1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.565 2009/07/02 11:44:01 nm4 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TK/01 Added patch by Johannes Berg that expands the main option
10 "spamd_servers" if it starts with a dollar sign.
12 TK/02 Write list of recipients to X-Envelope-Sender header when building
13 the mbox-format spool file for content scanning (suggested by Jakob
16 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
17 (http://www.dcc-servers.net/) support via dccifd. Activated by
18 setting EXPERIMENTAL_DCC=yes in Local/Makefile. Check out
19 experimental_spec.txt for more documentation.
21 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
22 by Mark Daniel Reidel <mr@df.eu>.
24 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
25 When building exim an external PCRE library is now needed -
26 PCRE is a system library on the majority of modern systems.
27 See entry on PCRE_LIBS in EDITME file.
29 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
30 conversation. Added nologin parameter to request.
31 Patch contributed by Kirill Miazine
33 TF/01 Do not log submission mode rewrites if they do not change the address.
35 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
37 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
38 log files in place. Contributed by Roberto Lima
40 NM/04 Bugzilla 667: close socket used by dovecot authenticator
42 TF/03 Bugzilla 615: When checking the local_parts router precondition
43 after a local_part_suffix or local_part_prefix option, Exim now
44 does not use the address's named list lookup cache, since this
45 contains cached lookups for the whole local part.
47 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
48 Robert Millan. Documentation is in experimental-spec.txt
50 TF/04 Bugzilla 668: Fix parallel build (make -j).
52 NM/05 Bugzilla 437: Prevent Maildix aux files being created with mode 000
54 NM/05 Bugzilla 598: Improvement to Dovecot authenticator handling.
55 Patch provided by Jan Srzednicki
57 TF/05 Leading white space used to be stripped from $spam_report which
58 wrecked the formatting. Now it is preserved.
60 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
61 that they are available at delivery time.
63 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
65 TF/08 TLS error reporting now respects the incoming_interface and
66 incoming_port log selectors.
68 TF/09 Produce a more useful error message if an SMTP transport's hosts
69 setting expands to an empty string.
71 NM/06 Bugzilla 744: EXPN did not work under TLS.
72 Patch provided by Phil Pennock
74 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
75 Patch provided by Richard Godbee
77 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
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/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
100 Patch provided by Graeme Fowler
102 NM/15 Bugzilla 851: Documentation example syntax fix.
108 TK/01 Add preliminary DKIM support. Currently requires a forked version of
109 ALT-N's libdkim that I have put here:
110 http://duncanthrax.net/exim-experimental/
112 Note to Michael Haardt: I had to rename some vars in sieve.c. They
113 were called 'true' and it seems that C99 defines that as a reserved
114 keyword to be used with 'bool' variable types. That means you could
115 not include C99-style headers which use bools without triggering
116 build errors in sieve.c.
118 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
119 as mailq or other aliases. Changed the --help handling significantly
120 to do whats expected. exim_usage() emits usage/help information.
122 SC/01 Added the -bylocaldomain option to eximstats.
124 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
126 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
128 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
134 PH/01 Another patch from the Sieve maintainer.
136 PH/02 When an IPv6 address is converted to a string for single-key lookup
137 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
138 dots are used instead of colons so that keys in lsearch files need not
139 contain colons. This was done some time before quoting was made available
140 in lsearch files. However, iplsearch files do require colons in IPv6 keys
141 (notated using the quote facility) so as to distinguish them from IPv4
142 keys. This meant that lookups for IP addresses in host lists did not work
143 for iplsearch lookups.
145 This has been fixed by arranging for IPv6 addresses to be expressed with
146 colons if the lookup type is iplsearch. This is not incompatible, because
147 previously such lookups could never work.
149 The situation is now rather anomolous, since one *can* have colons in
150 ordinary lsearch keys. However, making the change in all cases is
151 incompatible and would probably break a number of configurations.
153 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
156 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
157 conversion specification without a maximum field width, thereby enabling
158 a rogue spamd server to cause a buffer overflow. While nobody in their
159 right mind would setup Exim to query an untrusted spamd server, an
160 attacker that gains access to a server running spamd could potentially
161 exploit this vulnerability to run arbitrary code as the Exim user.
163 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
164 $primary_hostname instead of what libspf2 thinks the hosts name is.
166 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
167 a directory entry by the name of the lookup key. Previously, if a
168 symlink pointed to a non-existing file or a file in a directory that
169 Exim lacked permissions to read, a lookup for a key matching that
170 symlink would fail. Now it is enough that a matching directory entry
171 exists, symlink or not. (Bugzilla 503.)
173 PH/03 The body_linecount and body_zerocount variables are now exported in the
176 PH/04 Added the $dnslist_matched variable.
178 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
179 This means they are set thereafter only if the connection becomes
182 PH/06 Added the client_condition to authenticators so that some can be skipped
183 by clients under certain conditions.
185 PH/07 The error message for a badly-placed control=no_multiline_responses left
186 "_responses" off the end of the name.
188 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
190 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
191 (without spaces) instead of just copying the configuration text.
193 PH/10 Added the /noupdate option to the ratelimit ACL condition.
195 PH/11 Added $max_received_linelength.
197 PH/12 Added +ignore_defer and +include_defer to host lists.
199 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
200 way in which PCRE > 7.0 is built.
202 PH/14 Implemented queue_only_load_latch.
204 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
205 MAIL command. The effect was to mangle the value on 64-bit systems.
207 PH/16 Another patch from the Sieve maintainer.
209 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
211 PH/18 If a system quota error occurred while trying to create the file for
212 a maildir delivery, the message "Mailbox is full" was not appended to the
213 bounce if the delivery eventually timed out. Change 4.67/27 below applied
214 only to a quota excession during the actual writing of the file.
216 PH/19 It seems that peer DN values may contain newlines (and other non-printing
217 characters?) which causes problems in log lines. The DN values are now
218 passed through string_printing() before being added to log lines.
220 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
221 and InterBase are left for another time.)
223 PH/21 Added message_body_newlines option.
225 PH/22 Guard against possible overflow in moan_check_errorcopy().
227 PH/23 POSIX allows open() to be a macro; guard against that.
229 PH/24 If the recipient of an error message contained an @ in the local part
230 (suitably quoted, of course), incorrect values were put in $domain and
231 $local_part during the evaluation of errors_copy.
237 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
238 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
241 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
242 issue a MAIL command.
244 PH/02 In an ACL statement such as
246 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
248 if a client was not listed at all, or was listed with a value other than
249 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
250 the condition was not true (as it should be), so access was not denied.
251 The bug was that the ! inversion was incorrectly passed on to the second
252 item. This has been fixed.
254 PH/03 Added additional dnslists conditions == and =& which are different from
255 = and & when the dns lookup returns more than one IP address.
257 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
258 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
260 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
261 FSYNC, which compiles an option called disable_fsync that allows for
262 bypassing fsync(). The documentation is heavily laced with warnings.
264 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
266 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
267 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
268 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
269 including adding "make clean"; (3) Added -fPIC when compiling the test
270 dynamically loaded module, to get rid of a warning.
272 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
273 message fails, move_frozen_messages = true and ignore_bounce_errors_after
274 = 0s. The bug is otherwise harmless.
276 PH/07 There was a bug in the dovecot authenticator such that the value of
277 $auth1 could be overwritten, and so not correctly preserved, after a
278 successful authentication. This usually meant that the value preserved by
279 the server_setid option was incorrect.
281 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
283 PH/09 Installed PCRE release 7.0.
285 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
286 run for batched SMTP input. It is now run at the start of every message
287 in the batch. While fixing this I discovered that the process information
288 (output by running exiwhat) was not always getting set for -bs and -bS
289 input. This is fixed, and it now also says "batched" for BSMTP.
291 PH/11 Added control=no_pipelining.
293 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
294 patch, slightly modified), and move the expansion of helo_data till after
295 the connection is made in the smtp transport (so it can use these
298 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
300 PH/14 Added log_selector = +pid.
302 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
304 PH/16 Add ${if forany and ${if forall.
306 PH/17 Added dsn_from option to vary the From: line in DSNs.
308 PH/18 Flush SMTP output before performing a callout, unless control =
309 no_callout_flush is set.
311 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
312 was true (the default) a successful delivery failed to delete the retry
313 item, thus causing premature timeout of the address. The bug is now
316 PH/20 Added hosts_avoid_pipelining to the smtp transport.
318 PH/21 Long custom messages for fakedefer and fakereject are now split up
319 into multiline reponses in the same way that messages for "deny" and
320 other ACL rejections are.
322 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
323 with slight modification.
325 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
326 draft, changing the syntax and factoring some duplicate code".
328 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
329 for deliveries of the second and subsequent messages over the same SMTP
332 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
333 ${reduce, with only minor "tidies".
335 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
337 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
338 expansion side effects.
340 PH/27 When a message times out after an over-quota error from an Exim-imposed
341 quota, the bounce message says "mailbox is full". This message was not
342 being given when it was a system quota that was exceeded. It now should
345 MH/03 Made $recipients available in local_scan(). local_scan() already has
346 better access to the recipient list through recipients_list[], but
347 $recipients can be useful in postmaster-provided expansion strings.
349 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
350 in the case of a MAIL command with additional options following the
351 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
352 were accidentally chopped off.
354 PH/29 SMTP synchronization checks are implemented when a command is read -
355 there is a check that no more input is waiting when there shouldn't be
356 any. However, for some commands, a delay in an ACL can mean that it is
357 some time before the response is written. In this time, more input might
358 arrive, invalidly. So now there are extra checks after an ACL has run for
359 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
360 pipelining has not been advertised.
362 PH/30 MH's patch to allow iscntrl() characters to be list separators.
364 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
365 returned in the SMTP response when smtp_return_error_details was false.
368 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
369 instead of the C I/O that was originally supplied, because problems were
372 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
373 Exim which did not show up earlier: it was assuming that a call to
374 SSL_CTX_set_info_callback() might give an error value. In fact, there is
375 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
376 was a macro that became an assignment, so it seemed to work. This has
377 changed to a proper function call with a void return, hence the compile
378 error. Exim's code has been fixed.
380 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
383 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
385 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
387 PH/37 If a message is not accepted after it has had an id assigned (e.g.
388 because it turns out to be too big or there is a timeout) there is no
389 "Completed" line in the log. When some messages of this type were
390 selected by exigrep, they were listed as "not completed". Others were
391 picked up by some special patterns. I have improved the selection
392 criteria to be more general.
394 PH/38 The host_find_failed option in the manualroute router can now be set
395 to "ignore", to completely ignore a host whose IP address cannot be
396 found. If all hosts are ignored, the behaviour is controlled by the new
397 host_all_ignored option.
399 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
400 homing or because of multiple MX records with /mx) generated more than
401 one IP address, and the following item turned out to be the local host,
402 all the secondary addresses of the first item were incorrectly removed
403 from the list, along with the local host and any following hosts (which
404 is what is supposed to happen).
406 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
407 whoever called Exim into the -H file. In the case of the daemon it was
408 behaving confusingly. When first started, it used values for whoever
409 started the daemon, but after a SIGHUP it used the Exim user (because it
410 calls itself on a restart). I have changed the code so that it now always
413 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
414 message are rejected with the same error (e.g. no authentication or bad
415 sender address), and a DATA command is nevertheless sent (as can happen
416 with PIPELINING or a stupid MUA), the error message that was given to the
417 RCPT commands is included in the rejection of the DATA command. This is
418 intended to be helpful for MUAs that show only the final error to their
421 PH/42 Another patch from the Sieve maintainer.
423 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
424 Eximstats - Fixed some broken HTML links and added missing column headers
426 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
427 columns for Rejects, Temp Rejects, Ham, and Spam rows.
429 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
431 PH/43 Yet another patch from the Sieve maintainer.
433 PH/44 I found a way to check for a TCP/IP connection going away before sending
434 the response to the final '.' that terminates a message, but only in the
435 case where the client has not sent further data following the '.'
436 (unfortunately, this is allowed). However, in many cases there won't be
437 any further data because there won't be any more messages to send. A call
438 to select() can be used: if it shows that the input is "ready", there is
439 either input waiting, or the socket has been closed. An attempt to read
440 the next input character can distinguish the two cases. Previously, Exim
441 would have sent an OK response which the client would never have see.
442 This could lead to message repetition. This fix should cure that, at
443 least in a lot of common cases.
445 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
446 advertised in response to EHLO.
452 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
453 fixed by 4.65/MH/01 (is this a record?) are fixed:
455 (i) An empty string was always treated as zero by the numeric comparison
456 operators. This behaviour has been restored.
458 (ii) It is documented that the numeric comparison operators always treat
459 their arguments as decimal numbers. This was broken in that numbers
460 starting with 0 were being interpreted as octal.
462 While fixing these problems I realized that there was another issue that
463 hadn't been noticed. Values of message_size_limit (both the global option
464 and the transport option) were treated as octal if they started with 0.
465 The documentation was vague. These values are now always treated as
466 decimal, and I will make that clear in the documentation.
472 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
473 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
476 MH/01 Don't check that the operands of numeric comparison operators are
477 integers when their expansion is in "skipping" mode (fixes bug
478 introduced by 4.64-PH/07).
480 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
481 child addresses, Exim now panics and dies. Previously, because the count
482 is held in a short int, deliveries were likely to be lost. As such a
483 large number of recipients for a single message is ridiculous
484 (performance will be very, very poor), I have chosen to impose a limit
485 rather than extend the field.
491 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
492 leftover -K file (the existence of which was triggered by #402).
493 While we were at it, introduced process PID as part of the -K
494 filename. This should rule out race conditions when creating
497 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
498 processing considerably. Previous code took too long for large mails,
499 triggering a timeout which in turn triggers #401.
501 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
502 in the DK code in transports.c. sendfile() is not really portable,
503 hence the _LINUX specificness.
505 TF/01 In the add_headers option to the mail command in an Exim filter,
506 there was a bug that Exim would claim a syntax error in any
507 header after the first one which had an odd number of characters
510 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
511 callout verification, Exim cached a "reject" for the entire domain. This
512 is correct for most verifications, but it is not correct for a recipient
513 verification with use_sender or use_postmaster set, because in that case
514 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
515 case of MAIL FROM:<> rejection from other early rejections (e.g.
516 rejection of HELO). When verifying a recipient using a non-null MAIL
517 address, the cache is ignored if it shows MAIL FROM:<> rejection.
518 Whatever the result of the callout, the value of the domain cache is
519 left unchanged (for any other kind of callout, getting as far as trying
520 RCPT means that the domain itself is ok).
522 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
525 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
526 manifest itself as EPIPE rather than ECONNECT. When tidying away a
527 session, the daemon ignores ECONNECT errors and logs others; it now
528 ignores EPIPE as well.
530 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
531 (quoted-printable decoding).
533 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
534 later the small subsequent patch to fix an introduced bug.
536 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
538 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
540 PH/08 An error is now given if message_size_limit is specified negative.
542 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
543 to be given (somewhat) arbitrary names.
545 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
548 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
549 miscellaneous code fixes
551 PH/10 Added the log_reject_target ACL modifier to specify where to log
554 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
555 hostname. This is wrong, because it relates to the incoming message (and
556 probably the interface on which it is arriving) and not to the outgoing
557 callout (which could be using a different interface). This has been
558 changed to use the value of the helo_data option from the smtp transport
559 instead - this is what is used when a message is actually being sent. If
560 there is no remote transport (possible with a router that sets up host
561 addresses), $smtp_active_hostname is used.
563 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
564 tweaks were necessary in order to get it to work (see also 21 below):
565 (a) The code assumed that strncpy() returns a negative number on buffer
566 overflow, which isn't the case. Replaced with Exim's string_format()
568 (b) There were several signed/unsigned issues. I just did the minimum
569 hacking in of casts. There is scope for a larger refactoring.
570 (c) The code used strcasecmp() which is not a standard C function.
571 Replaced with Exim's strcmpic() function.
572 (d) The code set only $1; it now sets $auth1 as well.
573 (e) A simple test gave the error "authentication client didn't specify
574 service in request". It would seem that Dovecot has changed its
575 interface. Fortunately there's a specification; I followed it and
576 changed what the client sends and it appears to be working now.
578 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
581 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
582 address A is aliased to B and C, where B exists and C does not. Without
583 -v the output is "A verified" because verification stops after a
584 successful redirection if more than one address is generated. However,
585 with -v the child addresses are also verified. Exim was outputting "A
586 failed to verify" and then showing the successful verification for C,
587 with its parentage. It now outputs "B failed to verify", showing B's
588 parentage before showing the successful verification of C.
590 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
591 look up a TXT record in a specific list after matching in a combined
594 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
595 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
596 they consult the DNS. I had assumed they would set it the way they
597 wanted; and indeed my experiments on Linux seem to show that in some
598 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
599 To be on the safe side, however, I have now made the interface to
600 host_find_byname() similar to host_find_bydns(), with an argument
601 containing the DNS resolver options. The host_find_byname() function now
602 sets these options at its start, just as host_find_bydns() does. The smtp
603 transport options dns_qualify_single and dns_search_parents are passed to
604 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
605 of host_find_byname() use the default settings of RES_DEFNAMES
606 (qualify_single) but not RES_DNSRCH (search_parents).
