was caused by some obsolete
3387 code that should have been removed. The fix is to remove it!
3389 3. When running in the test harness, delay a bit after writing a bounce
3390 message to get a bit more predictability in the log output.
3392 4. Added a call to search_tidyup() just before forking a reception process. In
3393 theory, someone could use a lookup in the expansion of smtp_accept_max_
3394 per_host which, without the tidyup, could leave open a database connection.
3396 5. Added the variables $recipient_data and $sender_data which get set from a
3397 lookup success in an ACL "recipients" or "senders" condition, or a router
3398 "senders" option, similar to $domain_data and $local_part_data.
3400 6. Moved the writing of debug_print from before to after the "senders" test
3403 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3404 problems for message scanning, either using a data ACL, or using
3405 local_scan() because the Received: header was not generated till after they
3406 were called (in order to set the time as the time of reception completion).
3407 I have revised the way this works. The header is now generated after the
3408 body is received, but before the ACL or local_scan() are called. After they
3409 are run, the timestamp in the header is updated.
3415 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3416 before starting a queue runner without re-exec. This happened only when
3417 deliver_drop_privilege was set or when the Exim user was set to root. The
3418 effect of the bug was that timeouts during subsequent deliveries caused
3419 crashes instead of being properly handled. The handler is now left at its
3420 default (and expected) setting.
3422 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3423 message, again when deliver_drop_privilege is set or Exim is run as root.
3424 The bug described in (1) was not present in this case, but the tidying up
3425 of the other signals was missing. I have made the two cases consistent.
3427 3. The ignore_target_hosts setting on a manualroute router was being ignored
3428 for hosts that were looked up using the /MX notation.
3430 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3433 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3434 operated on the sender address. After changing the $sender_address to <>
3435 for the sender address verify, Exim was re-instated it as the original
3436 (before rewriting) address, but remembering that it had rewritten it, so it
3437 wasn't rewriting it again. This bug also had the effect of breaking the
3438 sender address verification caching when the sender address was rewritten.
3440 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3441 This has been changed so that if the ip literal address matches
3442 ignore_target_hosts, the router declines.
3444 7. Added expansion conditions match_domain, match_address, and match_local_
3445 part (NOT match_host).
3447 8. The placeholder for the Received: header didn't have a length field set.
3449 9. Added code to Exim itself and to exim_lock to test for a specific race
3450 condition that could lead to file corruption when using MBX delivery. The
3451 issue is with the lockfile that is created in /tmp. If this file is removed
3452 after a process has opened it but before that process has acquired a lock,
3453 there is the potential for a second process to recreate the file and also
3454 acquire a lock. This could lead to two Exim processes writing to the file
3455 at the same time. The added code performs the same test as UW imapd; it
3456 checks after acquiring the lock that its file descriptor still refers to
3457 the same named file.
3459 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3460 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3463 11. Added the smtp_active_hostname option. If used, this will typically be made
3464 to depend on the incoming interface address. Because $interface_address is
3465 not set up until the daemon has forked a reception process, error responses
3466 that can happen earlier (such as "too many connections") no longer contain
3469 12. If an expansion in a condition on a "warn" statement fails because a lookup
3470 defers, the "warn" statement is abandoned, and the next ACL statement is
3471 processed. Previously this caused the whole ACL to be aborted.
3473 13. Added the iplsearch lookup type.
3475 14. Added ident_timeout as a log selector.
3477 15. Added tls_certificate_verified as a log selector.
3479 16. Added a global option tls_require_ciphers (compare the smtp transport
3480 option of the same name). This controls incoming TLS connections.
3482 17. I finally figured out how to make tls_require_ciphers do a similar thing
3483 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3484 before starting the TLS session.
3486 18. Tabs are now shown as \t in -bP output.
3488 19. If the log selector return_path_on_delivery was set, Exim crashed when
3489 bouncing a message because it had too many Received: header lines.
3491 20. If two routers both had headers_remove settings, and the first one included
3492 a superfluous trailing colon, the final name in the first list and the
3493 first name in the second list were incorrectly joined into one item (with a
3494 colon in the middle).
3500 1. Added -C and -D options to the exinext utility, mainly to make it easier
3501 to include in the automated testing, but these could be helpful when
3502 multiple configurations are in use.
3504 2. The exinext utility was not formatting the output nicely when there was
3505 an alternate port involved in the retry record key, nor when there was a
3506 message id as well (for retries that were specific to a specific message
3507 and a specific host). It was also confused by IPv6 addresses, because of
3508 the additional colons they contain. I have fixed the IPv4 problem, and
3509 patched it up to do a reasonable job for IPv6.
3511 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3512 delivery, the log line now contains "pipelined" if PIPELINING was used.
3514 4. An SMTP transport process used to panic and die if the bind() call to set
3515 an explicit outgoing interface failed. This has been changed; it is now
3516 treated in the same way as a connect() failure.
3518 5. A reference to $sender_host_name in the part of a conditional expansion
3519 that was being skipped was still causing a DNS lookup. This no longer
3522 6. The def: expansion condition was not recognizing references to header lines
3523 that used bh_ and bheader_.
3525 7. Added the _cache feature to named lists.
3527 8. The code for checking quota_filecount in the appendfile transport was
3528 allowing one more file than it should have been.
3530 9. For compatibility with Sendmail, the command line option
3538 and sets the incoming protocol and host name (for trusted callers). The
3539 host name and its colon can be omitted when only the protocol is to be set.
3540 Note the Exim already has two private options, -pd and -ps, that refer to
3541 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3542 "s", but I don't think that's a major issue.
3544 10. A number of refactoring changes to the code, none of which should affect
3547 (a) The number of logging options was getting close to filling up the
3548 32-bit word that was used as a bit map. I have split them into two classes:
3549 those that are passed in the argument to log_write(), and those that are
3550 only ever tested independently outside of that function. These are now in
3551 separate 32-bit words, so there is plenty of room for expansion again.
3552 There is no change in the user interface or the logging behaviour.
3554 (b) When building, for example, log lines, the code previously used a
3555 macro that called string_cat() twice, in order to add two strings. This is
3556 not really sufficiently general. Furthermore, there was one instance where
3557 it was actually wrong because one of the argument was used twice, and in
3558 one call a function was used. (As it happened, calling the function twice
3559 did not affect the overall behaviour.) The macro has been replaced by a
3560 function that can join an arbitrary number of extra strings onto a growing
3563 (c) The code for expansion conditions now uses a table and a binary chop
3564 instead of a serial search (which was left over from when there were very
3565 few conditions). Also, it now recognizes conditions like "pam" even when
3566 the relevant support is not compiled in: a suitably worded error message is
3567 given if an attempt is made to use such a condition.
3569 11. Added ${time_interval:xxxxx}.
3571 12. A bug was causing one of the ddress fields not to be passed back correctly
3572 from remote delivery subprocesses. The field in question was not being
3573 subsequently used, so this caused to problems in practice.
3575 13. Added new log selectors queue_time and deliver_time.