608 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
609 spool_read_header() do less string testing, by means of a preliminary
610 switch on the second character of optional "-foo" lines. (This is
611 overdue, caused by the large number of possibilities that now exist.
612 Originally there were few.) While I was there, I also converted the
613 str(n)cmp tests so they don't re-test the leading "-" and the first
614 character, in the hope this might squeeze out yet more improvement.
616 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
617 flag allowing group syntax was set by the header_syntax check but not
618 turned off, possible causing trouble later; (2) The flag was not being
619 set at all for the header_verify test, causing "group"-style headers to
620 be rejected. I have now set it in this case, and also caused header_
621 verify to ignore an empty address taken from a group. While doing this, I
622 came across some other cases where the code for allowing group syntax
623 while scanning a header line wasn't quite right (mostly, not resetting
624 the flag correctly in the right place). These bugs could have caused
625 trouble for malformed header lines. I hope it is now all correct.
627 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
628 with the "reply" argument non-NULL. The code, however (which originally
629 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
630 but it didn't always do it. This confused somebody who was copying the
631 code for some other use. I have removed all the tests.
633 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
634 feature that was used to support insecure browsers during the U.S. crypto
635 embargo. It requires special client support, and Exim is probably the
636 only MTA that supported it -- and would never use it because real RSA is
637 always available. This code has been removed, because it had the bad
638 effect of slowing Exim down by computing (never used) parameters for the
639 RSA_EXPORT functionality.
641 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
642 authenticator to fail if there's a tab character in the incoming data
643 (there should never be unless someone is messing about, as it's supposed
644 to be base64-encoded). Also added, on Timo's advice, the "secured" option
645 if the connection is using TLS or if the remote IP is the same as the
646 local IP, and the "valid-client-cert option" if a client certificate has
649 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
650 authenticators. This can be used for authorization after authentication
651 succeeds. (In the case of plaintext, it servers for both authentication
654 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
655 if any retry times were supplied.
657 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
658 connection, where there is no client IP address to check. In this
659 situation, the verify now always succeeds.
661 PH/25 Applied John Jetmore's -Mset patch.
663 PH/26 Added -bem to be like -Mset, but loading a message from a file.
665 PH/27 In a string expansion for a processed (not raw) header when multiple
666 headers of the same name were present, leading whitespace was being
667 removed from all of them, but trailing whitespace was being removed only
668 from the last one. Now trailing whitespace is removed from each header
669 before concatenation. Completely empty headers in a concatenation (as
672 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
673 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
675 PH/29 [Removed. This was a change that I later backed out, and forgot to
676 correct the ChangeLog entry (that I had efficiently created) before
677 committing the later change.]
679 PH/30 Exim was sometimes attempting to deliver messages that had suffered
680 address errors (4xx response to RCPT) over the same connection as other
681 messages routed to the same hosts. Such deliveries are always "forced",
682 so retry times are not inspected. This resulted in far too many retries
683 for the affected addresses. The effect occurred only when there were more
684 hosts than the hosts_max_try setting in the smtp transport when it had
685 the 4xx errors. Those hosts that it had tried were not added to the list
686 of hosts for which the message was waiting, so if all were tried, there
687 was no problem. Two fixes have been applied:
689 (i) If there are any address or message errors in an SMTP delivery, none
690 of the hosts (tried or untried) are now added to the list of hosts
691 for which the message is waiting, so the message should not be a
692 candidate for sending over the same connection that was used for a
693 successful delivery of some other message. This seems entirely
694 reasonable: after all the message is NOT "waiting for some host".
695 This is so "obvious" that I'm not sure why it wasn't done
696 previously. Hope I haven't missed anything, but it can't do any
697 harm, as the worst effect is to miss an optimization.
699 (ii) If, despite (i), such a delivery is accidentally attempted, the
700 routing retry time is respected, so at least it doesn't keep
701 hammering the server.
703 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
704 in ${readsocket because some servers need this prod.
706 PH/32 Added some extra debug output when updating a wait-xxx database.
708 PH/33 The hint "could be header name not terminated by colon", which has been
709 given for certain expansion errors for a long time, was not being given
710 for the ${if def:h_colon_omitted{... case.
712 PH/34 The spec says: "With one important exception, whenever a domain list is
713 being scanned, $domain contains the subject domain." There was at least
714 one case where this was not true.
716 PH/35 The error "getsockname() failed: connection reset by peer" was being
717 written to the panic log as well as the main log, but it isn't really
718 panic-worthy as it just means the connection died rather early on. I have
719 removed the panic log writing for the ECONNRESET error when getsockname()
722 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
723 runs only) independently of the message's sender address. This meant
724 that, if the 4xx error was in fact related to the sender, a different
725 message to the same recipient with a different sender could confuse
726 things. In particualar, this can happen when sending to a greylisting
727 server, but other circumstances could also provoke similar problems.
728 I have changed the default so that the retry time for these errors is now
729 based a combination of the sender and recipient addresses. This change
730 can be overridden by setting address_retry_include_sender=false in the
733 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
734 remote server are returned as part of bounce messages. This was not
735 happening for LMTP over a pipe (the lmtp transport), but now it is the
736 same for both kinds of LMTP.
738 PH/38 Despite being documented as not happening, Exim was rewriting addresses
739 in header lines that were in fact CNAMEs. This is no longer the case.
741 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
742 and queue runs started by the daemon processed all messages. This has
743 been fixed so that -R and -S can now usefully be given with -q<time>.
745 PH/40 Import PCRE release 6.7 (fixes some bugs).
747 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
749 PH/42 Give an error if -q is specified more than once.
751 PH/43 Renamed the variables $interface_address and $interface_port as
752 $received_ip_address and $received_port, to make it clear that these
753 values apply to message reception, and not to the outgoing interface when
754 a message is delivered. (The old names remain recognized, of course.)
756 PH/44 There was no timeout on the connect() call when using a Unix domain
757 socket in the ${readsocket expansion. There now is.
759 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
760 be meaningful with "accept".
762 SC/01 Eximstats V1.43
763 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
765 SC/02 Eximstats V1.44
766 Use a glob alias rather than an array ref in the generated
767 parser. This improves both readability and performance.
769 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
770 Collect SpamAssassin and rejection statistics.
771 Don't display local sender or destination tables unless
772 there is data to show.
773 Added average volumes into the top table text output.
775 SC/04 Eximstats V1.46
776 Collect data on the number of addresses (recipients)
777 as well as the number of messages.
779 SC/05 Eximstats V1.47
780 Added 'Message too big' to the list of mail rejection
781 reasons (thanks to Marco Gaiarin).
783 SC/06 Eximstats V1.48
784 Mainlog lines which have GMT offsets and are too short to
785 have a flag are now skipped.
787 SC/07 Eximstats V1.49 (Alain Williams)
788 Added the -emptyok flag.
790 SC/08 Eximstats V1.50
791 Fixes for obtaining the IP address from reject messages.
793 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
794 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
795 whitesspace changes from 4.64-PH/27
797 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
800 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
801 are found, allow negative numbers in numeric criteria)
803 JJ/06 exipick.20061117.2, added new $message_body_missing variable
805 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
806 to match changes made in 4.64-PH/43
808 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
810 PH/47 Put in an explicit test for a DNS lookup of an address record where the
811 "domain" is actually an IP address, and force a failure. This locks out
812 those revolvers/nameservers that support "A-for-A" lookups, in
813 contravention of the specifications.
815 PH/48 When a host name was looked up from an IP address, and the subsequent
816 forward lookup of the name timed out, the host name was left in
817 $sender_host_name, contrary to the specification.
819 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
820 restricted to single-key lookups, Exim was not diagnosing an error if
821 * or *@ was used with a query-style lookup.
823 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
825 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
826 long ago, but noone interested enough thought of it. Let's just say that
827 the "1.1" means that there are some new functions that weren't there at
828 some point in the past.
830 PH/51 Error processing for expansion failure of helo_data from an smtp
831 transport during callout processing was broken.
833 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
834 tested/used via the -bh/-bhc/-bs options.
836 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
837 bug, fixed in subsequent PCRE releases).
839 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
840 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
842 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
848 SC/01 Use a glob alias rather than an array ref in eximstats generated
849 parser. This improves both readability and performance.
851 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
852 Don't display local sender or destination tables in eximstats unless
853 there is data to show.
854 Added average volumes into the eximstats top table text output.
856 SC/03 Collect data on the number of addresses (recipients) as well
857 as the number of messages in eximstats.
859 TF/01 Correct an error in the documentation for the redirect router. Exim
860 does (usually) call initgroups() when daemonizing.
862 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
863 with consistent privilege compared to when running as a daemon.
865 TF/03 Note in the spec that $authenticated_id is not set for local
866 submissions from trusted users.
868 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
869 Thanks to Dean Brooks <dean@iglou.com> for the patch.
871 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
872 by adding some example configuration directives to the default
873 configuration file. A little bit of work is required to uncomment the
874 directives and define how usernames and passwords are checked, but
875 there is now a framework to start from.
877 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
878 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
879 without this. I don't know how relevant this is to other LDAP libraries.
881 PH/02 Add the verb name to the "unknown ACL verb" error.
883 PH/03 Magnus Holmgren's patch for filter_prepend_home.
885 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
887 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
888 directory not expanded when it should be if an expanded home directory
889 was set for the address (which is overridden by the transport).
891 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
894 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
895 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
896 because it is too late at that time, and has no effect.
898 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
899 security issue with \' (bugzilla #107). I could not use the
900 PQescapeStringConn() function, because it needs a PGconn value as one of
903 PH/08 When testing addresses using -bt, indicate those final addresses that
904 are duplicates that would not cause an additional delivery. At least one
905 person was confused, thinking that -bt output corresponded to deliveries.
906 (Suppressing duplicates isn't a good idea as you lose the information
907 about possibly different redirections that led to the duplicates.)
909 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
910 systems where poll() doesn't work, in particular OS X.
912 PH/10 Added more information to debugging output for retry time not reached.
914 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
915 operations in malware.c.
917 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
920 PH/13 If write_rejectlog was set false when logging was sent to syslog with
921 syslog_duplication set false, log lines that would normally be written
922 both the the main log and to the reject log were not written to syslog at
925 PH/14 In the default configuration, change the use of "message" in ACL warn
926 statements to "add_header".
928 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
929 not followed by a command (e.g. "seen endif").
931 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
932 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
935 PH/17 Added extra conditions to the default value of delay_warning_condition
939 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
940 { match{$h_precedence:}{(?i)bulk|list|junk} } \
941 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
944 The Auto-Submitted: and various List- headers are standardised, whereas I
945 don't think Precedence: ever was.
947 PH/18 Refactored debugging code in route_finduser() to show more information,
948 in particular, the error code if getpwnam() issues one.
950 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
951 This is apparently needed in addition to the PH/07 change above to avoid
952 any possible encoding problems.
954 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
955 but not after initializing Perl.
957 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
958 output them only if debugging. By default they are written stderr,
959 apparently, which is not desirable.
961 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
964 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
967 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
969 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
970 authenticated or an ident call has been made. Suppress the default
971 values for $authenticated_id and $authenticated_sender (but permit -oMai
972 and -oMas) when testing with -bh.
974 PH/24 Re-jigged the order of the tests in the default configuration so that the
975 tests for valid domains and recipients precede the DNS black list and CSA
976 tests, on the grounds that those ones are more expensive.
978 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
979 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
980 This bug exists in every version of Exim that I still have, right back to
983 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
984 However, an attempt to turn on case-sensitivity in a regex key by
985 including (?-i) didn't work because the subject string was already
986 lowercased, and the effects were non-intuitive. It turns out that a
987 one-line patch can be used to allow (?-i) to work as expected.
993 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
994 other effects) broke the use of negated acl sub-conditions.
996 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
999 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1000 "Deny" causes Exim to reject the incoming connection with a 554 error.
1001 Unfortunately, if there is a major crisis, such as a disk failure,
1002 tcp-wrappers gives "deny", whereas what one would like would be some
1003 kind of temporary error. A kludge has been added to help with this.
1004 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1005 554 error is used if errno is still zero or contains ENOENT (which occurs
1006 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1009 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1011 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1012 errors. Otherwise a message that provokes a temporary error (when other
1013 messages do not) can cause a whole host to time out.
1015 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1016 addresses were routed directly to files or pipes from a redirect router.
1017 File deliveries just didn't batch; pipe deliveries might have suffered
1020 PH/06 A failure to get a lock for a hints database would erroneously always say
1021 "Failed to get write lock", even when it was really a read lock.
1023 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1024 of 0600. This has been changed to use the value of the lockfile_mode
1025 option (which defaults to 0600).
1027 PH/08 Applied small patch from the Sieve maintainer.
1029 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1030 folder from quota calculations, a direct delivery into this folder messed
1031 up the contents of the maildirsize file. This was because the regex was
1032 used only to exclude .Trash (or whatever) when the size of the mailbox
1033 was calculated. There was no check that a delivery was happening into an
1034 excluded directory. This bug has been fixed by ignoring all quota
1035 processing for deliveries into excluded directories.
1037 PH/10 Added the maildirfolder_create_regex option to appendfile.
1043 PH/01 The code for finding all the local interface addresses on a FreeBSD
1044 system running IPv6 was broken. This may well have applied to all BSD
1045 systems, as well as to others that have similar system calls. The broken
1046 code found IPv4 interfaces correctly, but gave incorrect values for the
1047 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1048 that it would not match correctly against @[] and not recognize the IPv6
1051 PH/02 The ipliteral router was not recognizing addresses of the form user@
1052 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1054 PH/03 Added disable_ipv6.
1056 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1057 decoded form, because it is most often used to construct To: headers
1058 lines in autoreplies, and the decoded form may well be syntactically
1059 invalid. However, $reply_address has leading white space removed, and all
1060 newlines turned into spaces so that the autoreply transport does not
1063 PH/05 If group was specified without a user on a router, and no group or user
1064 was specified on a transport, the group from the router was ignored.
1066 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1067 for visible compile-time settings that can be used to change these
1068 numbers, for those that want even more. Backwards compatibility with old
1069 spool files has been maintained. However, going back to a previous Exim
1070 release will lost any variables that are in spool files.
1072 PH/07 Two small changes when running in the test harness: increase delay when
1073 passing a TCP/IP connection to a new process, in case the original
1074 process has to generate a bounce, and remove special handling of
1075 127.0.0.2 (sic), which is no longer necessary.
1077 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1078 be the same on different OS.
1080 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1083 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1084 whether --show-vars was specified or not
1086 JJ/02 exipick: Added support for new ACL variable spool format introduced
1089 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1090 syntactically invalid From: or Reply-to: line, and a filter used this to
1091 generate an autoreply, and therefore failed to obtain an address for the
1092 autoreply, Exim could try to deliver to a non-existent relative file
1093 name, causing unrelated and misleading errors. What now happens is that
1094 it logs this as a hard delivery error, but does not attempt to create a
1097 PH/11 The exinext utility has a -C option for testing purposes, but although
1098 the given file was scanned by exinext itself; it wasn't being passed on
1099 when Exim was called.
1101 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1102 an end-of-file indication when reading a command response.
1104 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1105 compiled. In many other places in Exim, IPv6 addresses are always
1106 recognized, so I have changed this. It also means that IPv4 domain
1107 literals of the form [IPV4:n.n.n.n] are now always recognized.
1109 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1110 used if the router is not running as root, for example, when verifying at
1111 ACL time, or when using -bh. The debugging output from this situation was
1112 non-existent - all you got was a failure to exec. I have made two
1115 (a) Failures to set uid/gid, the current directory, or a process leader
1116 in a subprocess such as that created by queryprogram now generate
1117 suitable debugging ouput when -d is set.
1119 (b) The queryprogram router detects when it is not running as root,
1120 outputs suitable debugging information if -d is set, and then runs
1121 the subprocess without attempting to change uid/gid.
1123 PH/15 Minor change to Makefile for building test_host (undocumented testing
1126 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1127 additional section of a DNS packet that returns MX or SRV records.
1128 Instead, it always explicitly searches for A/AAAA records. This avoids
1129 major problems that occur when a DNS server includes only records of one
1130 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1131 fixed another bug: if SRV records were looked up and the corresponding
1132 address records were *not* found in the additional section, the port
1133 values from the SRV records were lost.
1135 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1136 using the correct key (the original address) when searching the retry
1137 rules in order to find which one to use for generating the retry hint.
1139 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1140 adding the default one. Similarly, if it contains a Reply-To: header, the
1141 errors_reply_to option, if set, is not used.
1143 PH/19 When calculating a retry time, Exim used to measure the "time since
1144 failure" by looking at the "first failed" field in the retry record. Now
1145 it does not use this if it is later than than the arrival time of the
1146 message. Instead it uses the arrival time. This makes for better
1147 behaviour in cases where some deliveries succeed, thus re-setting the
1148 "first failed" field. An example is a quota failure for a huge message
1149 when small messages continue to be delivered. Without this change, the
1150 "time since failure" will always be short, possible causing more frequent
1151 delivery attempts for the huge message than are intended.