3577 14. Might have fixed a bug in maildirsizefile handling that threw up
3578 "unexpected character" debug warnings, and recalculated the data
3579 unnecessarily. In any case, I expanded the warning message to give more
3582 15. Added the message "Restricted characters in address" to the statements in
3583 the default ACL that block characters like @ and % in local parts.
3585 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3586 Three changes have been made:
3588 (a) There was a serious bug; a negative response to MAIL caused the whole
3589 recipient domain to be cached as invalid, thereby blocking all messages
3590 to all local parts at the same domain, from all senders. This bug has
3591 been fixed. The domain is no longer cached after a negative response to
3592 MAIL if the sender used is not empty.
3594 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3597 (c) A new callout option, "use_sender" has been added for people who want
3598 the modified behaviour.
3604 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3607 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3608 indeed breaks things for older releases.
3610 3. Added additional logging to the case where there is a problem reading data
3611 from a filter that is running in a subprocess using a pipe, in order to
3612 try to track down a specific problem.
3614 4. Testing facility fudge: when running in the test harness and attempting
3615 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3616 getting "No route to host". Convert this to a timeout.
3618 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3621 6. Some OS don't have socklen_t but use size_t instead. This affects the
3622 fifth argument of getsockopt() amongst other things. This is now
3623 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3624 can be set for individual OS. I have set it for SunOS5, OSF1, and
3625 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3626 some earlier ones do not.
3628 7. Change 4.30/15 was not doing the test caselessly.
3630 8. The standard form for an IPv6 address literal was being rejected by address
3631 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3632 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3633 this, as well as the form without the "IPv6" on the front (but only when
3634 address literals are enabled, of course).
3636 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3638 10. Exim crashed if a message with an empty sender address specified by -f
3639 encountered a router with an errors_to setting. This could be provoked only
3640 by a command such as
3644 where an empty string was supplied; "<>" did not hit this bug.
3646 11. Installed PCRE release 4.5.
3648 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3649 remained set. It is now erased.
3651 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3652 times from message ids (which are base 36 rather than the normal 62).
3654 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3655 were being counted as actual protocol errors, and logged if the log
3656 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3657 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3658 and DATA following a set of rejected RCPTs do not count as protocol errors.
3659 In other words, Exim assumes they were pipelined, though this may not
3660 actually be the case. Of course, in all cases the client gets an
3661 appropriate error code.
3663 15. If a lookup fails in an ACL condition, a message about the failure may
3664 be available; it is used if testing the ACL cannot continue, because most
3665 such messages specify what the cause of the deferral is. However, some
3666 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3667 that caused an old message to be retained and used if a later statement
3668 caused a defer, replacing the real cause of the deferral.
3670 16. If an IP address had so many PTR records that the DNS lookup buffer
3671 was not large enough to hold them, Exim could crash while trying to process
3672 the truncated data. It now detects and logs this case.
3674 17. Further to 4.21/58, another change has been made: if (and only if) the
3675 first line of a message (the first header line) ends with CRLF, a bare LF
3676 in a subsequent header line has a space inserted after it, so as not to
3677 terminate the header.
3679 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3680 unnecessarily, used atoi() instead of strtol(), and didn't check the
3681 termination when getting file sizes from file names by regex.
3683 19. Completely re-implemented the support for maildirsize files, in the light
3684 of a number of problems with the previous contributed implementation
3685 (4.30/29). In particular:
3687 . If the quota is zero, the maildirsize file is maintained, but no quota is
3690 . If the maildir directory does not exist, it is created before any attempt
3691 to write a maildirsize file.
3693 . The quota value in the file is just a cache; if the quota is changed in
3694 the transport, the new value overrides.
3696 . A regular expression is available for excluding directories from the
3699 20. The autoreply transport checks the characters in options that define the
3700 message's headers; it allows continued headers, but it was checking with
3701 isspace() after an embedded newline instead of explicitly looking for a
3704 21. If all the "regular" hosts to which an address was routed had passed their
3705 expiry times, and had not reached their retry times, the address was
3706 bounced, even if fallback hosts were defined. Now Exim should go on to try
3709 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3710 equivalent code in the SMTP transport. Some hosts send humungous responses
3711 to HELO/EHLO, more than 1024 it seems.
3713 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3714 gives compiler warnings in some environments. I've now done it "properly",
3717 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3718 (because of problems with the built-in one) was declared to return uschar *
3719 instead of char *, causing compiler failure.
3721 25. Fixed a file descriptor leak when processing alias/forward files.
3723 26. Fixed a minor format string issue in dbfn.c.
3725 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3727 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3728 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3731 29. When a sender address is verified, it is cached, to save repeating the test
3732 when there is more than one recipient in a message. However, when the
3733 verification involves a callout, it is possible for different callout
3734 options to be set for different recipients. It is too complicated to keep
3735 track of this in the cache, so now Exim always runs a verification when a
3736 callout is required, relying on the callout cache for the optimization.
3737 The overhead is duplication of the address routing, but this should not be
3740 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3741 to be verified with callout=postmaster, and the main callout worked but the
3742 postmaster check failed, the verification correctly failed. However, if a
3743 subsequent RCPT command asked for sender verification *without* the
3744 postmaster check, incorrect caching caused this verification also to fail,
3747 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3748 it was not caching the DNS options (qualify_single, search_parents) that
3749 were used when the lookup failed. A subsequent lookup with different
3750 options therefore always gave the same answer, though there were cases
3751 where it should not have. (Example: a "domains = !$mx_any" option on a
3752 dnslookup router: the "domains" option is always processed without any
3753 widening, but the router might have qualify_single set.) Now Exim uses the
3754 cached value only when the same options are set.
3756 32. Added John Jetmore's "exipick" utility to the distribution.
3758 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3759 than a timeout (e.g. a certificate is required, and is not provided), an
3760 Exim server now closes the connection immediately. Previously it waited for
3761 the client to close - but if the client is SSL, it seems that they each
3762 wait for each other, leading to a delay before one of them times out.
3764 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3765 maintained 0.8.x compatibility because I don't think many are using it, and
3766 it is clearly obsolete.
3768 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3771 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3772 client certificate was expired. A simple patch fixes this, though I don't
3773 understand the full logic of why the verify callback is called multiple
3776 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3777 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3778 which causes problems with some clients (such as the Certicom SSL Plus
3779 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3780 disables the coutermeasure allowing Eudora to connect."
3782 38. Exim was not checking that a write() to a log file succeeded. This could
3783 lead to Bad Things if a log got too big, in particular if it hit a file
3784 size limit. Exim now panics and dies if it cannot write to a log file, just
3785 as it does if it cannot open a log file.
3787 39. Modified OS/Makefile-Linux so that it now contains
3789 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3791 The two -D definitions ensure that Exim is compiled with large file
3792 support, which makes it possible to handle log files that are bigger than
3795 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3796 instance) a domain was checked against a named list that involved a lookup,
3797 causing $domain_data to be set, then another domain was checked against the
3798 same list, then the first domain was re-checked, the value of $domain_data
3799 after the final check could be wrong. In particular, if the second check
3800 failed, it could be set empty. This bug probably also applied to
3803 41. The strip_trailing_dot option was not being applied to the address given
3804 with the -f command-line option.
3806 42. The code for reading a message's header from the spool was incrementing
3807 $received_count, but never initializing it. This meant that the value was
3808 incorrect (doubled) while delivering a message in the same process in which
3809 it was received. In the most common configuration of Exim, this never
3810 happens - a fresh exec is done - but it can happen when
3811 deliver_drop_privilege is set.
3813 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3814 it now includes up to 150 characters of the unexpected data in the log
3817 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3818 and building data strings. The size of both of these buffers was 10 000
3819 bytes - far larger than anybody would *ever* want, thought I. Needless to
3820 say, somebody hit the limit. I have increased the maximum line length to
3821 20 000 and the maximum data length of concatenated lines to 100 000. I have
3822 also fixed two bugs, because there was no checking on these buffers. Tsk,
3823 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3824 buffer is too small.
3826 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3827 lsearch lookups. Now it does.
3829 46. When parsing a route_list item in a manualroute router, a fixed-length
3830 buffer was used for the list of hosts. I made this 1024 bytes long,
3831 thinking that nobody would ever have a list of hosts that long. Wrong.
3832 Somebody had a whole pile of complicated expansion conditions, and the
3833 string was silently truncated, leading to an expansion error. It turns out
3834 that it is easier to change to an unlimited length (owing to other changes
3835 that have happened since this code was originally written) than to build
3836 structure for giving a limitation error. The length of the item that
3837 expands into the list of hosts is now unlimited.
3839 47. The lsearch lookup could not handle data where the length of text line was
3840 more than 4095 characters. Such lines were truncated, leading to shortened
3841 data being returned. It should now handle lines of any length.
3843 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3844 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3847 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3848 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3849 no longer happens in comment lines. A list of replacements is now placed
3850 at the head of all of the source files, except those whose only change is
3851 to replace PERL_COMMAND in the very first #! line.
3853 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3854 messages on the queue, with a bottom-up merge sort, using code contributed
3855 by Michael Haardt. This should make operations like -bp somewhat faster on
3856 large queues. It won't affect queue runners, except when queue_run_in_order
3859 51. Installed eximstats 1.31 in the distribution.
3861 52. Added support for SRV lookups to the dnslookup router.
3863 53. If an ACL referred to $message_body or $message_body_end, the value was not
3864 reset for any messages that followed in the same SMTP session.
3866 54. The store-handling optimization for building very long strings was not
3867 differentiating between the different store pools. I don't think this
3868 actually made any difference in practice, but I've tidied it.
3870 55. While running the routers to verify a sender address, $sender_address
3871 was still set to the sender address. This is wrong, because when routing to
3872 send a bounce to the sender, it would be empty. Therefore, I have changed
3873 it so that, while verifying a sender address, $sender_address is set to <>.
3874 (There is no change to what happens when verifying a recipient address.)
3876 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3877 target A or AAAA records (if not already returned) without resetting the
3878 qualify_single or search_parents options of the DNS resolver. These are
3879 inappropriate in this case because the targets of MX and SRV records must
3880 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3881 target that, when qualified, matched something in the local domain. These
3882 two options are now turned off when doing these lookups.
3884 57. It seems that at least some releases of Reiserfs (which does not have the
3885 concept of a fixed number of inodes) returns zero and not -1 for the
3886 number of available inodes. This interacted badly with check_spool_inodes,
3887 which assumed that -1 was the "no such thing" setting. What I have done is
3888 to check that the total number of inodes is greater than zero before doing
3889 the test of how many are available.
3891 58. When a "warn" ACL statement has a log_message modifier, the message is
3892 remembered, and not repeated. This is to avoid a lot of repetition when a
3893 message has many recipients that cause the same warning to be written.
3894 Howewer, Exim was preserving the list of already written lines for an
3895 entire SMTP session, which doesn't seem right. The memory is now reset if a
3896 new message is started.
3898 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3899 debugging output unless log_rewrite was also set.
3901 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3902 of (int)(handle) when we know that handle contains (void *)(-1).
3904 61. The Exim daemon panic-logs an error return when it closes the incoming
3905 connection. However "connection reset by peer" seems to be common, and
3906 isn't really an error worthy of noting specially, so that particular error
3909 62. When Exim is trying to find all the local interfaces, it used to panic and
3910 die if the ioctl to get the interface flags failed. However, it seems that
3911 on at least one OS (Solaris 9) it is possible to have an interface that is
3912 included in the list of interfaces, but for which you get a failure error
3913 for this call. This happens when the interface is not "plumbed" into a
3914 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3915 failure of the "get flags" call assumes that the interface is down.
3917 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3918 makes life easier for people who are doing arithmetic on fields extracted
3919 from dates, where you often get leading zeros that should not be
3920 interpreted as octal.
3922 64. Added qualify_domain to the redirect router, to override the global
3925 65. If a pathologically long header line contained very many addresses (the
3926 report of this problem mentioned 10 000) and each of them was rewritten,
3927 Exim could use up a very large amount of memory. (It kept on making new
3928 copies of the header line as it rewrote, and never released the old ones.)
3929 At the expense of a bit more processing, the header rewriting function has
3930 been changed so that it no longer eats memory in this way.
3932 66. The generation of the Received: header has been moved from the time that a
3933 message starts to be received, to the time that it finishes. The timestamp
3934 in the Received: header should now be very close to that of the <= log
3935 line. There are two side-effects of this change:
3937 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3938 logged header lines no longer include the local Received: line, because
3939 it has not yet been created. The same applies to a copy of the message
3940 that is returned to a non-SMTP sender when a message is rejected.
3942 (b) When a filter file is tested using -bf, no additional Received: header
3943 is added to the test message. After some thought, I decided that this
3946 This change does not affect the value of $received_for. It is still set
3947 after address rewriting, but before local_scan() is called.
3949 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3951 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3952 gave an unhelpful panic error message, and a defer error. I have managed to
3953 change this behaviour so that it now rejects any supplied certificate,
3954 which seems right, as the list of acceptable certificates is empty.
3956 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3957 gave an unhelpful defer error. I have not managed to make this reject any
3958 supplied certificates, but the error message it gives is "no certificate
3959 supplied", which is not helpful.
3961 70. exigrep's output now also includes lines that are not associated with any
3962 message, but which match the given pattern. Implemented by a patch from
3963 Martin Sluka, which also tidied up the Perl a bit.
3965 71. Recipient callout verification, like sender verification, was using <> in
3966 the MAIL FROM command. This isn't really the right thing, since the actual
3967 sender may affect whether the remote host accepts the recipient or not. I
3968 have changed it to use the actual sender in the callout; this means that
3969 the cache record is now keyed on a recipient/sender pair, not just the
3970 recipient address. There doesn't seem to be a real danger of callout loops,
3971 since a callout by the remote host to check the sender would use <>.