1152 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1154 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1155 $1, $2, $3) because the numerical variables can be reset during some
1156 expansion items (e.g. "match"), thereby losing the authentication data.
1158 PH/21 Make -bV show the size of off_t variables so that the test suite can
1159 decide whether to run tests for quotas > 2G.
1161 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1162 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1163 filecount value is greater than 2G or if a quota value is greater than 2G
1164 on a system where the size of off_t is not greater than 4, a panic error
1167 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1168 never match. The debug and -bh output now contains an explicit error
1169 message indicating a malformed IPv4 address or mask.
1171 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1172 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1173 PH/23 above applies.
1175 PH/25 Do not write to syslog when running in the test harness. The only
1176 occasion when this arises is a failure to open the main or panic logs
1177 (for which there is an explicit test).
1179 PH/26 Added the /no_tell option to "control=freeze".
1181 PH/27 If a host name lookup failed very early in a connection, for example, if
1182 the IP address matched host_lookup and the reverse lookup yielded a name
1183 that did not have a forward lookup, an error message of the form "no IP
1184 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1185 could be logged. Now it outputs the IP address instead of "NULL".
1187 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1188 allows the sender and the authenticated sender to be set when
1189 submitting a message from within Exim. Since child_open_exim() is
1190 documented for local_scan(), the new function should be too.
1192 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1193 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1194 results in an empty string is now treated as unset.
1196 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1198 PH/31 Added sender_verify_fail logging option.
1200 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1201 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1202 tidied the source and removed it altogether.
1204 PH/33 When a queue run was abandoned because the load average was too high, a
1205 log line was always written; now it is written only if the queue_run log
1206 selector is set. In addition, the log line for abandonment now contains
1207 information about the queue run such as the pid. This is always present
1208 in "start" and "stop" lines but was omitted from the "abandon" line.
1210 PH/34 Omit spaces between a header name and the colon in the error message that
1211 is given when verify = headers_syntax fails (if there are lots of them,
1212 the message gets confusing).
1214 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1215 names, as there are now some PTR records that contain slashes. This check
1216 is only to protect against broken name servers that fall over on strange
1217 characters, so the fact that it applies to all lookups doesn't matter.
1219 PH/36 Now that the new test suite is complete, we can remove some of the
1220 special code in Exim that was needed for the old test suite. For example,
1221 sorting DNS records because real resolvers return them in an arbitrary
1222 order. The new test suite's fake resolver always returns records in the
1225 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1226 bounces) except when queue_only is set, to avoid logging races between
1227 the different processes.
1229 PH/38 Panic-die if .include specifies a non-absolute path.
1231 PH/39 A tweak to the "H" retry rule from its user.
1233 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1234 a label. They prevented compilation on older perls.
1236 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1237 a warning to be raised on newish perls.
1239 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1240 on queue. Changes to match documented behaviour of showing count of
1241 messages matching specified criteria.
1243 PH/40 Changed the default ident timeout from 30s to 5s.
1245 PH/41 Added support for the use of login_cap features, on those BSD systems
1246 that have them, for controlling the resources used by pipe deliveries.
1248 PH/42 The content-scanning code uses fopen() to create files in which to put
1249 message data. Previously it was not paying any attention to the mode of
1250 the files. Exim runs with umask(0) because the rest of the code creates
1251 files with open(), and sets the required mode explicitly. Thus, these
1252 files were ending up world-writeable. This was not a big issue, because,
1253 being within the spool directory, they were not world-accessible. I have
1254 created a function called modefopen, which takes an additional mode
1255 argument. It sets umask(777), creates the file, chmods it to the required
1256 mode, then resets the umask. All the relevant calls to fopen() in the
1257 content scanning code have been changed to use this function.
1259 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1260 to 24 hours. This avoids potential overflow problems when processing G
1261 and H retry rules. I suspect nobody ever tinkers with this value.
1263 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1265 PH/45 When the plaintext authenticator is running as a client, the server's
1266 challenges are checked to ensure they are valid base64 strings. By
1267 default, the authentication attempt is cancelled if an invalid string is
1268 received. Setting client_ignore_invalid_base64 true ignores these errors.
1269 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1270 they are received. Thus, the responses can be made to depend on the
1271 challenges. If an invalid string is ignored, an empty string is placed in
1274 PH/46 Messages that are created by the autoreply transport now contains a
1275 References: header, in accordance with RFCs 2822 and 3834.
1277 PH/47 Added authenticated_sender_force to the smtp transport.
1279 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1281 PH/49 Installed latest patch from the Sieve maintainer.
1283 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1284 was also set, the appendfile transport was unnecessarily scanning a
1285 directory of message files (e.g. for maildir delivery) to find the count
1286 of files (along with the size), even though it did not need this
1287 information. It now does the scan only if it needs to find either the
1288 size of the count of files.
1290 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1292 PH/52 Two bugs concerned with error handling when the smtp transport is
1295 (i) Exim was not creating retry information for temporary errors given
1296 for individual recipients after the DATA command when the smtp transport
1297 was used in LMTP mode. This meant that they could be retried too
1298 frequently, and not timed out correctly.
1300 (ii) Exim was setting the flag that allows error details to be returned
1301 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1302 recipients that were returned after the DATA command.
1304 PH/53 This is related to PH/52, but is more general: for any failing address,
1305 when detailed error information was permitted to be returned to the
1306 sender, but the error was temporary, then after the final timeout, only
1307 "retry timeout exceeded" was returned. Now it returns the full error as
1308 well as "retry timeout exceeded".
1310 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1311 do this, and (what is worse) MTAs that accept it.
1313 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1314 will now be deprecated.
1316 PH/56 New os.c-cygwin from the Cygwin maintainer.
1318 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1319 formats (previously only available in exim formats via -bpr, -bpru,
1320 and -bpra. Now also available in native and exiqgrep formats)
1322 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1323 with very large, slow to parse queues
1325 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1327 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1329 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1330 responses to authentication challenges, though it was showing the
1331 challenges; (ii) I've removed the CR characters from the debug output for
1334 PH/58 Allow for the insertion of a newline as well as a space when a string
1335 is turned into more than one encoded-word during RFC 2047 encoding. The
1336 Sieve code now uses this.
1338 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1339 data_4xx, lost_connection, tls_required.
1341 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1342 message was being sent as an SMTP response.
1344 PH/61 Add -l and -k options to exicyclog.
1346 PH/62 When verifying, if an address was redirected to one new address, so that
1347 verification continued, and the new address failed or deferred after
1348 having set something in $address_data, the value of $address_data was not
1349 passed back to the ACL. This was different to the case when no
1350 redirection occurred. The value is now passed back in both cases.
1352 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1353 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1354 use login_cap.h, so on its own it isn't the distinguishing feature. The
1355 new name refers directly to the setclassresources() function.
1357 PH/65 Added configuration files for NetBSD3.
1359 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1361 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1362 is preferred over IPv4.
1364 PH/68 The bounce_return_message and bounce_return_body options were not being
1365 honoured for bounces generated during the reception of non-SMTP messages.
1366 In particular, this applied to messages rejected by the ACL. This bug has
1367 been fixed. However, if bounce_return_message is true and bounce_return_
1368 body is false, the headers that are returned for a non-SMTP message
1369 include only those that have been read before the error was detected.
1370 (In the case of an ACL rejection, they have all been read.)
1372 PH/69 The HTML version of the specification is now built in a directory called
1373 spec_html instead of spec.html, because the latter looks like a path with
1374 a MIME-type, and this confuses some software.
1376 PH/70 Catch two compiler warnings in sieve.c.
1378 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1379 function verify_get_ident() calls ip_connect() to connect a socket, but
1380 if the "connect()" function timed out, ip_connect() used to close the
1381 socket. However, verify_get_ident() also closes the socket later, and in
1382 between Exim writes to the log, which may get opened at this point. When
1383 the socket was closed in ip_connect(), the log could get the same file
1384 descriptor number as the socket. This naturally causes chaos. The fix is
1385 not to close the socket in ip_connect(); the socket should be closed by
1386 the function that creates it. There was only one place in the code where
1387 this was missing, in the iplookup router, which I don't think anybody now
1388 uses, but I've fixed it anyway.
1390 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1391 well as to direct DNS lookups. Otherwise the handling of names in host
1392 lists is inconsistent and therefore confusing.
1398 PH/01 Two changes to the default runtime configuration:
1400 (1) Move the checks for relay_from_hosts and authenticated clients from
1401 after to before the (commented out) DNS black list checks.
1403 (2) Add control=submission to the relay_from_hosts and authenticated
1404 clients checks, on the grounds that messages accepted by these
1405 statements are most likely to be submissions.
1407 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1409 (1) Generate an error if the third argument for the ${prvs expansion is
1412 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1415 (3) Reset the variables that are obtained from the first argument of
1416 ${prvscheck and used in the second argument before leaving the code,
1417 because their memory is reclaimed, so using them afterwards may do
1420 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1421 one (it's much easier than Tom thought :-).
1423 (5) Because of (4), we can now allow for the use of $prvscheck_result
1424 inside the third argument.
1426 PH/03 For some reason, the default setting of PATH when running a command from
1427 a pipe transport was just "/usr/bin". I have changed it to
1430 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1431 anything to be listed in the output from -bV.
1433 PH/05 When a filter generated an autoreply, the entire To: header line was
1434 quoted in the delivery log line, like this:
1436 => >A.N.Other <ano@some.domain> <original@ddress> ...
1438 This has been changed so that it extracts the operative address. There
1439 may be more than one such address. If so, they are comma-separated, like
1442 => >ano@some.domain,ona@other.domain <original@ddress> ...
1444 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1445 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1446 not being looked up in the rDNS to get a host name, Exim was showing the
1447 IP address twice in Received: lines, even though the IP addresses were
1448 identical. For example:
1450 Received: from [1.2.3.4] (helo=[1.2.3.4])
1452 However, if the real host name was known, it was omitting the HELO data
1453 if it matched the actual IP address. This has been tidied up so that it
1454 doesn't show the same IP address twice.
1456 PH/07 When both +timestamp and +memory debugging was on, the value given by
1457 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1458 called by the debug printing, thereby overwriting the timestamp buffer.
1459 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1461 PH/08 When the original message was included in an autoreply transport, it
1462 always said "this is a copy of the message, including all the headers",
1463 even if body_only or headers_only was set. It now gives an appropriate
1466 PH/09 Applied a patch from the Sieve maintainer which:
1468 o fixes some comments
1469 o adds the (disabled) notify extension core
1470 o adds some debug output for the result of if/elsif tests
1471 o points to the current vacation draft in the documentation
1472 and documents the missing references header update
1476 o fixes a bug in processing the envelope test (when testing
1477 multiple envelope elements, the last element determinted the
1480 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1481 Electronic Mail") by including:
1483 Auto-submitted: auto-generated
1485 in the messages that it generates (bounce messages and others, such as
1486 warnings). In the case of bounce messages for non-SMTP mesages, there was
1487 also a typo: it was using "Auto_submitted" (underscore instead of
1488 hyphen). Since every message generated by Exim is necessarily in response
1489 to another message, thes have all been changed to:
1491 Auto-Submitted: auto-replied
1493 in accordance with these statements in the RFC:
1495 The auto-replied keyword:
1497 - SHOULD be used on messages sent in direct response to another
1498 message by an automatic process,
1500 - MUST NOT be used on manually-generated messages,
1502 - MAY be used on Delivery Status Notifications (DSNs) and Message
1503 Disposition Notifications (MDNs),
1505 - MUST NOT be used on messages generated by automatic or periodic
1506 processes, except for messages which are automatic responses to
1509 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1510 to the default Received: header definition.
1512 PH/12 Added log selector acl_warn_skipped (default on).
1514 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1515 variables because (a) they are in the wrong storage pool and (b) even if
1516 they were copied, it wouldn't work properly because of the caching.
1518 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1519 checking when decoding. Apparently there are clients that generate
1520 overlong encoded strings. Why am I not surprised?
1522 PH/15 If the first argument of "${if match_address" was not empty, but did not
1523 contain an "@" character, Exim crashed. Now it writes a panic log message
1524 and treats the condition as false.
1526 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1528 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1529 extension "envelope-auth". The code is finished and in agreement with
1530 other implementations, but there is no documentation so far and in fact,
1531 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1532 not changing the active code.
1534 Print executed "if" and "elsif" statements when debugging is used. This
1535 helps a great deal to understand what a filter does.
1537 Document more things not specified clearly in RFC3028. I had all this
1538 sorted out, when out of a sudden new issues came to my mind. Oops."
1540 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1543 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1544 canonical form (as documented). However, after a host name lookup from
1545 the IP address, check_host() was doing a simple string comparison with
1546 addresses acquired from the DNS when checking that the found name did
1547 have the original IP as one of its addresses. Since any found IPv6
1548 addresses are likely to be in abbreviated form, the comparison could
1549 fail. Luckily, there already exists a function for doing the comparison
1550 by converting both addresses to binary, so now that is used instead of
1551 the text comparison.
1553 PH/20 There was another similar case to PH/19, when a complete host name was
1554 given in a host list; looking up its IP address could give an abbreviated
1555 form, whereas the current host's name might or might not be abbreviated.
1556 The same fix has been applied.
1562 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1563 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1566 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1567 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1569 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1571 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1572 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1573 Also, for "Auto-Submitted", treat anything other than "no" as
1574 non-personal, in accordance with RFC 3834. (Previously it treated
1575 anything starting "auto-" as non-personal.)
1577 TF/01 The control=submission/name=... option had a problem with syntax
1578 errors if the name included a slash character. The /name= option
1579 now slurps the rest of the string, so it can include any characters
1580 but it must come last in the list of options (after /sender_retain
1583 PH/05 Some modifications to the interface to the fake nameserver for the new
1591 TK/01 Added the "success_on_redirect" address verification option. See
1592 NewStuff for rationale and an example.
1594 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1596 PH/02 Patch to exigrep to allow it to work on syslog lines.
1598 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1599 fread() to skip over the body file's header line, because in Cygwin the
1600 header line is locked and is inaccessible.
1602 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1603 co-exist for some time) to make it clear that it is the Exim ID that is
1604 referenced, not the Message-ID: header line.
1606 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1607 string_format() function, because snprintf() does not exist on all
1610 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1611 ${stat: expansion item.
1613 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1614 protocol synchronization error", to keep the pedants happy.
1616 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1617 well as for IRIX systems, when gcc is being used. See the host.c source
1620 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1622 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1625 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1626 transport and to the smtp transport in LMTP mode.
1628 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1630 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1631 run a filter in a subprocess. This could lead to confusion in subsequent
1632 lookups in the parent process. There should also be a search_tidyup() at
1633 the end of the subprocess.
1635 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1636 only if the host matched helo_try_verify_hosts, which caused the
1637 verification to occur when the EHLO/HELO command was issued. The ACL just
1638 tested the remembered result. Now, if a previous verification attempt has
1639 not happened, "verify = helo" does it there and then.
1641 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1643 TK/03 Fix log output including CR from clamd.
1645 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1646 exist provoked a memory error which could cause a segfault.
1648 PH/15 Installed PCRE 6.2
1650 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1652 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1653 of the problem. Specifically, suggested +O2 rather than +O1 for the
1656 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1658 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1659 record, the port was not correctly logged when the outgoing_port log
1660 selector was set (it logged the transort's default port).
1662 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1663 fallback_hosts, and "hosts" in the smtp transport.
1665 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1666 host errors such as "Connection refused".
1668 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1669 authentication with radiusclient 0.4.9:
1671 - Error returned from rc_read_config was caught wrongly
1672 - Username/password not passed on to radius server due to wrong length.
1674 The presumption is that some radiusclient API changes for 4.51/PH/17
1675 were not taken care of correctly. The code is still untested by me (my
1676 Linux distribution still has 0.3.2 of radiusclient), but it was
1677 contributed by a Radius user.
1679 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1680 expanding the "port" option of the smtp transport.
1682 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1683 while reading a MIME header. Thanks to Tom Hughes for a patch.
1685 PH/24 Include config.h inside local_scan.h so that configuration settings are
1688 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1689 that in an ACL for RCPT (for example), you can examine exactly what was
1692 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1693 commands, but it was not correctly comparing the address with the actual
1694 client host address. Thus, it would show the EHLO address in Received:
1695 header lines when this was not necessary.
1697 PH/27 Added the % operator to ${eval:}.
1699 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1700 it should be ignoring failures (because with -C, for example, it has lost
1701 privilege). It wasn't ignoring creation failures other than "already
1704 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1707 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1708 because an input line was too long, either on its own, or by virtue of
1709 too many continuations, the temporary file was not being removed, and the
1710 return code was incorrect.
1712 PH/31 Missing "BOOL" in function definition in filtertest.c.
1714 PH/32 Applied Sieve patches from the maintainer.
1716 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1718 PH/33 Added "verify = not_blind".