3972 [SEE ABOVE: changed after hitting problems.]
3974 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3975 temporary errors. However, in the case of such a code being given after
3976 the end of a data transmission (i.e. after ".") Exim was failing to write
3977 a retry record for the message. (Yes, there was some broken host that was
3978 actually sending 8xx at this point.)
3980 73. An unknown lookup type in a host list could cause Exim to panic-die when
3981 the list was checked. (An example that provoked this was putting <; in the
3982 middle of a list instead of at the start.) If this happened during a DATA
3983 ACL check, a -D file could be left lying around. This kind of configuration
3984 error no longer causes Exim to die; instead it causes a defer errror. The
3985 incident is still logged to the main and panic logs.
3987 74. Buglet left over from Exim 3 conversion. The message "too many messages
3988 in one connection" was written to the rejectlog but not the mainlog, except
3989 when address rewriting (yes!) was being logged.
3991 75. Added write_rejectlog option.
3993 76. When a system filter was run not as root (that is, when system_filter_user
3994 was set), the values of the $n variables were not being returned to the
3995 main process; thus, they were not subsequently available in the $sn
3998 77. Added +return_path_on_delivery log selector.
4000 78. A connection timeout was being treated differently from recipients deferred
4001 when testing hosts_max_try with a message that was older than the host's
4002 retry timeout. (The host should not be counted, thus allowing all hosts to
4003 be tried at least once before bouncing.) This may have been the cause of an
4004 occasionally reported bug whereby a message would remain on the queue
4005 longer than the retry timeout, but would be bounced if a delivery was
4006 forced. I say "may" because I never totally pinned down the problem;
4007 setting up timeout/retry tests is difficult. See also the next item.
4009 79. The ultimate address timeout was not being applied to errors that involved
4010 a combination of host plus message (for example, a timeout on a MAIL
4011 command). When an address resolved to a number of possible hosts, and they
4012 were not all tried for each delivery (e.g. because of hosts_max_try), a
4013 message could remain on the queue longer than the retry timeout.
4015 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4018 81. Fixed an obscure SMTP outgoing bug which required at least the following
4019 conditions: (a) there was another message waiting for the same server;
4020 (b) the server returned 5xx to all RCPT commands in the first message so
4021 that the message was not completed; (c) the server dropped the connection
4022 or gave a negative response to the RSET that Exim sends to abort the
4023 transaction. The observed case was a dropped connection after DATA that had
4024 been sent in pipelining mode. That is, the server had advertised PIPELINING
4025 but was not implementing it correctly. The effect of the bug was incorrect
4026 behaviour, such as trying another host, and this could lead to a crash.
4032 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4033 and daemon.c were passed as pointers to ints; they should have been
4034 pointers to socklen_t variables (which are typically unsigned ints).
4036 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4039 3. Fixed a really odd bug that affected only the testing scheme; patching a
4040 certain fixed string in the binary changed the value of another string that
4041 happened to be identical to the end of the original first string.
4043 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4044 name", it returns that address as the IP address. On some operating
4045 systems (e.g. Solaris), it also passes back the IP address string as the
4046 "host name". However, on others (e.g. Linux), it passes back an empty
4047 string. Exim wasn't checking for this, and was changing the host name to an
4048 empty string, assuming it had been canonicized.
4050 5. Although rare, it is permitted to have more than one PTR record for a given
4051 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4052 all the names associated with an address, because they do in Solaris.
4053 However, it seems that they do not in Linux for data that comes from the
4054 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4055 I found this out when I moved to a new Linux workstation and tried to run
4056 the Exim test suite.
4058 To get round this problem I have changed the code so that it now does its
4059 own call to the DNS to look up PTR records when searching for a host name.
4060 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4061 addresses that are only in /etc/hosts are still found.
4063 This behaviour is, however, controlled by an option called host_lookup_
4064 order, which defaults to "bydns:byaddr". If people want to use the other
4065 order, or indeed, just use one or the other means of lookup, they can
4066 specify it in this variable.
4068 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4069 some operating systems, this comes back from gethostbyaddr() as an empty
4070 string, and this is what Exim used to test for. However, it seems that in
4071 other systems, "." is yielded. Exim now tests for this case too.
4073 7. The values of check_spool_space and check_log_space are now held internally
4074 as a number of kilobytes instead of an absolute number of bytes. If a
4075 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4076 kilobyte. This means that much larger values can be stored.
4078 8. Exim monitor: an attempt to get the action menu when not actually pointing
4079 at a message produces an empty menu entitled "No message selected". This
4080 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4081 no entries in it ("Shell widget menu has zero width and/or height"). So I
4082 have added a single, blank menu entry in this case.
4084 9. Added ${quote_local_part.
4086 10. MIME decoding is now applied to the contents of Subject: header lines when
4089 11. Now that a reference to $sender_host_address automatically causes a reverse
4090 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4091 host lookup before query-style lookups in lists that might use this
4092 variable. This has therefore been abolished, and the "net-" prefix is no
4093 longer necessary for query-style lookups.
4095 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4096 have been a typo for LFLAGS, so it has been changed.
4098 13. The install script calls Exim with "-C /dev/null" in order to find the
4099 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4100 to be output. Howeve, since Exim outputs its version number before the
4101 error, it didn't break the script. It just looked ugly. I fixed this by
4102 always allowing "-C /dev/null" if the caller is root.
4104 14. Ignore overlarge ACL variable number when reading spool file - insurance
4105 against a later release with more variables having written the file.
4107 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4108 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4109 this, as well as the form without the "IPv6" on the front.
4111 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4112 OS/Makefile-Darwin file.
4114 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4115 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4117 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4118 variable. However, some OS use ints for inodes. I've added cast to long int
4119 to get rid of the compiler warning.
4121 19. I had forgotten to lock out "/../" in configuration file names when
4122 ALT_CONFIG_PREFIX was set.
4124 20. Routers used for verification do not need to specify transports. However,
4125 if such a router generated a host list, and callout was configured, Exim
4126 crashed, because it could not find a port number from the (non-existent)
4127 transport. It now assumes port 25 in this circumstance.
4129 21. Added the -t option to exigrep.
4131 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4132 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4133 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4134 them is set, LOOKUP_LSEARCH is forced.
4136 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4138 24. Added sender and host information to the "rejected by local_scan()" log
4139 line; previously there was no indication of these.
4141 25. Added .include_if_exists.
4143 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4144 Linux. It turns out that not all file systems support this. Apparently some
4145 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4146 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4147 not supported on the file descriptor, is now ignored when Exim is trying to
4148 sync a directory. This applies only to Linux.
4150 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4152 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4153 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4156 29. Added support for maildirsize files from supplied patch (modified a bit).
4158 30. The use of :fail: followed by an empty string could lead Exim to respond to
4159 sender verification failures with (e.g.):
4161 550 Verification failed for <xxx>
4162 550 Sender verify failed
4164 where the first response line was missing the '-' that indicates it is not
4165 the final line of the response.
4167 31. The loop for finding the name of the user that called Exim had a hardwired
4168 limit of 10; it now uses the value of finduser_retries, which is used for
4169 all other user lookups.
4171 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4174 33. Exim was neglecting to zero errno before one call of strtol() when
4175 expanding a string and expecting an integer value. On some systems this
4176 resulted in spurious "integer overflow" errors. Also, it was casting the
4177 result into an int without checking.
4179 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4180 did not work. The code looks as if it has *never* worked, though it appears
4181 to have been documented since at least releast 1.62. I have made it work.
4183 35. The "timeout_DNS" error in retry rules, also documented since at least
4184 1.62, also never worked. As it isn't clear exactly what this means, and
4185 clearly it isn't a major issue, I have abolished the feature by treating it
4186 as "timeout", and writing a warning to the main and panic logs.
4188 36. The display of retry rules for -brt wasn't always showing the error code
4191 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4192 timeout_connect_A, timeout_connect_MX.
4194 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4195 to the empty sender.
4197 39. The daemon was not analysing the content of -oX till after it had closed
4198 stderr and disconnected from the controlling terminal. This meant that any
4199 syntax errors were only noted on the panic log, and the return code from
4200 the command was 0. By re-arranging the code a little, I've made the
4201 decoding happen first, so such errors now appear on stderr, and the return
4202 code is 1. However, the actual setting up of the sockets still happens in
4203 the disconnected process, so errors there are still only recorded on the
4206 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4207 connections (as happens on some IP stacks) was logged at start up time as
4208 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4209 it from "IPv6 and IPv4", which means that two separate sockets are being
4212 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4213 says whether AF_INET or AF_INET6 was passed as an argument.
4215 42. Exiwhat output was messed up when time zones were included in log
4218 43. Exiwhat now gives more information about the daemon's listening ports,
4219 and whether -tls-on-connect was used.
4221 44. The "port" option of the smtp transport is now expanded.
4223 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4224 silently ignored. Now an error message is written to the main and panic
4227 46. There's a new ACL modifier called "logwrite" which writes to a log file
4228 as soon as it is encountered.
4230 47. Added $local_user_uid and $local_user_gid at routing time.
4232 48. Exim crashed when trying to verify a sender address that was being
4235 49. Exim was recognizing only a space character after ".include". It now also
4236 recognizes a tab character.
4238 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4239 extracting the relevant information from the specification. The man page no
4240 longer contains scrambled data for the -d option, and I've added a section
4241 at the front about calling Exim under different names.
4243 51. Added "extra_headers" argument to the "mail" command in filter files.
4245 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4248 53. Installed eximstats 1.29.
4250 54. Added transport_filter_timeout as a generic transport option.
4252 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4253 Cc: header lines. This was required by RFC 822, but it not required by RFC
4256 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4257 incoming messages that did not have them. Now it does so only if the
4258 message originates locally, that is, if there is no associated remote host
4259 address. When Resent- header lines are present, this applies to the Resent-
4260 lines rather than the non-Resent- lines.
4262 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4263 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4265 58. Messages for configuration errors now include the name of the main
4266 configuration file - useful now that there may be more than one file in a
4267 list (.included file names were always shown).
4269 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4270 for those rare installations that do not start the daemon as root or run it
4271 setuid root. I've cut out the call to initgroups() if the daemon is not
4274 60. The Exim user and group can now be bound into the binary as text strings
4275 that are looked up at the start of Exim's processing.
4277 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4279 62. Added $mailstore_basename variable.
4281 63. Installed patch to sieve.c from Michael Haardt.
4283 64. When Exim failed to open the panic log after failing to open the main log,
4284 the original message it was trying to log was written to stderr and debug
4285 output, but if they were not available (the usual case in production), it
4286 was lost. Now it is written to syslog before the two lines that record the
4287 failures to open the logs.
4289 65. Users' Exim filters run in subprocesses under the user's uid. It is
4290 possible for a "deliver" command or an alias in a "personal" command to
4291 provoke an address rewrite. If logging of address rewriting is configured,
4292 this fails because the process is not running as root or exim. There may be
4293 a better way of dealing with this, but for the moment (because 4.30 needs
4294 to be released), I have disabled address rewrite logging when running a
4295 filter in a non-root, non-exim process.
4301 1. The buildconfig auxiliary program wasn't quoting the value set for
4302 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4303 not defined. This bug was masked in 4.22 by the effect that was fixed in
4306 2. Some messages that were rejected after a message id was allocated were
4307 shown as "incomplete" by exigrep. It no longer does this for messages that
4308 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4310 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4311 have allow_domain_literals set, the ID did not get logged in the <= line.
4312 Domain literals are now always recognized in Message-ID: header lines.
4314 4. The first argument for a ${extract expansion item is the key name or field
4315 number. Leading and trailing spaces in this item were not being ignored,
4316 causing some misleading effects.
4318 5. When deliver_drop_privilege was set, single queue runner processes started
4319 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4320 same command in the process it spins off) were not dropping privilege.
4322 6. When the daemon running as "exim" started a queue runner, it always
4323 re-executed Exim in the spun-off process. This is a waste of effort when
4324 deliver_drop_privilege is set. The new process now just calls the
4325 queue-runner function directly.
4331 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4334 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4335 ignored. Though the use of -oP was forcing the writing of a pid file, it
4336 was always written to the default place.
4338 3. If the message "no IP address found for host xxxx" is generated during
4339 incoming verification, it is now followed by identification of the incoming
4340 connection (so you can more easily find what provoked it).
4342 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4344 5. Added some features to "harden" Exim a bit more against certain attacks:
4346 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4347 be put in Local/Makefile. This is like the never_users runtime option,
4348 but it cannot be overridden. The default setting is "root".
4350 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4351 prefix string with which any file named in a -C command line option
4354 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4355 is retained for -C and -D only if the caller of Exim is root. Without
4356 it, the exim user may also use -C and -D and retain privilege.
4358 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4359 command line option is disabled.
4361 6. Macro names set by the -D option must start with an upper case letter, just
4362 like macro names defined in the configuration file.
4364 7. Added "dereference=" facility to LDAP.
4366 8. Two instances of the typo "uknown" in the source files are fixed.
4368 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4369 the Configure-Makefile script screwed up while processing it.
4371 10. Incorporated PCRE 4.4.
4373 11. The SMTP synchronization check was not operating right at the start of an
4374 SMTP session. For example, it could not catch a HELO sent before the client
4375 waited for the greeting. There is now a check for outstanding input at the
4376 point when the greeting is written. Because of the duplex, asynchronous
4377 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4378 way, but not yet received, when the check is performed.
4380 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4381 on TCP/IP sockets, because this apparently causes some broken clients to
4384 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4385 unchanged) from the Cygwin maintainer.
4387 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4388 was supported instead of testing for mbx. Effectively a typo.
4390 15. The spa authenticator server code was not checking that the input it
4391 received was valid base64.