1720 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1721 Local/Makefile (with some defaults set). These are used in built scripts
1722 such as exicyclog, but they have never been used in the exim_install
1723 script (though there are many overriding facilities there). I have
1724 arranged that the exim_install script now takes note of these two
1727 PH/35 Installed configuration files for Dragonfly.
1729 PH/36 When a locally submitted message by a trusted user did not contain a
1730 From: header, and the sender address was obtained from -f or from an SMTP
1731 MAIL command, and the trusted user did not use -F to supply a sender
1732 name, $originator_name was incorrectly used when constructing a From:
1733 header. Furthermore, $originator_name was used for submission mode
1734 messages from external hosts without From: headers in a similar way,
1735 which is clearly wrong.
1737 PH/37 Added control=suppress_local_fixups.
1739 PH/38 When log_selector = +received_sender was set, and the addition of the
1740 sender made the log line's construction buffer exactly full, or one byte
1741 less than full, an overflow happened when the terminating "\n" was
1744 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1745 when the result of a list match is failure because a DNS lookup failed.
1747 PH/40 RM_COMMAND is now used in the building process.
1749 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1750 the "build-* directories that it finds.
1752 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1753 address in a domain literal was a prefix of an interface address.
1755 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1756 when verifying a sender address, unless rewrite_headers is false.
1758 PH/44 Wrote a long comment about why errors_to addresses are verified as
1759 recipients, not senders.
1761 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1762 the ratelimit ACL was added.
1764 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1766 PH/46 Added extra information about PostgreSQL errors to the error string.
1768 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1769 avoiding the need to install special zones in a real server. This is
1770 backwards compatible; if it can't find the fake resolver, it drops back.
1771 Thus, both old and new test suites can be run.
1773 TF/02 Added util/ratelimit.pl
1775 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1776 clock is set back in time.
1778 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1779 Candler <B.Candler@pobox.com>.
1781 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1782 OK for addresses that are the result of redirections.
1784 PH/48 A number of further additions for the benefit of the new test suite,
1785 including a fake gethostbyname() that interfaces to the fake DNS resolver
1788 TF/06 The fix for widen_domains has also been applied to qualify_single and
1789 search_parents which are the other dnslookup options that can cause
1792 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1795 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1797 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1798 service" if it encountered an MX record with an empty target hostname.
1799 The message is now "an MX or SRV record indicated no SMTP service".
1801 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1802 if the DNS of the sending site is misconfigured. This is quite a
1803 common situation. This change restores the behaviour of treating a
1804 helo verification defer as a failure.
1806 PH/51 If self=fail was set on a router, the bounce message did not include the
1807 actual error message.
1813 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1815 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1816 command itself ended in error, the underlying message about the transport
1817 filter timeout was being overwritten with the pipe command error. Now the
1818 underlying error message should be appended to the second error message.
1820 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1822 PH/02 Reduce the amount of output that "make" produces by default. Full output
1823 can still be requested.
1825 PH/03 The warning log line about a condition test deferring for a "warn" verb
1826 was being output only once per connection, rather than after each
1827 occurrence (because it was using the same function as for successful
1828 "warn" verbs). This seems wrong, so I have changed it.
1830 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1831 it should not have, which might have caused a crash in the right
1832 circumstances, but probably never did.
1834 PH/04 Installed a modified version of Tony Finch's patch to make submission
1835 mode fix the return path as well as the Sender: header line, and to
1836 add a /name= option so that you can make the user's friendly name appear
1839 TF/03 Added the control = fakedefer ACL modifier.
1841 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1842 Mark Lowes for thorough testing.
1844 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1846 TK/03 Merged latest SRS patch from Miles Wilton.
1848 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1849 with the definition in sysexits.h (which is #included earlier).
1850 Fortunately, Exim does not actually use EX_OK. The code used to try to
1851 preserve the sysexits.h value, by assumimg that macro definitions were
1852 scanned for macro replacements. I have been disabused of this notion,
1853 so now the code just undefines EX_OK before #including unistd.h.
1855 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1856 in the smtp transport. When a block could not be written in a single
1857 write() function, the timeout was being re-applied to each part-write.
1858 This seems wrong - if the receiver was accepting one byte at a time it
1859 would take for ever. The timeout is now adjusted when this happens. It
1860 doesn't have to be particularly precise.
1862 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1863 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1865 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1866 without a domain if the check to <postmaster@domain> fails.
1868 SC/01 Eximstats: added -xls and the ability to specify output files
1869 (patch written by Frank Heydlauf).
1871 SC/02 Eximstats: use FileHandles for outputing results.
1873 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1875 SC/04 Eximstats: fixed display of large numbers with -nvr option
1877 SC/05 Eximstats: fixed merging of reports with empty tables.
1879 SC/06 Eximstats: added the -include_original_destination flag
1881 SC/07 Eximstats: removed tabs and trailing whitespace.
1883 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1885 TK/06 MBOX spool code: Add real "From " MBOX separator line
1886 so the .eml file is really in mbox format (even though
1887 most programs do not really care). Patch from Alex Miller.
1889 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1890 The latter is generated from $received_to and is only set if the
1891 message has one envelope recipient. SA can use these headers,
1892 obviously out-of-the-box. Patch from Alex Miller.
1894 PH/08 The ${def test on a variable was returning false if the variable's
1895 value was "0", contrary to what the specification has always said!
1896 The result should be true unless the variable is empty.
1898 PH/09 The syntax error of a character other than { following "${if
1899 def:variable_name" (after optional whitespace) was not being diagnosed.
1900 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1901 accidental colon was present, for example, could give incorrect results.
1903 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1904 result is used (not including appendfile, where other changes are about
1907 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1908 This involved changing a lot of size variables from int to off_t. It
1909 should work with maildirs and everything.
1911 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1912 spamd dying while we are connected to it.
1914 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1917 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1918 function for BDB 4.3.
1920 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1922 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1923 the sources for PCRE have been re-arranged and more files are now
1926 PH/15 The code I had for printing potentially long long variables in PH/11
1927 above was not the best (it lost precision). The length of off_t variables
1928 is now inspected at build time, and an appropriate printing format (%ld
1929 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1930 to be "long long int" or "long int". This is needed for the internal
1931 formatting function string_vformat().
1933 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1934 the configuration file to be ":syslog", then the script "guesses" where
1935 the logs files are, rather than using the compiled in default. In our
1936 case the guess is not the same as the compiled default, so the script
1937 suddenly stopped working when I started to use syslog. The patch checks
1938 to see if log_file_path is "". If so, it attempts to read it from exim
1939 with no configuration file to get the compiled in version, before it
1940 falls back to the previous guessing code."
1942 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1943 implementing BATV in an Exim configuration. See NewStuff for the gory
1946 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1947 Makefile that are specific to HP-UX.
1949 PH/18 If the "use_postmaster" option was set for a recipient callout together
1950 with the "random" option, the postmaster address was used as the MAIL
1951 FROM address for the random test, but not for the subsequent recipient
1952 test. It is now used for both.
1954 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1955 patch removes a few documentation additions to RFC 3028, because the
1956 latest draft now contains them. It adds the new en;ascii-case comparator
1957 and a new error check for 8bit text in MIME parts. Comparator and
1958 require names are now matched exactly. I enabled the subaddress
1959 extension, but it is not well tested yet (read: it works for me)."
1961 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1962 rework some of the code of TK/09 above to avoid the hardwired use of
1963 "%lld" and "long long". Replaced the call to snprintf() with a call to
1966 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1967 records point to non-existent hosts", "retry timeout exceeded", and
1968 "retry time not reached for any host after a long failure period".
1970 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1971 experimental DomainKeys support:
1973 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1974 (2) On an error such as an illegally used "control", the wrong name for
1975 the control was given.
1977 These problems did NOT occur unless DomainKeys support was compiled.
1979 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1981 PH/24 Added ${if match_ip condition.
1983 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1984 databases so that it will be absolutely obvious if a crash occurs in the
1985 DB library. This is a regular occurrence (often caused by mis-matched
1988 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1989 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1990 picked up on a user's system that detects such things. There doesn't seem
1991 to be a gcc warning option for this - only an attribute that has to be
1992 put on the function's prototype. It seems that in Fedora Core 4 they have
1993 set this on a number of new functions. No doubt there will be more in due
1996 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1997 specify a transport. However, if an address that was verified by such a
1998 router was the subject of a callout, Exim crashed because it tried to
1999 read the rcpt_include_affixes from the non-existent transport. Now it
2000 just assumes that the setting of that option is false. This bug was
2001 introduced by 4.51/PH/31.
2003 PH/28 Changed -d+all to exclude +memory, because that information is very
2004 rarely of interest, but it makes the output a lot bigger. People tend to
2005 do -d+all out of habit.
2007 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2008 code in os-type was giving problems when libc.so lives in lib64, like on
2011 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2012 aren't the modern standard, and it seems that some systems' include files
2013 don't always have them. Exim was already checking for some of the newer
2014 ones like T_AAAA, and defining it itself. I've added checks for all the
2015 record types that Exim uses.
2017 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2018 not automatically generating a new one, as it is supposed to. This
2019 prevented TLS from working. If the file did exist, but contained invalid
2020 data, a new version was generated, as expected. It was only the case of a
2021 non-existent file that was broken.
2023 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2024 with a change in libdomainkeys > 0.64.
2026 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2027 from DNS. If the selector record carries the flag, it now has
2028 precedence over the domain-wide flag.
2030 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2032 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2033 the use of a port name that isn't defined in /etc/services) occurred, the
2034 message was deferred as in a normal delivery, and thus remained on the
2035 spool, instead of being failed because of the mua_wrapper setting. This
2036 is now fixed, and I tidied up some of the mua_wrapper messages at the
2039 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2040 the messages in a hash of arrays rather than using individual hashes.
2041 This is a bit cleaner and results in dramatic memory savings, albeit
2042 at a slight CPU cost.
2044 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2045 as requested by Marc Sherman.
2047 SC/10 Eximstats: added histograms for user specified patterns as requested
2050 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2052 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2053 fopen() in the content-scanning modules that did not already have it.
2059 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2060 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2062 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2064 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2066 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2067 to test Sieve filters that use "vacation".
2069 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2070 that changes the way the GnuTLS parameters are stored in the cache file.
2071 The new format can be generated externally. For backward compatibility,
2072 if the data in the cache doesn't make sense, Exim assumes it has read an
2073 old-format file, and it generates new data and writes a new file. This
2074 means that you can't go back to an older release without removing the
2077 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2078 work if there are any delivery delays because "one_time" forces the
2079 parent to be marked "delivered", so its unseen clone is never tried
2080 again. For this reason, Exim now forbids the simultaneous setting of
2083 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2084 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2085 entry if you want to know the details. The fix, however, neglected to
2086 consider the case when local delivery batching is involved. The test for
2087 "previously delivered" was not happening when checking to see if an
2088 address could be batched with a previous (undelivered) one; under
2089 certain circumstances this could lead to multiple deliveries to the same
2092 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2093 in its include files, and this causes problems building Exim.
2095 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2096 header_syntax) but Exim was just ignoring anything given after a slash.
2097 In particular, this caused confusion with an attempt to use "verify =
2098 reverse_host_lookup/defer_ok". An error is now given when options are
2099 supplied for verify items that do not have them. (Maybe reverse_host_
2100 lookup should have a defer_ok option, but that's a different point.)
2102 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2103 defined by RFC 821) to 2048, because there were problems with some AUTH
2104 commands, and RFC 1869 says the size should be increased for extended
2105 SMTP commands that take arguments.
2107 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2110 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2111 "unknown" error; now it says that the functionality isn't in the binary.
2113 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2114 an address' error message when a string expansion fails (syntax or
2115 whatever). Otherwise the password may appear in the log. Following change
2116 PH/42 below, there is no longer a chance of it appearing in a bounce
2119 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2121 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2122 of its addresses was ever tried. (Bugzilla bug #2.)
2124 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2125 the result incorrectly in the debug output. (It correctly added a newline
2126 to what was transported.)
2128 TF/01 Added $received_time.
2130 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2131 commented out examples of how to interface to a virus scanner and to
2132 SpamAssassin. Also added commented examples of av_scanner and
2133 spamd_address settings.
2135 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2136 and controls are allowed in which ACLs. There were a couple of minor
2137 errors. Some of the entries in the conditions table (which is a table of
2138 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2139 negation of where the condition IS allowed.
2141 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2143 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2144 header file does not have a version number, so I've had to invent a new
2145 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2146 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2147 radiusclient), but it was contributed by a Radius user.
2149 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2150 files or directories, for OpenSSL.
2152 PH/19 When an Exim process that is running as root has to create an Exim log
2153 file, it does so in a subprocess that runs as exim:exim so as to get the
2154 ownership right at creation (otherwise, other Exim processes might see
2155 the file with the wrong ownership). There was no test for failure of this
2156 fork() call, which would lead to the process getting stuck as it waited
2157 for a non-existent subprocess. Forks do occasionally fail when resources
2158 run out. I reviewed all the other calls to fork(); they all seem to check
2161 PH/20 When checking for unexpected SMTP input at connect time (before writing
2162 the banner), Exim was not dealing correctly with a non-positive return
2163 from the read() function. If the client had disconnected by this time,
2164 the result was a log entry for a synchronization error with an empty
2165 string after "input=" when read() returned zero. If read() returned -1
2166 (an event I could not check), uninitialized data bytes were printed.
2167 There were reports of junk text (parts of files, etc) appearing after
2170 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2172 PH/22 Added support for macro redefinition, and (re)definition in between
2173 driver and ACL definitions.
2175 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2176 forgetting to use the resulting value; it was using the unexpanded value.
2178 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2179 hadn't been configured. The fix is from Juergen Kreileder, who
2180 understands it better than I do:
2182 "Here's what I see happening with three configured cyrus_sasl
2183 authenticators configured (plain, login, cram-md5):
2185 On startup auth_cyrus_sasl_init() gets called for each of these.
2186 This means three calls to sasl_listmech() without a specified mech_list.
2187 => SASL tests which mechs of all available mechs actually work
2188 => three warnings about OTP not working
2189 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2191 With the patch, sasl_listmech() also gets called three times. But now
2192 SASL's mech_list option is set to the server_mech specified in the the
2193 authenticator. Or in other words, the answer from sasl_listmech()
2194 gets limited to just the mech you're testing for (which is different
2196 => the return list contains just 'plain' or 'login', 'cram-md5' or
2197 nothing depending on the value of ob->server_mech.
2199 I've just tested the patch: Authentication still works fine,
2200 unavailable mechs specified in the exim configuration are still
2201 caught, and the auth.log warnings about OTP are gone."
2203 PH/25 When debugging is enabled, the contents of the command line are added
2204 to the debugging output, even when log_selector=+arguments is not
2207 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2208 answer is "GNU", and only if the return is "GNU/something" is the answer
2211 PH/27 $acl_verify_message is now set immediately after the failure of a
2212 verification in an ACL, and so is available in subsequent modifiers. In
2213 particular, the message can be preserved by coding like this:
2215 warn !verify = sender
2216 set acl_m0 = $acl_verify_message
2218 Previously, $acl_verify_message was set only while expanding "message"
2219 and "log_message" when a very denied access.
2221 PH/28 Modified OS/os.c-Linux with
2223 -#ifndef OS_LOAD_AVERAGE
2224 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2226 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2227 nomenclature these days.)
2229 PH/29 Installed patch from the Sieve maintainer that adds the options
2230 sieve_useraddress and sieve_subaddress to the redirect router.
2232 PH/30 In these circumstances:
2233 . Two addresses routed to the same list of hosts;
2234 . First host does not offer TLS;
2235 . First host accepts first address;
2236 . First host gives temporary error to second address;
2237 . Second host offers TLS and a TLS session is established;
2238 . Second host accepts second address.
2239 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2240 and peerdn, if requested) that were in fact used only for the second
2243 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2244 attention to any local part prefix or suffix that was matched by the
2245 router that accepted the address. It now behaves in the same way as it
2246 does for delivery: the affixes are removed from the local part unless
2247 rcpt_include_affixes is set on the transport.
2249 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2250 timeout during the DATA phase of an incoming message.
2252 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2253 applied a patch sanctioned by the Sieve maintainer.
2255 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2256 the uid or gid is negative. A case of a negative gid caused this to be
2257 noticed. The fix allows for either to be negative.
2259 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2260 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2263 PH/36 The change PH/12 above was broken. Fixed it.
2265 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2266 the grounds that routing the same address twice would always produce the
2267 same answer. This might have been true once, but it is certainly no
2268 longer true now. Routing a child address may depend on the previous
2269 routing that produced that child. Some complicated redirection strategies
2270 went wrong when messages had multiple recipients, and made Exim's
2271 behaviour dependent on the order in which the addresses were given.
2273 I have moved the duplicate checking until after the routing is complete.
2274 Exim scans the addresses that are assigned to local and remote
2275 transports, and removes any duplicates. This means that more work will be
2276 done, as duplicates will always all be routed, but duplicates are
2277 presumably rare, so I don't expect this is of any significance.
2279 For deliveries to pipes, files, and autoreplies, the duplicate checking
2280 still happens during the routing process, since they are not going to be
2283 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2284 It corrects a timeout issue with spamd. This is Ian's comment: "The
2285 background is that sometimes spamd either never reads data from a
2286 connection it has accepted, or it never writes response data. The exiscan
2287 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2288 blindly assumes that writes won't block so it may never time out."