4393 16. The debug output line for the "set" modifier in ACLs was not showing the
4394 name of the variable that was being set.
4396 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4398 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4399 lookup on its own. Something else was needed to trigger the lookup. For
4400 example, a match in host_lookup or the need for a host name in a host list.
4401 Now, if $sender_host_name is referenced and the host name has not yet been
4402 looked up, a lookup is performed. If the lookup fails, the variable remains
4403 empty, and $host_lookup_failed is set to "1".
4405 19. Added "eqi" as a case-independent comparison operator.
4407 20. The saslauthd authentication condition could segfault if neither service
4408 nor realm was specified.
4410 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4411 error message that was logged was misleading, and incoming SMTP
4412 connections were dropped. The message is now more accurate, and temporary
4413 errors are given to SMTP connections.
4415 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4416 (except RSET) with a 421 error, until QUIT is received. However, it was
4417 failing to send a response to QUIT.
4419 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4420 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4421 ACL is now run after the helo_try_verify_hosts code.
4423 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4424 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4425 case-independent, so other case variants are also recognized). Apparently
4426 some systems use these upper case variants.
4428 25. If more than two messages were waiting for the same host, and a transport
4429 filter was specified for the transport, Exim sent two messages over the
4430 same TCP/IP connection, and then failed with "socket operation on non-
4431 socket" when it tried to send the third.
4433 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4435 27. The extern definition of crypt16() in expand.c was not being excluded when
4436 the OS had its own crypt16() function.
4438 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4439 as a preferred synonym for return_size_limit, both as an option and as an
4442 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4444 30. Changed the default configuration ACL to relax the local part checking rule
4445 for addresses that are not in any local domains. For these addresses,
4446 slashes and pipe symbols are allowed within local parts, but the sequence
4447 /../ is explicitly forbidden.
4449 31. SPA server authentication was not clearing the challenge buffer before
4452 32. log_message in a "warn" ACL statement was writing to the reject log as
4453 well as to the main log, which contradicts the documentation and doesn't
4454 seem right (because no rejection is happening). So I have stopped it.
4456 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4457 However, I am unable to do any testing of this.
4459 34. Fixed an infelicity in the appendfile transport. When checking directories
4460 for a mailbox, to see if any needed to be created, it was accidentally
4461 using path names with one or more superfluous leading slashes; tracing
4462 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4464 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4465 contents are added to the log line that is written for every discarded
4466 recipient. (Previously a log_message setting was ignored.)
4468 36. The ${quote: operator now quotes the string if it is empty.
4470 37. The install script runs exim in order to find its version number. If for
4471 some reason other than non-existence or emptiness, which it checks, it
4472 could not run './exim', it was installing it with an empty version number,
4473 i.e. as "exim-". This error state is now caught, and the installation is
4476 38. An argument was missing from the function that creates an error message
4477 when Exim fails to connect to the socket for saslauthd authentication.
4478 This could cause Exim to crash, or give a corrupted message.
4480 39. Added isip, isip4, and isip6 to ${if conditions.
4482 40. The ACL variables $acl_xx are now saved with the message, and can be
4483 accessed later in routers, transports, and filters.
4485 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4486 strings in the file are not string-expanded.
4488 42. If a MAIL command specified a SIZE value that was too large to fit into an
4489 int variable, the check against message_size_limit failed. Such values are
4490 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4491 day this will have to be increased, but I don't think I want to be around
4492 when emails are that large.
4499 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4500 iconv() is not standard in FreeBSD.
4502 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4503 IPv6 enabled. The observed symptom was a segmentation fault on return from
4504 the function os_common_find_running_interfaces() in src/os.c.
4506 3. In the check_special_case() function in daemon.c I had used "errno" as an
4507 argument name, which causes warnings on some systems. This was basically a
4508 typo, since it was named "eno" in the comments!
4510 4. The code that waits for the clock to tick (at a resolution of some fraction
4511 of a second) so as to ensure message-id uniqueness was always waiting for
4512 at least one whole tick, when it could have waited for less. [This is
4513 almost certainly not relevant at current processor speeds, where it is
4514 unlikely to ever wait at all. But we try to future-proof.]
4516 5. The function that sleeps for a time interval that includes fractions of a
4517 second contained a race. It did not block SIGALRM between setting the
4518 timer, and suspending (a couple of lines later). If the interval was short
4519 and the sigsuspend() was delayed until after it had expired, the suspension
4520 never ended. On busy systems this could lead to processes getting stuck for
4523 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4524 process, before it forks any delivery processes. The open lookup caching
4525 mechanism meant that the open file or database connection was passed into
4526 the delivery process. The problem was that delivery processes always tidy
4527 up cached lookup data. This could cause a problem for the next delivery
4528 process started by the queue runner, because the external queue runner
4529 process does not know about the closure. So the next delivery process
4530 still has data in the lookup cache. In the case of a file lookup, there was
4531 no problem because closing a file descriptor in a subprocess doesn't affect
4532 the parent. However, if the lookup was caching a connection to a database,
4533 the connection was closed, and the second delivery process was likely to
4534 see errors such as "PGSQL: query failed: server closed the connection
4535 unexpectedly". The problem has been fixed by closing all cached lookups
4536 in a queue runner before running a delivery process.
4538 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4539 seem to have the "const" qualifier which it has on other OS. I've
4542 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4543 *of the same type* (client or server) with the same public name that an
4544 error should be diagnosed.
4546 9. When Exim looked up a host name for an IP address, but failed to find the
4547 original IP address when looking up the host name (a safety check), it
4548 output the message "<ip address> does not match any IP for NULL", which was
4549 confusing, to say the least. The bug was that the host name should have
4550 appeared instead of "NULL".
4552 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4553 user that is built into the binary, and the -C or -D options is used, root
4554 privilege is dropped before the configuration file is read. In addition,
4555 logging is switched to stderr instead of the normal log files. If the
4556 configuration then re-defines the Exim user, the unprivileged environment
4557 is probably not what is expected, so Exim logs a panic warning message (but
4560 However, if deliver_drop_privilege is set, the unprivileged state may well
4561 be exactly what is intended, so the warning has been cut out in that case,
4562 and Exim is allowed to try to write to its normal log files.
4568 1. smtp_return_error_details was not giving details for temporary sender
4569 or receiver verification errors.
4571 2. Diagnose a configuration error if two authenticators have the same public
4574 3. Exim used not to create the message log file for a message until the first
4575 delivery attempt. This could be confusing when incoming messages were held
4576 for policy or load reasons. The message log file is now created at the time
4577 the message is received, and an initial "Received" line is written to it.
4579 4. The automatically generated man page for command line options had a minor
4580 bug that caused no ill effects; however, a more serious problem was that
4581 the procedure for building the man page automatically didn't always
4582 operate. Consequently, release 4.20 contains an out-of-date version. This
4583 shouldn't happen again.
4585 5. When building Exim with embedded Perl support, the script that builds the
4586 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4587 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4589 6. The freeze_tell option was not being used for messages that were frozen on
4590 arrival, either by an ACL or by local_scan().
4592 7. Added the smtp_incomplete_transaction log selector.
4594 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4595 was accepting AUTH without a new EHLO.
4597 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4598 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4601 10. Logging of TCP/IP connections (when configured) now happens in the main
4602 daemon process instead of the child process, so that the TCP/IP connection
4603 count is more accurate (but it can never be perfect).
4605 11. The use of "drop" in a nested ACL was not being handled correctly in the
4606 outer ACL. Now, if condition failure induced by the nested "drop" causes
4607 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4608 or "require"), the connection is dropped.
4610 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4611 that yield "discard" can now be used with an "accept" or a "discard" verb,
4612 but an error is generated for any others (because I can't see a useful way
4613 to define what should happen).
4615 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4616 are now processed in the same way as lines in the Exim configuration file.
4617 In particular, continuation lines are supported.
4619 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4621 15. Added -ti meaning -t -i.
4623 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4624 domains, and warn by logging if others are found.
4626 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4627 returned from GIFCONF when Exim is trying to find the list of interfaces on
4628 a host. The code in os.c has been modified to copy each ifreq to an aligned
4629 structure in all cases.
4631 Also, in some cases, the returned ifreq's were being copied to a 'struct
4632 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4633 means the last couple of bytes of an IPv6 address could be chopped if the
4634 ifreq contained only a normal sockaddr (14 bytes storage).
4636 18. Named domain lists were not supported in the hosts_treat_as_local option.
4637 An entry such as +xxxx was not recognized, and was treated as a literal
4640 19. Ensure that header lines added by a DATA ACL are included in the reject log
4641 if the ACL subsequently rejects the message.
4643 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4644 MD5 (which is deprecated).
4646 21. When testing a filter file using -bf, Exim was writing a message when it
4647 took the sender from a "From " line in the message, but it was not doing so
4648 when it took $return_path from a Return-Path: header line. It now does.
4650 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4651 with a valid header line field name (a series of printing characters
4652 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4654 23. Changed "disc" in the source to "disk" to conform to the documentation and
4655 the book and for uniformity.
4657 24. Ignore Sendmail's -Ooption=value command line item.
4659 25. When execve() failed while trying to run a command in a pipe transport,
4660 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4661 could be confused with a return value of 69 from the command itself. This
4662 has been changed to 127, the value the shell returns if it is asked to run
4663 a non-existent command. The wording for the related log line suggests a
4664 non-existent command as the problem.
4666 26. If received_header_text expands to an empty string, do not add a Received:
4667 header line to the message. (Well, it adds a token one on the spool, but
4668 marks it "old" so that it doesn't get used or transmitted.)
4670 27. Installed eximstats 1.28 (addition of -nt option).
4672 28. There was no check for failure on the call to getsockname() in the daemon
4673 code. This can fail if there is a shortage of resources on the system, with
4674 ENOMEM, for example. A temporary error is now given on failure.
4676 29. Contrary to the C standard, it seems that in some environments, the
4677 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4678 program. Exim now does this explicitly; it affects the formatting of
4679 timestamps using strftime().
4681 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4682 complaints. I've now initialized all the variables, to avoid this.
4684 32. Header lines added by a system filter were not being "seen" during
4685 transport-time rewrites.
4687 33. The info_callback() function passed to OpenSSL is set up with type void
4688 (*)(SSL *, int, int), as described somewhere. However, when calling the
4689 function (actually a macro) that sets it up, the type void(*)() is
4690 expected. I've put in a cast to prevent warnings from picky compilers.
4692 34. If a DNS black list lookup found a CNAME record, but there were no A
4693 records associated with the domain it pointed at, Exim crashed.
4695 35. If a DNS black list lookup returned more than one A record, Exim ignored
4696 all but the first. It now scans all returned addresses if a particular IP
4697 value is being sought. In this situation, the contents of the
4698 $dnslist_value variable are a list of all the addresses, separated by a
4701 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4702 to accept a host name and all its aliases if the forward lookup for any of
4703 them yielded the IP address of the incoming connection. Now it accepts only
4704 those names whose forward lookup yields the correct IP address. Any other
4705 names are discarded. This closes a loophole whereby a rogue DNS
4706 administrator could create reverse DNS records to break through a
4707 wildcarded host restriction in an ACL.
4709 37. If a user filter or a system filter that ran in a subprocess used any of
4710 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4711 the wrong values were passed to the pipe command ($thisaddress had the
4712 value of $0, $0 had the value of $1, etc). This bug was introduced by
4713 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4715 38. Improved the line breaking for long SMTP error messages from ACLs.
4716 Previously, if there was no break point between 40 and 75 characters, Exim
4717 left the rest of the message alone. Two changes have been made: (a) I've
4718 reduced the minimum length to 35 characters; (b) if it can't find a break
4719 point between 35 and 75 characters, it looks ahead and uses the first one
4720 that it finds. This may give the occasional overlong line, but at least the
4721 remaining text gets split now.
4723 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4724 file descriptors might be low, and that setting 1000 would always raise it.
4725 It turns out that in some environments, the limit is already over 1000 and
4726 that lowering it causes trouble. So now Exim takes care not to decrease it.
4728 40. When delivering a message, the value of $return_path is set to $sender_
4729 address at the start of routing (routers may change the value). By an
4730 oversight, this default was not being set up when an address was tested by
4731 -bt or -bv, which affected the outcome if any router or filter referred to
4734 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4735 log only when "message" or "log_message" are set. However, if one of the
4736 conditions was an address verification, or a call to a nested ACL, the
4737 messages generated by the underlying test were being passed through. This
4738 no longer happens. The underlying message is available in $acl_verify_
4739 message for both "message" and "log_message" expansions, so it can be
4740 passed through if needed.
4742 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4743 new expansion $bh_ to give the encoded byte string without charset
4744 translation. Translation happens only if iconv() is available; HAVE_ICONV
4745 indicates this at build time. HEADERS_CHARSET gives the charset to
4746 translate to; headers_charset can change it in the configuration, and
4747 "headers charset" can change it in an individual filter file.
4749 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4750 that creates RFC 2047 encoded "words" labels them as that charset instead
4751 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4752 expansion operator; (ii) when Exim creates a From: line for a local
4753 message; (iii) when a header line is rewritten to include a "phrase" part.
4755 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4756 buggy, causing it to skip the first lines of messages whose message ID
4757 ended in 'D'. This would not have bitten before Exim release 4.14, because
4758 message IDs were unlikely to end in 'D' before then. The effect was to have
4759 incorrect size information for certain domains.
4761 45. #include "config.h" was missing at the start of the crypt16.c module. This
4762 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4765 46. If there was a timeout during a "random" callout check, Exim treated it as
4766 a failure of the random address, and carried on sending RSET and the real
4767 address. If the delay was just some slowness somewhere, the response to the
4768 original RCPT would be taken as a response to RSET and so on, causing
4769 mayhem of various kinds.