2290 PH/39 Allow G after quota size as well as K and M.
2292 PH/40 The value set for $authenticated_id in an authenticator may not contain
2293 binary zeroes or newlines because the value is written to log lines and
2294 to spool files. There was no check on this. Now the value is run through
2295 the string_printing() function so that such characters are converted to
2296 printable escape sequences.
2298 PH/41 $message_linecount is a new variable that contains the total number of
2299 lines in the message. Compare $body_linecount, which is the count for the
2302 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2303 bounce and delay warning messages, except in certain special cases, which
2306 (a) An SMTP error message from a remote host;
2307 (b) A message specified in a :fail: redirection;
2308 (c) A message specified in a "fail" command in a system filter;
2309 (d) A message specified in a FAIL return from the queryprogram router;
2310 (e) A message specified by the cannot_route_message router option.
2312 In these cases only, Exim does include the error details in bounce and
2313 warning messages. There are also a few cases where bland messages such
2314 as "unrouteable address" or "local delivery error" are given.
2316 PH/43 $value is now also set for the "else" part of a ${run expansion.
2318 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2319 being worked on, but at least Exim now implements the latest version to
2322 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2323 process to complete was treated as a delivery failure, a timeout while
2324 writing the message to the pipe was logged, but erroneously treated as a
2325 successful delivery. Such timeouts include transport filter timeouts. For
2326 consistency with the overall process timeout, these timeouts are now
2327 treated as errors, giving rise to delivery failures by default. However,
2328 there is now a new Boolean option for the pipe transport called
2329 timeout_defer, which, if set TRUE, converts the failures into defers for
2330 both kinds of timeout. A transport filter timeout is now identified in
2333 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2334 systems where "make" and "gmake" are different, calling "gmake" at top
2335 level broke things. I've arranged for the value of $(MAKE) to be passed
2336 from the Makefile to this script so that it can call the same version of
2340 A note about Exim versions 4.44 and 4.50
2341 ----------------------------------------
2343 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2344 changes of various kinds. As a consequence, a big documentation update was
2345 needed. This delayed the release for rather longer than seemed good, especially
2346 in the light of a couple of (minor) security issues. Therefore, the changes
2347 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2348 release. So 4.44 and 4.50 are in effect two different branches that both start
2351 I have left the 4.50 change log unchanged; it contains all the changes since
2352 4.43. The change log for 4.44 is below; many of its items are identical to
2353 those for 4.50. This seems to be the most sensible way to preserve the
2354 historical information.
2360 1. Minor wording change to the doc/README.SIEVE file.
2362 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2363 computation of the current number of files was incorrect.
2365 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2366 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2369 4. Give more explanation in the error message when the command for a transport
2370 filter fails to execute.
2372 5. There are several places where Exim runs a non-Exim command in a
2373 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2374 was being done only for the command run by the queryprogram router. It is
2375 now done for all such subprocesses. The other cases are: ${run, transport
2376 filters, and the commands run by the lmtp and pipe transports.
2378 6. Added CONFIGURE_GROUP build-time option.
2380 7. Some older OS have a limit of 256 on the maximum number of file
2381 descriptors. Exim was using setrlimit() to set 1000 as a large value
2382 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2383 systems. I've change it so that if it can't get 1000, it tries for 256.
2385 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2386 was an oversight, and furthermore, ever since the addition of extra
2387 controls (e.g. 4.43/32), the checks on when to allow different forms of
2388 "control" were broken. There should now be diagnostics for all cases when a
2389 control that does not make sense is encountered.
2391 9. Added the /retain_sender option to "control=submission".
2393 10. $recipients is now available in the predata ACL (oversight).
2395 11. Tidy the search cache before the fork to do a delivery from a message
2396 received from the command line. Otherwise the child will trigger a lookup
2397 failure and thereby defer the delivery if it tries to use (for example) a
2398 cached ldap connection that the parent has called unbind on.
2400 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2401 of $address_data from the recipient verification was clobbered by the
2402 sender verification.
2404 13. The value of address_data from a sender verification is now available in
2405 $sender_address_data in subsequent conditions in the ACL statement.
2407 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2409 15. Added a new option "connect=<time>" to callout options, to set a different
2412 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2413 was its contents. (It was OK if the option was not defined at all.)
2415 17. A "Completed" log line is now written for messages that are removed from
2416 the spool by the -Mrm option.
2418 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2419 information about exactly what failed.
2421 19. Added -dd to debug only the daemon process.
2423 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2424 handles timeouts, both on the server side and network timeouts. Renamed the
2425 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2427 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2428 It is now set to "smtps".
2430 22. $host_address is now set to the target address during the checking of
2431 ignore_target_hosts.
2433 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2434 being passed; this would have caused $sender_host_name to have been used if
2435 matching the list had actually called for a host name (not very likely,
2436 since this list is usually IP addresses). A host name is now passed as
2439 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2440 code that specifies a non-restarting handler (typically sigaction() in
2441 modern systems) in an attempt to fix a rare and obscure crash bug.
2443 25. Narrowed the window for a race in the daemon that could cause it to ignore
2444 SIGCHLD signals. This is not a major problem, because they are used only to
2445 wake it up if nothing else does.
2447 26. A malformed maildirsize file could cause Exim to calculate negative values
2448 for the mailbox size or file count. Odd effects could occur as a result.
2449 The maildirsize information is now recalculated if the size or filecount
2452 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2453 support for a long time. Removed HAVE_SYS_VFS_H.
2455 28. Installed the latest version of exipick from John Jetmore.
2457 29. In an address list, if the pattern was not a regular expression, an empty
2458 subject address (from a bounce message) matched only if the pattern was an
2459 empty string. Non-empty patterns were not even tested. This was the wrong
2460 because it is perfectly reasonable to use an empty address as part of a
2461 database query. An empty address is now tested by patterns that are
2462 lookups. However, all the other forms of pattern expect the subject to
2463 contain a local part and a domain, and therefore, for them, an empty
2464 address still always fails if the pattern is not itself empty.
2466 30. Exim went into a mad DNS loop when attempting to do a callout where the
2467 host was specified on an smtp transport, and looking it up yielded more
2468 than one IP address.
2470 31. Re-factored the code for checking spool and log partition space into a
2471 function that finds that data and another that does the check. The former
2472 is then used to implement four new variables: $spool_space, $log_space,
2473 $spool_inodes, and $log_inodes.
2475 32. The RFC2047 encoding function was originally intended for short strings
2476 such as real names; it was not keeping to the 75-character limit for
2477 encoded words that the RFC imposes. It now respects the limit, and
2478 generates multiple encoded words if necessary. To be on the safe side, I
2479 have increased the buffer size for the ${rfc2047: expansion operator from
2482 33. It is now permitted to omit both strings after an "if" condition; if the
2483 condition is true, the result is "true". As before, when the second string
2484 is omitted, a false condition yields an empty string. This makes it less
2485 cumbersome to write custom ACL and router conditions.
2487 34. Failure to deliver a bounce message always caused it to be frozen, even if
2488 there was an errors_to setting on the router. The errors_to setting is now
2491 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2492 canonical form (fully expanded) before being placed in
2493 $sender_host_address.
2495 36. The table in the code that translates DNS record types into text (T_A to
2496 "A" for instance) was missing entries for NS and CNAME. It is just possible
2497 that this could have caused confusion if both these types were looked up
2498 for the same domain, because the text type is used as part of Exim's
2499 per-process caching. But the chance of anyone hitting this buglet seems
2502 37. The dnsdb lookup has been extended in a number of ways.
2504 (1) There is a new type, "zns", which walks up the domain tree until it
2505 finds some nameserver records. It should be used with care.
2507 (2) There is a new type, "mxh", which is like "mx" except that it returns
2508 just the host names, not the priorities.
2510 (3) It is now possible to give a list of domains (or IP addresses) to be
2511 looked up. The behaviour when one of the lookups defers can be
2512 controlled by a keyword.
2514 (4) It is now possible to specify the separator character for use when
2515 multiple records are returned.
2517 38. The dnslists ACL condition has been extended: it is now possible to supply
2518 a list of IP addresses and/or domains to be looked up in a particular DNS
2521 39. Added log_selector=+queue_time_overall.
2523 40. When running the queue in the test harness, wait just a tad after forking a
2524 delivery process, to get repeatability of debugging output.
2526 41. Include certificate and key file names in error message when GnuTLS fails
2527 to set them up, because the GnuTLS error message doesn't include the name
2528 of the failing file when there is a problem reading it.
2530 42. Allow both -bf and -bF in the same test run.
2532 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2534 44. The "Exiscan patch" is now merged into the mainline Exim source.
2536 45. Sometimes the final signoff response after QUIT could fail to get
2537 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2538 < 0 before doing a fflush(). This bug looks as though it goes back to the
2539 introduction of TLS in release 3.20, but "sometimes" must have been rare
2540 because the tests only now provoked it.
2542 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2543 (this can affect the format of dates).
2545 47. exim_tidydb, when checking for the continued existence of a message for
2546 which it has found a message-specific retry record, was not finding
2547 messages that were in split spool directories. Consequently, it was
2548 deleting retry records that should have stayed in existence.
2550 48. Steve fixed some bugs in eximstats.
2552 49. The SPA authentication driver was not abandoning authentication and moving
2553 on to the next authenticator when an expansion was forced to fail,
2554 contradicting the general specification for all authenticators. Instead it
2555 was generating a temporary error. It now behaves as specified.
2557 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2558 (the order specifies the preference for clients). The order is now AES256,
2559 AES128, 3DES, ARCFOUR128.
2561 51. Small patch to Sieve code - explicitly set From: when generating an
2564 52. Exim crashed if a remote delivery caused a very long error message to be
2565 recorded - for instance if somebody sent an entire SpamAssassin report back
2566 as a large number of 550 error lines. This bug was coincidentally fixed by
2567 increasing the size of one of Exim's internal buffers (big_buffer) that
2568 happened as part of the Exiscan merge. However, to be on the safe side, I
2569 have made the code more robust (and fixed the comments that describe what
2572 53. Now that there can be additional text after "Completed" in log lines (if
2573 the queue_time_overall log selector is set), a one-byte patch to exigrep
2574 was needed to allow it to recognize "Completed" as not the last thing in
2577 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2578 patch that reportedly fixes this has been added. I am not expert enough to
2579 create a test for it. This is what the patch creator wrote:
2581 "I found a little strange behaviour of ldap code when working with
2582 Windows 2003 AD Domain, where users was placed in more than one
2583 Organization Units. When I tried to give exim partial DN, the exit code
2584 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2585 But simultaneously result of request was absolutely normal ldap result,
2586 so I produce this patch..."
2588 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2589 REFERENCE, so I have modified the code to exclude the patch when that macro
2592 55. Some experimental protocols are using DNS PTR records for new purposes. The
2593 keys for these records are domain names, not reversed IP addresses. The
2594 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2595 leaves it alone. Component reversal etc. now happens only for IP addresses.
2598 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2600 57. Double the size of the debug message buffer (to 2048) so that more of very
2601 long debug lines gets shown.
2603 58. The exicyclog utility now does better if the number of log files to keep
2604 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2606 59. Two changes related to the smtp_active_hostname option:
2608 (1) $smtp_active_hostname is now available as a variable.
2609 (2) The default for smtp_banner uses $smtp_active_hostname instead
2610 of $primary_hostname.
2612 60. The host_aton() function is supposed to be passed a string that is known
2613 to be a valid IP address. However, in the case of IPv6 addresses, it was
2614 not checking this. This is a hostage to fortune. Exim now panics and dies
2615 if the condition is not met. A case was found where this could be provoked
2616 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2617 components; fortuitously, this particular loophole had already been fixed
2618 by change 4.50/55 above.
2620 If there are any other similar loopholes, the new check in host_aton()
2621 itself should stop them being exploited. The report I received stated that
2622 data on the command line could provoke the exploit when Exim was running as
2623 exim, but did not say which command line option was involved. All I could
2624 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2625 running as the user.
2628 61. There was a buffer overflow vulnerability in the SPA authentication code
2629 (which came originally from the Samba project). I have added a test to the
2630 spa_base64_to_bits() function which I hope fixes it.
2633 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2634 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2636 63. The daemon start-up calls getloadavg() while still root for those OS that
2637 need the first call to be done as root, but it missed one case: when
2638 deliver_queue_load_max is set with deliver_drop_privilege. This is
2639 necessary for the benefit of the queue runner, because there is no re-exec
2640 when deliver_drop_privilege is set.
2642 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2643 This has been fixed.
2645 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2646 was in use, was not putting the data itself into the right store pool;
2647 consequently, it could be overwritten for a subsequent message in the same
2648 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2651 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2653 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2654 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2655 the function were treating the return as a boolean value, which happened to
2656 work because 0=false and not-0=true, but is not correct code.
2658 68. The host_aton() function was not handling scoped IPv6 addresses (those
2659 with, for example, "%eth0" on the end) correctly.
2661 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2662 negated items (that is, ~something) in unsigned ints. Some compilers
2663 apparently mutter when there is no cast.
2665 70. If an address verification called from an ACL failed, and did not produce a
2666 user-specific message (i.e. there was only a "system" message), nothing was
2667 put in $acl_verify_message. In this situation, it now puts the system
2670 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2671 session; change 4.31/43 added the unwanted input to the log line - except
2672 that it did not do this in the start of session case. It now does.
2674 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2675 This is wrong and can cause the other end to generate a synchronization
2676 error if it is another Exim or anything else that does the synchronization
2677 check. A QUIT command is no longer sent after a timeout.
2679 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2680 during host lookups.
2682 74. The defer_ok option of callout verification was not working if it was used
2683 when verifying addresses in header lines, that is, for this case:
2685 verify = header_sender/callout=defer_ok
2687 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2688 those file descriptors could be used for SMTP connections. If anything
2689 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2690 could be sent to the SMTP client, causing chaos. The daemon now opens
2691 stdin, stdout, and stderr to /dev/null when it puts itself into the
2694 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2695 log by default. The user can override this with suitable Perl magic.
2697 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2698 the log message when discard triggers, was not working for the DATA ACL or
2699 for the non-SMTP ACL.
2701 78. Error message wording change in sieve.c.
2703 79. If smtp_accept_max_per_host was set, the number of connections could be
2704 restricted to fewer than expected, because the daemon was trying to set up
2705 a new connection before checking whether the processes handling previous
2706 connections had finished. The check for completed processes is now done
2707 earlier. On busy systems, this bug wouldn't be noticed because something
2708 else would have woken the daemon, and it would have reaped the completed
2711 80. If a message was submitted locally by a user whose login name contained one
2712 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2713 It caused a spool format error. I have fixed the spool reading code. A
2714 related problem was that the "from" clause in the Received: line became
2715 illegal because of the space(s). It is now covered by ${quote_local_part.
2717 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2720 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2722 83. Patch from Sieve maintainer to fix unterminated string problem in
2723 "vacation" handling.
2725 84. Some minor changes to the Linux configuration files to help with other
2726 OS variants using glibc.
2728 85. One more patch for Sieve to update vacation handling to latest spec.
2731 ----------------------------------------------------
2732 See the note above about the 4.44 and 4.50 releases.
2733 ----------------------------------------------------
2739 1. Change 4.43/35 introduced a bug that caused file counts to be
2740 incorrectly computed when quota_filecount was set in an appendfile
2743 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2744 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2747 3. Give more explanation in the error message when the command for a transport
2748 filter fails to execute.
2750 4. There are several places where Exim runs a non-Exim command in a
2751 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2752 was being done only for the command run by the queryprogram router. It is
2753 now done for all such subprocesses. The other cases are: ${run, transport
2754 filters, and the commands run by the lmtp and pipe transports.
2756 5. Some older OS have a limit of 256 on the maximum number of file
2757 descriptors. Exim was using setrlimit() to set 1000 as a large value
2758 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2759 systems. I've change it so that if it can't get 1000, it tries for 256.
2761 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2762 was an oversight, and furthermore, ever since the addition of extra
2763 controls (e.g. 4.43/32), the checks on when to allow different forms of
2764 "control" were broken. There should now be diagnostics for all cases when a
2765 control that does not make sense is encountered.
2767 7. $recipients is now available in the predata ACL (oversight).
2769 8. Tidy the search cache before the fork to do a delivery from a message
2770 received from the command line. Otherwise the child will trigger a lookup
2771 failure and thereby defer the delivery if it tries to use (for example) a
2772 cached ldap connection that the parent has called unbind on.
2774 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2775 of $address_data from the recipient verification was clobbered by the
2776 sender verification.
2778 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2779 was its contents. (It was OK if the option was not defined at all.)
2781 11. A "Completed" log line is now written for messages that are removed from
2782 the spool by the -Mrm option.
2784 12. $host_address is now set to the target address during the checking of
2785 ignore_target_hosts.
2787 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2788 being passed; this would have caused $sender_host_name to have been used if
2789 matching the list had actually called for a host name (not very likely,
2790 since this list is usually IP addresses). A host name is now passed as
2793 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2794 code that specifies a non-restarting handler (typically sigaction() in
2795 modern systems) in an attempt to fix a rare and obscure crash bug.
2797 15. Narrowed the window for a race in the daemon that could cause it to ignore
2798 SIGCHLD signals. This is not a major problem, because they are used only to
2799 wake it up if nothing else does.
2801 16. A malformed maildirsize file could cause Exim to calculate negative values
2802 for the mailbox size or file count. Odd effects could occur as a result.
2803 The maildirsize information is now recalculated if the size or filecount
2806 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2807 support for a long time. Removed HAVE_SYS_VFS_H.
2809 18. Updated exipick to current release from John Jetmore.
2811 19. Allow an empty sender to be matched against a lookup in an address list.
2812 Previously the only cases considered were a regular expression, or an
2815 20. Exim went into a mad DNS lookup loop when doing a callout where the
2816 host was specified on the transport, if the DNS lookup yielded more than
2819 21. The RFC2047 encoding function was originally intended for short strings
2820 such as real names; it was not keeping to the 75-character limit for
2821 encoded words that the RFC imposes. It now respects the limit, and
2822 generates multiple encoded words if necessary. To be on the safe side, I
2823 have increased the buffer size for the ${rfc2047: expansion operator from
2826 22. Failure to deliver a bounce message always caused it to be frozen, even if
2827 there was an errors_to setting on the router. The errors_to setting is now
2830 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2831 canonical form (fully expanded) before being placed in
2832 $sender_host_address.
2834 24. Updated eximstats to version 1.33
2836 25. Include certificate and key file names in error message when GnuTLS fails
2837 to set them up, because the GnuTLS error message doesn't include the name
2838 of the failing file when there is a problem reading it.
2840 26. Expand error message when OpenSSL has problems setting up cert/key files.
2843 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2844 (this can affect the format of dates).
2846 28. exim_tidydb, when checking for the continued existence of a message for
2847 which it has found a message-specific retry record, was not finding
2848 messages that were in split spool directories. Consequently, it was
2849 deleting retry records that should have stayed in existence.
2851 29. eximstats updated to version 1.35
2852 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2853 1.35 - bugfix such that pie charts by volume are generated correctly
2855 30. The SPA authentication driver was not abandoning authentication and moving
2856 on to the next authenticator when an expansion was forced to fail,
2857 contradicting the general specification for all authenticators. Instead it
2858 was generating a temporary error. It now behaves as specified.
2860 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2861 (the order specifies the preference for clients). The order is now AES256,
2862 AES128, 3DES, ARCFOUR128.
2864 31. Small patch to Sieve code - explicitly set From: when generating an
2867 32. Exim crashed if a remote delivery caused a very long error message to be
2868 recorded - for instance if somebody sent an entire SpamAssassin report back
2869 as a large number of 550 error lines. This bug was coincidentally fixed by
2870 increasing the size of one of Exim's internal buffers (big_buffer) that
2871 happened as part of the Exiscan merge. However, to be on the safe side, I
2872 have made the code more robust (and fixed the comments that describe what
2875 33. Some experimental protocols are using DNS PTR records for new purposes. The
2876 keys for these records are domain names, not reversed IP addresses. The
2877 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2878 leaves it alone. Component reversal etc. now happens only for IP addresses.
2881 34. The host_aton() function is supposed to be passed a string that is known
2882 to be a valid IP address. However, in the case of IPv6 addresses, it was
2883 not checking this. This is a hostage to fortune. Exim now panics and dies
2884 if the condition is not met. A case was found where this could be provoked
2885 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2886 components; fortuitously, this particular loophole had already been fixed
2887 by change 4.50/55 or 4.44/33 above.
2889 If there are any other similar loopholes, the new check in host_aton()
2890 itself should stop them being exploited. The report I received stated that
2891 data on the command line could provoke the exploit when Exim was running as
2892 exim, but did not say which command line option was involved. All I could
2893 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2894 running as the user.
2897 35. There was a buffer overflow vulnerability in the SPA authentication code
2898 (which came originally from the Samba project). I have added a test to the
2899 spa_base64_to_bits() function which I hope fixes it.
2902 36. The daemon start-up calls getloadavg() while still root for those OS that
2903 need the first call to be done as root, but it missed one case: when
2904 deliver_queue_load_max is set with deliver_drop_privilege. This is
2905 necessary for the benefit of the queue runner, because there is no re-exec
2906 when deliver_drop_privilege is set.
2908 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2909 was in use, was not putting the data itself into the right store pool;
2910 consequently, it could be overwritten for a subsequent message in the same
2911 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2914 38. Sometimes the final signoff response after QUIT could fail to get
2915 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2916 < 0 before doing a fflush(). This bug looks as though it goes back to the
2917 introduction of TLS in release 3.20, but "sometimes" must have been rare
2918 because the tests only now provoked it.
2924 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2925 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2926 Now it returns an integer. A number of calls were still expecting a T/F
2927 return. Fortuitously, in all cases, the tests worked in OK situations,
2928 which is the norm. However, things would have gone wrong on any write
2929 failures on the smtp file descriptor. This function is used when sending
2930 messages over SMTP and also when doing verify callouts.
2932 2. When Exim is called to do synchronous delivery of a locally submitted
2933 message (the -odf or -odi options), it no longer closes stderr before doing
2936 3. Implemented the mua_wrapper option.
2938 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2940 5. Implemented the functions header_remove(), header_testname(),
2941 header_add_at_position(), and receive_remove_recipient(), and exported them
2944 6. If an ACL "warn" statement specified the addition of headers, Exim already
2945 inserted X-ACL-Warn: at the start if there was no header name. However, it
2946 was not making this test for the second and subsequent header lines if
2947 there were newlines in the string. This meant that an invalid header could
2948 be inserted if Exim was badly configured.
2950 7. Allow an ACL "warn" statement to add header lines at the start or after all
2951 the Received: headers, as well as at the end.
2953 8. Added the rcpt_4xx retry error code.
2955 9. Added postmaster_mailfrom=xxx to callout verification option.
2957 10. Added mailfrom=xxxx to the callout verification option, for verify=
2960 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2961 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2963 12. Inserted some casts to stop certain compilers warning when using pointer
2964 differences as field lengths or precisions in printf-type calls (mostly
2965 affecting debugging statements).
2967 13. Added optional readline() support for -be (dynamically loaded).
2969 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2970 same clock tick as a message's arrival, so that its received time was the
2971 same as the "first fail" time on the retry record, and that message
2972 remained on the queue past the ultimate address timeout, every queue runner
2973 would try a delivery (because it was past the ultimate address timeout) but
2974 after another failure, the ultimate address timeout, which should have then
2975 bounced the address, did not kick in. This was a "< instead of <=" error;
2976 in most cases the first failure would have been in the next clock tick
2977 after the received time, and all would be well.
2979 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2980 being recognized when the domain list was tested by the match_domain
2981 condition in an expansion string.
2983 16. Added the ${str2b64: operator.
2985 17. Exim was always calling setrlimit() to set a large limit for the number of
2986 processes, without checking whether the existing limit was already
2987 adequate. (It did check for the limit on file descriptors.) Furthermore,
2988 errors from getrlimit() and setrlimit() were being ignored. Now they are
2989 logged to the main and panic logs, but Exim does carry on, to try to do its
2990 job under whatever limits there are.
2992 18. Imported PCRE 5.0.
2994 19. Trivial typo in log message " temporarily refused connection" (the leading
2997 20. If the log selector return_path_on_delivery was set and an address was
2998 redirected to /dev/null, the delivery process crashed because it assumed
2999 that a return path would always be set for a "successful" delivery. In this
3000 case, the whole delivery is bypassed as an optimization, and therefore no
3003 21. Internal re-arrangement: the function for sending a challenge and reading
3004 a response while authentication was assuming a zero-terminated challenge
3005 string. It's now changed to take a pointer and a length, to allow for
3006 binary data in such strings.
3008 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3010 23. Exim was not respecting finduser_retries when seeking the login of the
3011 uid under which it was called; it was always trying 10 times. (The default
3012 setting of finduser_retries is zero.) Also, it was sleeping after the final
3013 failure, which is pointless.
3015 24. Implemented tls_on_connect_ports.
3017 25. Implemented acl_smtp_predata.
3019 26. If the domain in control=submission is set empty, Exim assumes that the
3020 authenticated id is a complete email address when it generates From: or
3021 Sender: header lines.
3023 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3024 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3025 chown and chgrp in /bin and hostname in /usr/bin.
3027 28. Exim was keeping the "process log" file open after each use, just as it
3028 does for the main log. This opens the possibility of it remaining open for
3029 long periods when the USR1 signal hits a daemon. Occasional processlog
3030 errors were reported, that could have been caused by this. Anyway, it seems
3031 much more sensible not to leave this file open at all, so that is what now
3034 29. The long-running daemon process does not normally write to the log once it
3035 has entered its main loop, and it closes the log before doing so. This is
3036 so that log files can straightforwardly be renamed and moved. However,
3037 there are a couple of unusual error situations where the daemon does write
3038 log entries, and I had neglected to close the log afterwards.
3040 30. The text of an SMTP error response that was received during a remote
3041 delivery was being truncated at 512 bytes. This is too short for some of
3042 the long messages that one sometimes sees. I've increased the limit to
3045 31. It is now possible to make retry rules that apply only when a message has a
3046 specific sender, in particular, an empty sender.
3048 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3049 it possible to be selective about when SMTP synchronization is enforced.
3051 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3053 32. Implemented hosts_connection_nolog.
3055 33. Added an ACL for QUIT.
3057 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3060 35. Added mailbox_size and mailbox_filecount to appendfile.
3062 36. Added control = no_multiline_responses to ACLs.
3064 37. There was a bug in the logic of the code that waits for the clock to tick
3065 in the case where the clock went backwards by a substantial amount such
3066 that the microsecond fraction of "now" was more than the microsecond
3067 fraction of "then" (but the whole seconds number was less).
3069 38. Added support for the libradius Radius client library this is found on
3070 FreeBSD (previously only the radiusclient library was supported).
3076 1. When certain lookups returned multiple values in the form name=value, the
3077 quoting of the values was not always being done properly. Specifically:
3078 (a) If the value started with a double quote, but contained no whitespace,
3080 (b) If the value contained whitespace other than a space character (i.e.
3081 tabs or newlines or carriage returns) it was not quoted.
3082 This fix has been applied to the mysql and pgsql lookups by writing a
3083 separate quoting function and calling it from the lookup code. The fix
3084 should probably also be applied to nisplus, ibase and oracle lookups, but
3085 since I cannot test any of those, I have not disturbed their existing code.
3087 2. A hit in the callout cache for a specific address caused a log line with no
3088 reason for rejecting RCPT. Now it says "Previous (cached) callout
3089 verification failure".
3091 3. There was an off-by-one bug in the queryprogram router. An over-long
3092 return line was truncated at 256 instead of 255 characters, thereby
3093 overflowing its buffer with the terminating zero. As well as fixing this, I
3094 have increased the buffer size to 1024 (and made a note to document this).
3096 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3097 when Exim is waiting for an SMTP response from a remote server, Exim
3098 restarts its select() call on the socket, thereby resetting its timeout.
3099 This is not a problem when such interrupts are rare. Somebody set up a cron
3100 job to run exiwhat every 2 minutes, which is less than the normal select()
3101 timeout (5 or 10 minutes). This meant that the select() timeout never
3102 kicked in because it was always reset. I have fixed this by comparing the
3103 time when an interrupt arrives with the time at the start of the first call
3104 to select(). If more time than the timeout has elapsed, the interrupt is
3105 treated as a timeout.
3107 5. Some internal re-factoring in preparation for the addition of Sieve
3108 extensions (by MH). In particular, the "personal" test is moved to a
3109 separate function, and given an option for scanning Cc: and Bcc: (which is
3110 not set for Exim filters).
3112 6. When Exim created an email address using the login of the caller as the
3113 local part (e.g. when creating a From: or Sender: header line), it was not
3114 quoting the local part when it contained special characters such as @.
3116 7. Installed new OpenBSD configuration files.
3118 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3119 try to make them clearer.
3121 9. Callout options, other than the timeout value, were being ignored when
3122 verifying sender addresses in header lines. For example, when using
3124 verify = header_sender/callout=no_cache
3126 the cache was (incorrectly) being used.
3128 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3129 only the Cygwin environment.
3131 11. When return_path_on_delivery was set as a log selector, if different remote
3132 addresses in the same message used different return paths and parallel
3133 remote delivery occurred, the wrong values would sometimes be logged.
3134 (Whenever a remote delivery process finished, the return path value from
3135 the most recently started remote delivery process was logged.)
3137 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3138 lines when AUTH and/or TLS are in use. This is the "received protocol"
3139 field. Exim used to use "asmtp" for authenticated SMTP, without any
3140 indication (in the protocol name) for TLS use. Now it follows the RFC and
3141 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3142 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3143 names appear in log lines as well as in Received: header lines.
3145 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3146 extensions, and comparison tests, and to fix some bugs.
3148 14. Changes to the "personal" filter test:
3150 (1) The test was buggy in that it was just doing the equivalent of
3151 "contains" tests on header lines. For example, if a user's address was
3152 anne@some.where, the "personal" test would incorrectly be true for
3154 To: susanne@some.where
3156 This test is now done by extracting each address from the header in turn,
3157 and checking the entire address. Other tests that are part of "personal"
3158 are now done using regular expressions (for example, to check local parts
3159 of addresses in From: header lines).
3161 (2) The list of non-personal local parts in From: addresses has been
3162 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3163 taken from the Sieve specification recommendations.
3165 (3) If the message contains any header line starting with "List-" it is
3166 treated as non-personal.
3168 (4) The test for "circular" in the Subject: header line has been removed
3169 because it now seems ill-conceived.
3171 15. Minor typos in src/EDITME comments corrected.
3173 16. Installed latest exipick from John Jetmore.
3175 17. If headers_add on a router specified a text string that was too long for
3176 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3177 of string_sprintf() is now avoided.
3179 18. $message_body_size was not set (it was always zero) when running the DATA
3180 ACL and the local_scan() function.
3182 19. For the "mail" command in an Exim filter, no default was being set for
3183 the once_repeat time, causing a random time value to be used if "once" was
3184 specified. (If the value happened to be <= 0, no repeat happened.) The
3185 default is now 0s, meaning "never repeat". The "vacation" command was OK
3186 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3187 (I found it when inspecting the code).
3189 20. There is now an overall timeout for performing a callout verification. It
3190 defaults to 4 times the callout timeout, which applies to individual SMTP
3191 commands during the callout. The overall timeout applies when there is more
3192 than one host that can be tried. The timeout is checked before trying the
3193 next host. This prevents very long delays if there are a large number of
3194 hosts and all are timing out (e.g. when the network connections are timing
3195 out). The value of the overall timeout can be changed by specifying an
3196 additional sub-option for "callout", called "maxwait". For example:
3198 verify = sender/callout=5s,maxwait=20s
3200 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3201 to the end before writing, but this should make it even safer).
3203 22. Exim was forgetting that it had advertised PIPELINING for the second and
3204 subsequent messages on an SMTP connection. It was also not resetting its
3205 memory on STARTTLS and an internal HELO.
3207 23. When Exim logs an SMTP synchronization error within a session, it now
3208 records whether PIPELINING has been advertised or not.
3210 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3211 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3212 rather than long int.
3214 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3216 26. Added the never_mail option to autoreply.
3222 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3223 crash if the getsockname() call failed; this can happen if a connection is
3224 closed very soon after it is established. The problem was simply in the
3225 order in which certain operations were done, causing Exim to try to write
3226 to the SMTP stream before it had set up the file descriptor. The bug has
3227 been fixed by making things happen in the correct order.
3233 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3234 before the connection was closed, thus losing the rejection response.
3236 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3237 some early Solaris releases, but causes trouble in current releases where
3238 socklen_t is defined.
3240 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3243 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3246 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3247 that releasing the top of it at the end releases what was used for sub-
3248 expansions (unless the block got too big). However, discard this block if
3249 the first thing is a variable or header, so that we can use its block when
3250 it is dynamic (useful for very large $message_headers, for example).
3252 6. Lookups now cache *every* query, not just the most recent. A new, separate
3253 store pool is used for this. It can be recovered when all lookup caches are
3254 flushed. Lookups now release memory at the end of their result strings.
3255 This has involved some general refactoring of the lookup sources.
3257 7. Some code has been added to the store_xxx() functions to reduce the amount
3258 of flapping under certain conditions.
3260 8. log_incoming_interface used to affect only the <= reception log lines. Now
3261 it causes the local interface and port to be added to several more SMTP log
3262 lines, for example "SMTP connection from", and rejection lines.
3264 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3266 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3268 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3269 could be overwritten at the end of the current message (or the start of a
3270 new message if it was set in a HELO ACL). The value is now preserved for
3271 the duration of the SMTP connection.
3273 12. If a transport had a headers_rewrite setting, and a matching header line
3274 contained an unqualified address, that address was qualified, even if it
3275 did not match any rewriting rules. The underlying bug was that the values
3276 of the flags that permit the existence of unqualified sender and recipient
3277 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3278 non-local messages, and by -bnq for local messages) were not being
3279 preserved with the message after it was received.
3281 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3282 "next input=" as part of the text comprising the host identity instead of
3283 the correct text. The code was using the same buffer for two different
3284 strings. However, depending on which order the printing function evaluated
3285 its arguments, the bug did not always show up. Under Linux, for example, my
3286 test suite worked just fine.
3288 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3289 doesn't work with some older versions of Perl. It has been changed to "my",
3290 which in any case is probably the better facility to use.
3292 15. A really picky compiler found some instances of statements for creating
3293 error messages that either had too many or two few arguments for the format
3296 16. The size of the buffer for calls to the DNS resolver has been increased
3297 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3298 for addresses that have a lot of PTR records. This alleviates a problem; it
3299 does not fully solve it.
3301 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3302 buffer now truncates the list and logs the incident, which is the same
3303 action as happens when Exim is looking up a host name and its aliases.
3304 Previously in this situation something unpredictable would happen;
3305 sometimes it was "internal error: store_reset failed".
3307 18. If a server dropped the connection unexpectedly when an Exim client was
3308 using GnuTLS and trying to read a response, the client delivery process
3309 crashed while trying to generate an error log message.
3311 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3312 string, for example:
3314 warn message = H1: something\nH2: something
3316 the text was added as a single header line from Exim's point of view
3317 though it ended up OK in the delivered message. However, searching for the
3318 second and subsequent header lines using $h_h2: did not work. This has been
3319 fixed. Similarly, if a system filter added multiple headers in this way,
3320 the routers could not see them.
3322 20. Expanded the error message when iplsearch is called with an invalid key to
3323 suggest using net-iplsearch in a host list.
3325 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3326 ACLs are no longer actually imposed (and a message to that effect is
3329 22. If a "gecos" field in a passwd entry contained escaped characters, in
3330 particular, if it contained a \" sequence, Exim got it wrong when building
3331 a From: or a Sender: header from that name. A second bug also caused
3332 incorrect handling when an unquoted " was present following a character
3333 that needed quoting.
3335 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3336 was not being matched caselessly.
3338 24. Arranged for all hyphens in the exim.8 source to be escaped with
3341 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3342 properly. Recipient callout cache records were still being keyed to include
3343 the sender, even when use_sender was set false. This led to far more
3344 callouts that were necessary. The sender is no longer included in the key
3345 when use_sender is false.
3347 26. Added "control = submission" modifier to ACLs.
3349 27. Added the ${base62d: operator to decode base 62 numbers.
3351 28. dnsdb lookups can now access SRV records.
3353 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3354 the configuration file.
3356 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3357 (-v) mode. This makes the output for a verbose queue run more intelligible.
3359 31. Added a use_postmaster feature to recipient callouts.
3361 32. Added the $body_zerocount variable, containing the number of binary zero
3362 bytes in the message body.
3364 33. The time of last modification of the "new" subdirectory is now used as the
3365 "mailbox time last read" when there is a quota error for a maildir
3368 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3370 35. Added +ignore_unknown as a special item in host lists.
3372 36. Code for decoding IPv6 addresses in host lists is now included, even if
3373 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3374 address was recognized as an IP address, but was then not correctly decoded
3375 into binary, causing unexpected and incorrect effects when compared with
3382 1. Very minor rewording of debugging text in manualroute to say "list of
3383 hosts" instead of "hostlist".
3385 2. If verify=header_syntax was set, and a header line with an unqualified
3386 address (no domain) and a large number of spaces between the end of the
3387 name and the colon was received, the reception process suffered a buffer
3388 overflow, and (when I tested it) crashed. This was caused by some obsolete
3389 code that should have been removed. The fix is to remove it!
3391 3. When running in the test harness, delay a bit after writing a bounce
3392 message to get a bit more predictability in the log output.
3394 4. Added a call to search_tidyup() just before forking a reception process. In
3395 theory, someone could use a lookup in the expansion of smtp_accept_max_
3396 per_host which, without the tidyup, could leave open a database connection.
3398 5. Added the variables $recipient_data and $sender_data which get set from a
3399 lookup success in an ACL "recipients" or "senders" condition, or a router
3400 "senders" option, similar to $domain_data and $local_part_data.
3402 6. Moved the writing of debug_print from before to after the "senders" test
3405 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3406 problems for message scanning, either using a data ACL, or using
3407 local_scan() because the Received: header was not generated till after they
3408 were called (in order to set the time as the time of reception completion).
3409 I have revised the way this works. The header is now generated after the
3410 body is received, but before the ACL or local_scan() are called. After they
3411 are run, the timestamp in the header is updated.
3417 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3418 before starting a queue runner without re-exec. This happened only when
3419 deliver_drop_privilege was set or when the Exim user was set to root. The
3420 effect of the bug was that timeouts during subsequent deliveries caused
3421 crashes instead of being properly handled. The handler is now left at its
3422 default (and expected) setting.
3424 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3425 message, again when deliver_drop_privilege is set or Exim is run as root.
3426 The bug described in (1) was not present in this case, but the tidying up
3427 of the other signals was missing. I have made the two cases consistent.
3429 3. The ignore_target_hosts setting on a manualroute router was being ignored
3430 for hosts that were looked up using the /MX notation.
3432 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3435 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3436 operated on the sender address. After changing the $sender_address to <>
3437 for the sender address verify, Exim was re-instated it as the original
3438 (before rewriting) address, but remembering that it had rewritten it, so it
3439 wasn't rewriting it again. This bug also had the effect of breaking the
3440 sender address verification caching when the sender address was rewritten.
3442 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3443 This has been changed so that if the ip literal address matches
3444 ignore_target_hosts, the router declines.
3446 7. Added expansion conditions match_domain, match_address, and match_local_
3447 part (NOT match_host).
3449 8. The placeholder for the Received: header didn't have a length field set.
3451 9. Added code to Exim itself and to exim_lock to test for a specific race
3452 condition that could lead to file corruption when using MBX delivery. The
3453 issue is with the lockfile that is created in /tmp. If this file is removed
3454 after a process has opened it but before that process has acquired a lock,
3455 there is the potential for a second process to recreate the file and also
3456 acquire a lock. This could lead to two Exim processes writing to the file
3457 at the same time. The added code performs the same test as UW imapd; it
3458 checks after acquiring the lock that its file descriptor still refers to
3459 the same named file.
3461 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3462 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3465 11. Added the smtp_active_hostname option. If used, this will typically be made
3466 to depend on the incoming interface address. Because $interface_address is
3467 not set up until the daemon has forked a reception process, error responses
3468 that can happen earlier (such as "too many connections") no longer contain
3471 12. If an expansion in a condition on a "warn" statement fails because a lookup
3472 defers, the "warn" statement is abandoned, and the next ACL statement is
3473 processed. Previously this caused the whole ACL to be aborted.
3475 13. Added the iplsearch lookup type.
3477 14. Added ident_timeout as a log selector.
3479 15. Added tls_certificate_verified as a log selector.
3481 16. Added a global option tls_require_ciphers (compare the smtp transport
3482 option of the same name). This controls incoming TLS connections.
3484 17. I finally figured out how to make tls_require_ciphers do a similar thing
3485 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3486 before starting the TLS session.
3488 18. Tabs are now shown as \t in -bP output.
3490 19. If the log selector return_path_on_delivery was set, Exim crashed when
3491 bouncing a message because it had too many Received: header lines.
3493 20. If two routers both had headers_remove settings, and the first one included
3494 a superfluous trailing colon, the final name in the first list and the
3495 first name in the second list were incorrectly joined into one item (with a
3496 colon in the middle).
3502 1. Added -C and -D options to the exinext utility, mainly to make it easier
3503 to include in the automated testing, but these could be helpful when
3504 multiple configurations are in use.
3506 2. The exinext utility was not formatting the output nicely when there was
3507 an alternate port involved in the retry record key, nor when there was a
3508 message id as well (for retries that were specific to a specific message
3509 and a specific host). It was also confused by IPv6 addresses, because of
3510 the additional colons they contain. I have fixed the IPv4 problem, and
3511 patched it up to do a reasonable job for IPv6.
3513 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3514 delivery, the log line now contains "pipelined" if PIPELINING was used.
3516 4. An SMTP transport process used to panic and die if the bind() call to set
3517 an explicit outgoing interface failed. This has been changed; it is now
3518 treated in the same way as a connect() failure.
3520 5. A reference to $sender_host_name in the part of a conditional expansion
3521 that was being skipped was still causing a DNS lookup. This no longer
3524 6. The def: expansion condition was not recognizing references to header lines
3525 that used bh_ and bheader_.
3527 7. Added the _cache feature to named lists.
3529 8. The code for checking quota_filecount in the appendfile transport was
3530 allowing one more file than it should have been.
3532 9. For compatibility with Sendmail, the command line option
3540 and sets the incoming protocol and host name (for trusted callers). The
3541 host name and its colon can be omitted when only the protocol is to be set.
3542 Note the Exim already has two private options, -pd and -ps, that refer to
3543 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3544 "s", but I don't think that's a major issue.
3546 10. A number of refactoring changes to the code, none of which should affect
3549 (a) The number of logging options was getting close to filling up the
3550 32-bit word that was used as a bit map. I have split them into two classes:
3551 those that are passed in the argument to log_write(), and those that are
3552 only ever tested independently outside of that function. These are now in
3553 separate 32-bit words, so there is plenty of room for expansion again.
3554 There is no change in the user interface or the logging behaviour.
3556 (b) When building, for example, log lines, the code previously used a
3557 macro that called string_cat() twice, in order to add two strings. This is
3558 not really sufficiently general. Furthermore, there was one instance where
3559 it was actually wrong because one of the argument was used twice, and in
3560 one call a function was used. (As it happened, calling the function twice
3561 did not affect the overall behaviour.) The macro has been replaced by a
3562 function that can join an arbitrary number of extra strings onto a growing
3565 (c) The code for expansion conditions now uses a table and a binary chop
3566 instead of a serial search (which was left over from when there were very
3567 few conditions). Also, it now recognizes conditions like "pam" even when
3568 the relevant support is not compiled in: a suitably worded error message is
3569 given if an attempt is made to use such a condition.
3571 11. Added ${time_interval:xxxxx}.
3573 12. A bug was causing one of the ddress fields not to be passed back correctly
3574 from remote delivery subprocesses. The field in question was not being
3575 subsequently used, so this caused to problems in practice.
3577 13. Added new log selectors queue_time and deliver_time.
3579 14. Might have fixed a bug in maildirsizefile handling that threw up
3580 "unexpected character" debug warnings, and recalculated the data
3581 unnecessarily. In any case, I expanded the warning message to give more
3584 15. Added the message "Restricted characters in address" to the statements in
3585 the default ACL that block characters like @ and % in local parts.
3587 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3588 Three changes have been made:
3590 (a) There was a serious bug; a negative response to MAIL caused the whole
3591 recipient domain to be cached as invalid, thereby blocking all messages
3592 to all local parts at the same domain, from all senders. This bug has
3593 been fixed. The domain is no longer cached after a negative response to
3594 MAIL if the sender used is not empty.
3596 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3599 (c) A new callout option, "use_sender" has been added for people who want
3600 the modified behaviour.
3606 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3609 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3610 indeed breaks things for older releases.
3612 3. Added additional logging to the case where there is a problem reading data
3613 from a filter that is running in a subprocess using a pipe, in order to
3614 try to track down a specific problem.
3616 4. Testing facility fudge: when running in the test harness and attempting
3617 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3618 getting "No route to host". Convert this to a timeout.
3620 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3623 6. Some OS don't have socklen_t but use size_t instead. This affects the
3624 fifth argument of getsockopt() amongst other things. This is now
3625 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3626 can be set for individual OS. I have set it for SunOS5, OSF1, and
3627 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3628 some earlier ones do not.
3630 7. Change 4.30/15 was not doing the test caselessly.
3632 8. The standard form for an IPv6 address literal was being rejected by address
3633 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3634 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3635 this, as well as the form without the "IPv6" on the front (but only when
3636 address literals are enabled, of course).
3638 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3640 10. Exim crashed if a message with an empty sender address specified by -f
3641 encountered a router with an errors_to setting. This could be provoked only
3642 by a command such as
3646 where an empty string was supplied; "<>" did not hit this bug.
3648 11. Installed PCRE release 4.5.
3650 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3651 remained set. It is now erased.
3653 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3654 times from message ids (which are base 36 rather than the normal 62).
3656 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3657 were being counted as actual protocol errors, and logged if the log
3658 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3659 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3660 and DATA following a set of rejected RCPTs do not count as protocol errors.
3661 In other words, Exim assumes they were pipelined, though this may not
3662 actually be the case. Of course, in all cases the client gets an
3663 appropriate error code.
3665 15. If a lookup fails in an ACL condition, a message about the failure may
3666 be available; it is used if testing the ACL cannot continue, because most
3667 such messages specify what the cause of the deferral is. However, some
3668 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3669 that caused an old message to be retained and used if a later statement
3670 caused a defer, replacing the real cause of the deferral.
3672 16. If an IP address had so many PTR records that the DNS lookup buffer
3673 was not large enough to hold them, Exim could crash while trying to process
3674 the truncated data. It now detects and logs this case.
3676 17. Further to 4.21/58, another change has been made: if (and only if) the
3677 first line of a message (the first header line) ends with CRLF, a bare LF
3678 in a subsequent header line has a space inserted after it, so as not to
3679 terminate the header.
3681 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3682 unnecessarily, used atoi() instead of strtol(), and didn't check the
3683 termination when getting file sizes from file names by regex.
3685 19. Completely re-implemented the support for maildirsize files, in the light
3686 of a number of problems with the previous contributed implementation
3687 (4.30/29). In particular:
3689 . If the quota is zero, the maildirsize file is maintained, but no quota is
3692 . If the maildir directory does not exist, it is created before any attempt
3693 to write a maildirsize file.
3695 . The quota value in the file is just a cache; if the quota is changed in
3696 the transport, the new value overrides.
3698 . A regular expression is available for excluding directories from the
3701 20. The autoreply transport checks the characters in options that define the
3702 message's headers; it allows continued headers, but it was checking with
3703 isspace() after an embedded newline instead of explicitly looking for a
3706 21. If all the "regular" hosts to which an address was routed had passed their
3707 expiry times, and had not reached their retry times, the address was
3708 bounced, even if fallback hosts were defined. Now Exim should go on to try
3711 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3712 equivalent code in the SMTP transport. Some hosts send humungous responses
3713 to HELO/EHLO, more than 1024 it seems.
3715 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3716 gives compiler warnings in some environments. I've now done it "properly",
3719 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3720 (because of problems with the built-in one) was declared to return uschar *
3721 instead of char *, causing compiler failure.
3723 25. Fixed a file descriptor leak when processing alias/forward files.
3725 26. Fixed a minor format string issue in dbfn.c.
3727 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3729 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3730 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3733 29. When a sender address is verified, it is cached, to save repeating the test
3734 when there is more than one recipient in a message. However, when the
3735 verification involves a callout, it is possible for different callout
3736 options to be set for different recipients. It is too complicated to keep
3737 track of this in the cache, so now Exim always runs a verification when a
3738 callout is required, relying on the callout cache for the optimization.
3739 The overhead is duplication of the address routing, but this should not be
3742 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3743 to be verified with callout=postmaster, and the main callout worked but the
3744 postmaster check failed, the verification correctly failed. However, if a
3745 subsequent RCPT command asked for sender verification *without* the
3746 postmaster check, incorrect caching caused this verification also to fail,
3749 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3750 it was not caching the DNS options (qualify_single, search_parents) that
3751 were used when the lookup failed. A subsequent lookup with different
3752 options therefore always gave the same answer, though there were cases
3753 where it should not have. (Example: a "domains = !$mx_any" option on a
3754 dnslookup router: the "domains" option is always processed without any
3755 widening, but the router might have qualify_single set.) Now Exim uses the
3756 cached value only when the same options are set.
3758 32. Added John Jetmore's "exipick" utility to the distribution.
3760 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3761 than a timeout (e.g. a certificate is required, and is not provided), an
3762 Exim server now closes the connection immediately. Previously it waited for
3763 the client to close - but if the client is SSL, it seems that they each
3764 wait for each other, leading to a delay before one of them times out.
3766 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3767 maintained 0.8.x compatibility because I don't think many are using it, and
3768 it is clearly obsolete.
3770 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3773 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3774 client certificate was expired. A simple patch fixes this, though I don't
3775 understand the full logic of why the verify callback is called multiple
3778 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3779 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3780 which causes problems with some clients (such as the Certicom SSL Plus
3781 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3782 disables the coutermeasure allowing Eudora to connect."
3784 38. Exim was not checking that a write() to a log file succeeded. This could
3785 lead to Bad Things if a log got too big, in particular if it hit a file
3786 size limit. Exim now panics and dies if it cannot write to a log file, just
3787 as it does if it cannot open a log file.
3789 39. Modified OS/Makefile-Linux so that it now contains
3791 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3793 The two -D definitions ensure that Exim is compiled with large file
3794 support, which makes it possible to handle log files that are bigger than
3797 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3798 instance) a domain was checked against a named list that involved a lookup,
3799 causing $domain_data to be set, then another domain was checked against the
3800 same list, then the first domain was re-checked, the value of $domain_data
3801 after the final check could be wrong. In particular, if the second check
3802 failed, it could be set empty. This bug probably also applied to
3805 41. The strip_trailing_dot option was not being applied to the address given
3806 with the -f command-line option.
3808 42. The code for reading a message's header from the spool was incrementing
3809 $received_count, but never initializing it. This meant that the value was
3810 incorrect (doubled) while delivering a message in the same process in which
3811 it was received. In the most common configuration of Exim, this never
3812 happens - a fresh exec is done - but it can happen when
3813 deliver_drop_privilege is set.
3815 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3816 it now includes up to 150 characters of the unexpected data in the log
3819 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3820 and building data strings. The size of both of these buffers was 10 000
3821 bytes - far larger than anybody would *ever* want, thought I. Needless to
3822 say, somebody hit the limit. I have increased the maximum line length to
3823 20 000 and the maximum data length of concatenated lines to 100 000. I have
3824 also fixed two bugs, because there was no checking on these buffers. Tsk,
3825 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3826 buffer is too small.
3828 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3829 lsearch lookups. Now it does.
3831 46. When parsing a route_list item in a manualroute router, a fixed-length
3832 buffer was used for the list of hosts. I made this 1024 bytes long,
3833 thinking that nobody would ever have a list of hosts that long. Wrong.
3834 Somebody had a whole pile of complicated expansion conditions, and the
3835 string was silently truncated, leading to an expansion error. It turns out
3836 that it is easier to change to an unlimited length (owing to other changes
3837 that have happened since this code was originally written) than to build
3838 structure for giving a limitation error. The length of the item that
3839 expands into the list of hosts is now unlimited.
3841 47. The lsearch lookup could not handle data where the length of text line was
3842 more than 4095 characters. Such lines were truncated, leading to shortened
3843 data being returned. It should now handle lines of any length.
3845 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3846 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3849 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3850 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3851 no longer happens in comment lines. A list of replacements is now placed
3852 at the head of all of the source files, except those whose only change is
3853 to replace PERL_COMMAND in the very first #! line.
3855 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3856 messages on the queue, with a bottom-up merge sort, using code contributed
3857 by Michael Haardt. This should make operations like -bp somewhat faster on
3858 large queues. It won't affect queue runners, except when queue_run_in_order
3861 51. Installed eximstats 1.31 in the distribution.
3863 52. Added support for SRV lookups to the dnslookup router.
3865 53. If an ACL referred to $message_body or $message_body_end, the value was not
3866 reset for any messages that followed in the same SMTP session.
3868 54. The store-handling optimization for building very long strings was not
3869 differentiating between the different store pools. I don't think this
3870 actually made any difference in practice, but I've tidied it.
3872 55. While running the routers to verify a sender address, $sender_address
3873 was still set to the sender address. This is wrong, because when routing to
3874 send a bounce to the sender, it would be empty. Therefore, I have changed
3875 it so that, while verifying a sender address, $sender_address is set to <>.
3876 (There is no change to what happens when verifying a recipient address.)
3878 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3879 target A or AAAA records (if not already returned) without resetting the
3880 qualify_single or search_parents options of the DNS resolver. These are
3881 inappropriate in this case because the targets of MX and SRV records must
3882 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3883 target that, when qualified, matched something in the local domain. These
3884 two options are now turned off when doing these lookups.
3886 57. It seems that at least some releases of Reiserfs (which does not have the
3887 concept of a fixed number of inodes) returns zero and not -1 for the
3888 number of available inodes. This interacted badly with check_spool_inodes,
3889 which assumed that -1 was the "no such thing" setting. What I have done is
3890 to check that the total number of inodes is greater than zero before doing
3891 the test of how many are available.
3893 58. When a "warn" ACL statement has a log_message modifier, the message is
3894 remembered, and not repeated. This is to avoid a lot of repetition when a
3895 message has many recipients that cause the same warning to be written.
3896 Howewer, Exim was preserving the list of already written lines for an
3897 entire SMTP session, which doesn't seem right. The memory is now reset if a
3898 new message is started.
3900 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3901 debugging output unless log_rewrite was also set.
3903 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3904 of (int)(handle) when we know that handle contains (void *)(-1).
3906 61. The Exim daemon panic-logs an error return when it closes the incoming
3907 connection. However "connection reset by peer" seems to be common, and
3908 isn't really an error worthy of noting specially, so that particular error
3911 62. When Exim is trying to find all the local interfaces, it used to panic and
3912 die if the ioctl to get the interface flags failed. However, it seems that
3913 on at least one OS (Solaris 9) it is possible to have an interface that is
3914 included in the list of interfaces, but for which you get a failure error
3915 for this call. This happens when the interface is not "plumbed" into a
3916 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3917 failure of the "get flags" call assumes that the interface is down.
3919 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3920 makes life easier for people who are doing arithmetic on fields extracted
3921 from dates, where you often get leading zeros that should not be
3922 interpreted as octal.
3924 64. Added qualify_domain to the redirect router, to override the global
3927 65. If a pathologically long header line contained very many addresses (the
3928 report of this problem mentioned 10 000) and each of them was rewritten,
3929 Exim could use up a very large amount of memory. (It kept on making new
3930 copies of the header line as it rewrote, and never released the old ones.)
3931 At the expense of a bit more processing, the header rewriting function has
3932 been changed so that it no longer eats memory in this way.
3934 66. The generation of the Received: header has been moved from the time that a
3935 message starts to be received, to the time that it finishes. The timestamp
3936 in the Received: header should now be very close to that of the <= log
3937 line. There are two side-effects of this change:
3939 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3940 logged header lines no longer include the local Received: line, because
3941 it has not yet been created. The same applies to a copy of the message
3942 that is returned to a non-SMTP sender when a message is rejected.
3944 (b) When a filter file is tested using -bf, no additional Received: header
3945 is added to the test message. After some thought, I decided that this
3948 This change does not affect the value of $received_for. It is still set
3949 after address rewriting, but before local_scan() is called.
3951 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3953 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3954 gave an unhelpful panic error message, and a defer error. I have managed to
3955 change this behaviour so that it now rejects any supplied certificate,
3956 which seems right, as the list of acceptable certificates is empty.
3958 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3959 gave an unhelpful defer error. I have not managed to make this reject any
3960 supplied certificates, but the error message it gives is "no certificate
3961 supplied", which is not helpful.
3963 70. exigrep's output now also includes lines that are not associated with any
3964 message, but which match the given pattern. Implemented by a patch from
3965 Martin Sluka, which also tidied up the Perl a bit.
3967 71. Recipient callout verification, like sender verification, was using <> in
3968 the MAIL FROM command. This isn't really the right thing, since the actual
3969 sender may affect whether the remote host accepts the recipient or not. I
3970 have changed it to use the actual sender in the callout; this means that
3971 the cache record is now keyed on a recipient/sender pair, not just the
3972 recipient address. There doesn't seem to be a real danger of callout loops,
3973 since a callout by the remote host to check the sender would use <>.
3974 [SEE ABOVE: changed after hitting problems.]
3976 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3977 temporary errors. However, in the case of such a code being given after
3978 the end of a data transmission (i.e. after ".") Exim was failing to write
3979 a retry record for the message. (Yes, there was some broken host that was
3980 actually sending 8xx at this point.)
3982 73. An unknown lookup type in a host list could cause Exim to panic-die when
3983 the list was checked. (An example that provoked this was putting <; in the
3984 middle of a list instead of at the start.) If this happened during a DATA
3985 ACL check, a -D file could be left lying around. This kind of configuration
3986 error no longer causes Exim to die; instead it causes a defer errror. The
3987 incident is still logged to the main and panic logs.
3989 74. Buglet left over from Exim 3 conversion. The message "too many messages
3990 in one connection" was written to the rejectlog but not the mainlog, except
3991 when address rewriting (yes!) was being logged.
3993 75. Added write_rejectlog option.
3995 76. When a system filter was run not as root (that is, when system_filter_user
3996 was set), the values of the $n variables were not being returned to the
3997 main process; thus, they were not subsequently available in the $sn
4000 77. Added +return_path_on_delivery log selector.
4002 78. A connection timeout was being treated differently from recipients deferred
4003 when testing hosts_max_try with a message that was older than the host's
4004 retry timeout. (The host should not be counted, thus allowing all hosts to
4005 be tried at least once before bouncing.) This may have been the cause of an
4006 occasionally reported bug whereby a message would remain on the queue
4007 longer than the retry timeout, but would be bounced if a delivery was
4008 forced. I say "may" because I never totally pinned down the problem;
4009 setting up timeout/retry tests is difficult. See also the next item.
4011 79. The ultimate address timeout was not being applied to errors that involved
4012 a combination of host plus message (for example, a timeout on a MAIL
4013 command). When an address resolved to a number of possible hosts, and they
4014 were not all tried for each delivery (e.g. because of hosts_max_try), a
4015 message could remain on the queue longer than the retry timeout.
4017 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4020 81. Fixed an obscure SMTP outgoing bug which required at least the following
4021 conditions: (a) there was another message waiting for the same server;
4022 (b) the server returned 5xx to all RCPT commands in the first message so
4023 that the message was not completed; (c) the server dropped the connection
4024 or gave a negative response to the RSET that Exim sends to abort the
4025 transaction. The observed case was a dropped connection after DATA that had
4026 been sent in pipelining mode. That is, the server had advertised PIPELINING
4027 but was not implementing it correctly. The effect of the bug was incorrect
4028 behaviour, such as trying another host, and this could lead to a crash.
4034 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4035 and daemon.c were passed as pointers to ints; they should have been
4036 pointers to socklen_t variables (which are typically unsigned ints).
4038 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4041 3. Fixed a really odd bug that affected only the testing scheme; patching a
4042 certain fixed string in the binary changed the value of another string that
4043 happened to be identical to the end of the original first string.
4045 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4046 name", it returns that address as the IP address. On some operating
4047 systems (e.g. Solaris), it also passes back the IP address string as the
4048 "host name". However, on others (e.g. Linux), it passes back an empty
4049 string. Exim wasn't checking for this, and was changing the host name to an
4050 empty string, assuming it had been canonicized.
4052 5. Although rare, it is permitted to have more than one PTR record for a given
4053 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4054 all the names associated with an address, because they do in Solaris.
4055 However, it seems that they do not in Linux for data that comes from the
4056 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4057 I found this out when I moved to a new Linux workstation and tried to run
4058 the Exim test suite.
4060 To get round this problem I have changed the code so that it now does its
4061 own call to the DNS to look up PTR records when searching for a host name.
4062 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4063 addresses that are only in /etc/hosts are still found.
4065 This behaviour is, however, controlled by an option called host_lookup_
4066 order, which defaults to "bydns:byaddr". If people want to use the other
4067 order, or indeed, just use one or the other means of lookup, they can
4068 specify it in this variable.
4070 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4071 some operating systems, this comes back from gethostbyaddr() as an empty
4072 string, and this is what Exim used to test for. However, it seems that in
4073 other systems, "." is yielded. Exim now tests for this case too.
4075 7. The values of check_spool_space and check_log_space are now held internally
4076 as a number of kilobytes instead of an absolute number of bytes. If a
4077 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4078 kilobyte. This means that much larger values can be stored.
4080 8. Exim monitor: an attempt to get the action menu when not actually pointing
4081 at a message produces an empty menu entitled "No message selected". This
4082 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4083 no entries in it ("Shell widget menu has zero width and/or height"). So I
4084 have added a single, blank menu entry in this case.
4086 9. Added ${quote_local_part.
4088 10. MIME decoding is now applied to the contents of Subject: header lines when
4091 11. Now that a reference to $sender_host_address automatically causes a reverse
4092 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4093 host lookup before query-style lookups in lists that might use this
4094 variable. This has therefore been abolished, and the "net-" prefix is no
4095 longer necessary for query-style lookups.
4097 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4098 have been a typo for LFLAGS, so it has been changed.
4100 13. The install script calls Exim with "-C /dev/null" in order to find the
4101 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4102 to be output. Howeve, since Exim outputs its version number before the
4103 error, it didn't break the script. It just looked ugly. I fixed this by
4104 always allowing "-C /dev/null" if the caller is root.
4106 14. Ignore overlarge ACL variable number when reading spool file - insurance
4107 against a later release with more variables having written the file.
4109 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4110 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4111 this, as well as the form without the "IPv6" on the front.
4113 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4114 OS/Makefile-Darwin file.
4116 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4117 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4119 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4120 variable. However, some OS use ints for inodes. I've added cast to long int
4121 to get rid of the compiler warning.
4123 19. I had forgotten to lock out "/../" in configuration file names when
4124 ALT_CONFIG_PREFIX was set.
4126 20. Routers used for verification do not need to specify transports. However,
4127 if such a router generated a host list, and callout was configured, Exim
4128 crashed, because it could not find a port number from the (non-existent)
4129 transport. It now assumes port 25 in this circumstance.
4131 21. Added the -t option to exigrep.
4133 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4134 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4135 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4136 them is set, LOOKUP_LSEARCH is forced.
4138 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4140 24. Added sender and host information to the "rejected by local_scan()" log
4141 line; previously there was no indication of these.
4143 25. Added .include_if_exists.
4145 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4146 Linux. It turns out that not all file systems support this. Apparently some
4147 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4148 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4149 not supported on the file descriptor, is now ignored when Exim is trying to
4150 sync a directory. This applies only to Linux.
4152 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4154 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4155 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4158 29. Added support for maildirsize files from supplied patch (modified a bit).
4160 30. The use of :fail: followed by an empty string could lead Exim to respond to
4161 sender verification failures with (e.g.):
4163 550 Verification failed for <xxx>
4164 550 Sender verify failed
4166 where the first response line was missing the '-' that indicates it is not
4167 the final line of the response.
4169 31. The loop for finding the name of the user that called Exim had a hardwired
4170 limit of 10; it now uses the value of finduser_retries, which is used for
4171 all other user lookups.
4173 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4176 33. Exim was neglecting to zero errno before one call of strtol() when
4177 expanding a string and expecting an integer value. On some systems this
4178 resulted in spurious "integer overflow" errors. Also, it was casting the
4179 result into an int without checking.
4181 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4182 did not work. The code looks as if it has *never* worked, though it appears
4183 to have been documented since at least releast 1.62. I have made it work.
4185 35. The "timeout_DNS" error in retry rules, also documented since at least
4186 1.62, also never worked. As it isn't clear exactly what this means, and
4187 clearly it isn't a major issue, I have abolished the feature by treating it
4188 as "timeout", and writing a warning to the main and panic logs.
4190 36. The display of retry rules for -brt wasn't always showing the error code
4193 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4194 timeout_connect_A, timeout_connect_MX.
4196 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4197 to the empty sender.
4199 39. The daemon was not analysing the content of -oX till after it had closed
4200 stderr and disconnected from the controlling terminal. This meant that any
4201 syntax errors were only noted on the panic log, and the return code from
4202 the command was 0. By re-arranging the code a little, I've made the
4203 decoding happen first, so such errors now appear on stderr, and the return
4204 code is 1. However, the actual setting up of the sockets still happens in
4205 the disconnected process, so errors there are still only recorded on the
4208 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4209 connections (as happens on some IP stacks) was logged at start up time as
4210 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4211 it from "IPv6 and IPv4", which means that two separate sockets are being
4214 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4215 says whether AF_INET or AF_INET6 was passed as an argument.
4217 42. Exiwhat output was messed up when time zones were included in log
4220 43. Exiwhat now gives more information about the daemon's listening ports,
4221 and whether -tls-on-connect was used.
4223 44. The "port" option of the smtp transport is now expanded.
4225 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4226 silently ignored. Now an error message is written to the main and panic
4229 46. There's a new ACL modifier called "logwrite" which writes to a log file
4230 as soon as it is encountered.
4232 47. Added $local_user_uid and $local_user_gid at routing time.
4234 48. Exim crashed when trying to verify a sender address that was being
4237 49. Exim was recognizing only a space character after ".include". It now also
4238 recognizes a tab character.
4240 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4241 extracting the relevant information from the specification. The man page no
4242 longer contains scrambled data for the -d option, and I've added a section
4243 at the front about calling Exim under different names.
4245 51. Added "extra_headers" argument to the "mail" command in filter files.
4247 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4250 53. Installed eximstats 1.29.
4252 54. Added transport_filter_timeout as a generic transport option.
4254 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4255 Cc: header lines. This was required by RFC 822, but it not required by RFC
4258 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4259 incoming messages that did not have them. Now it does so only if the
4260 message originates locally, that is, if there is no associated remote host
4261 address. When Resent- header lines are present, this applies to the Resent-
4262 lines rather than the non-Resent- lines.
4264 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4265 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4267 58. Messages for configuration errors now include the name of the main
4268 configuration file - useful now that there may be more than one file in a
4269 list (.included file names were always shown).
4271 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4272 for those rare installations that do not start the daemon as root or run it
4273 setuid root. I've cut out the call to initgroups() if the daemon is not
4276 60. The Exim user and group can now be bound into the binary as text strings
4277 that are looked up at the start of Exim's processing.
4279 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4281 62. Added $mailstore_basename variable.
4283 63. Installed patch to sieve.c