4771 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4772 when I implemented it. It didn't allow for the fact that some option values
4773 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4774 the right test for positive values.
4776 48. Domain names in DNS records are case-independent. Exim always looks them up
4777 in lower case. Some resolvers return domain names in exactly the case they
4778 appear in the zone file, that is, they may contain uppercase letters. Not
4779 all resolvers do this - some return always lower case. Exim was treating a
4780 change of case by a resolver as a change of domain, similar to a widening
4781 of a domain abbreviation. This triggered its re-routing code and so it was
4782 trying to route what was effectively the same domain again. This normally
4783 caused routing to fail (because the router wouldn't handle the domain
4784 twice). Now Exim checks for this case specially, and just changes the
4785 casing of the domain that it ultimately uses when it transmits the message
4788 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4791 50. If a filter generated a file delivery with a non-absolute name (possible if
4792 no home directory exists for the router), the forbid_file option was not
4795 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4796 the existing equality matching.
4798 52. Exim was using ints instead of ino_t variables in some places where it was
4799 dealing with inode numbers.
4801 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4802 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4803 called TMPDIR, and if it finds it is different, it changes its value.
4805 54. The smtp_printf() function is now made available to local_scan() so
4806 additional output lines can be written before returning. There is also an
4807 smtp_fflush() function to enable the detection of a dropped connection.
4808 The variables smtp_input and smtp_batched_input are exported to
4811 55. Changed the default runtime configuration: the message "Unknown user"
4812 has been removed from the ACL, and instead placed on the localuser router,
4813 using the cannot_route_message feature. This means that any verification
4814 failures that generate their own messages won't get overridden. Similarly,
4815 the "Unrouteable address" message that was in the ACL for unverifiable
4816 relay addresses has also been removed.
4818 56. Added hosts_avoid_esmtp to the smtp transport.
4820 57. The exicyclog script was not checking for the esoteric option
4821 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4822 will work only if exicyclog is run under the appropriate euid.
4824 58. Following a discussion on the list, the rules by which Exim recognises line
4825 endings on incoming messages have been changed. The -dropcr and drop_cr
4826 options are now no-ops, retained only for backwards compatibility. The
4827 following line terminators are recognized: LF CRLF CR. However, special
4828 processing applies to CR:
4830 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4831 nor a local message in the state where . is a terminator.
4833 (ii) If a bare CR is encountered in a header line, an extra space is added
4834 after the line terminator so as not to end the header. The reasoning
4835 behind this is that bare CRs in header lines are most likely either
4836 to be mistakes, or people trying to play silly games.
4838 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4839 was being incorrectly given as 18 bytes larger than it should have been.
4840 This is a VOB (very old bug).
4842 60. This may never have affected anything current, but just in case it has:
4843 When the local host is found other than at the start of a list of hosts,
4844 the local host, those with the same MX, and any that follow, are discarded.
4845 When the list in question was part of a longer list of hosts, the following
4846 hosts (not currently being processed) were also being discarded. This no
4847 longer happens. I'm not sure if this situation could ever has previously
4850 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4853 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4854 header. This is something that is recommended in a new Internet Draft, and
4855 is something that is documented as being done by Sendmail. There are two
4856 possible values. For messages generated by the autoreply transport, Exim
4859 Auto-Submitted: auto-replied
4861 whereas for all other generated messages (e.g. bounces) it adds
4863 Auto-Submitted: auto-generated
4865 63. The "personal" condition in filters now includes a test for the
4866 Auto-Submitted: header. If it contains the string "auto-" the message it
4867 not considered personal.
4869 64. Added rcpt_include_affixes as a generic transport option.
4871 65. Added queue_only_override (default true).
4873 66. Added the syslog_duplication option.
4875 67. If what should have been the first header line of a message consisted of
4876 a space followed by a colon, Exim was mis-interpreting it as a header line.
4877 It isn't of course - it is syntactically invalid and should therefore be
4878 treated as the start of the message body. The misbehaviour could have
4879 caused a number of strange effects, including loss of data in subsequent
4880 header lines, and spool format errors.
4882 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4883 client host had authenticated. This control can now be exercised by an ACL
4884 for more flexibility.
4886 69. By default, callouts do not happen when testing with -bh. There is now a
4887 variant, -bhc, which does actually run the callout code, including
4888 consulting and updating the callout cache.
4890 70. Added support for saslauthd authentication, courtesy of Alexander
4893 71. If statvfs() failed on the spool or log directories while checking their
4894 size for availability, Exim confusingly gave the error "space shortage".
4895 Furthermore, in debugging mode it crashed with a floating point exception.
4896 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4897 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4898 serious problem, Exim now writes to the main and panic logs when this
4899 happens, with details of the failure. It then refuses to accept the
4900 incoming message, giving the message "spool directory problem" or "log
4901 directory problem" with a 421 code for SMTP messages.
4903 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4904 0, 1, and 2 exist, because some OS complain for execs without them (see
4905 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4906 descriptors. However, the code omitted to check that the open succeeded,
4907 causing mysterious errors if for some reason the permissions on /dev/null
4908 got screwed. Now Exim writes a message to the main and panic logs, and
4909 bombs out if it can't open /dev/null.
4911 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4912 interact so that it is all more flexible. It is supposed to remain
4913 backwards compatible. Also added extra_local_interfaces.
4915 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4916 to bomb out with an assertion failure - to the client this appears as a
4917 connection drop. This problem occurs in the part of the code that was taken
4918 from the Samba project. Fortunately, the assertion is in a very simple
4919 function, so I have fixed this by reproducing the function inline in the
4920 one place where it is called, and arranging for authentication to fail
4921 instead of killing the process with assert().
4923 75. The SPA client code was not working when the server requested OEM rather
4924 than Unicode encoding.
4926 76. Added code to make require_files with a specific uid setting more usable in
4927 the case where statting the file as root fails - usually a non-root-mounted
4928 NFS file system. When this happens and the failure is EACCES, Exim now
4929 forks a subprocess and does the per-uid checking as the relevant uid.
4931 77. Added process_log_path.
4933 78. If log_file_path was not explicitly set, a setting of check_log_space or
4934 check_log_inodes was ignored.
4936 79. If a space check for the spool or log partitions fails, the incident is now
4937 logged. Of course, in the latter case the data may get lost...
4939 80. Added the %p formatting code to string_format() so that it can be used to
4940 print addresses in debug_print(). Adjusted all the address printing in the
4941 debugging in store.c to use %p rather than %d.
4943 81. There was a concern that a line of code in smtp_in.c could overflow a
4944 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4945 initially expressed, the concern was not well-founded, because trailing
4946 spaces are removed early. However, if the trailing spaces were followed by
4947 a NULL, they did not get removed, so the overflow was possible. Two fixes
4950 (a) I re-wrote the offending code in a cleaner fashion.
4951 (b) If an incoming SMTP command contains a NULL character, it is rejected
4954 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4955 runs initgroups(), so that if the Exim user is in any additional groups,
4956 they will be used during message reception.
4962 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob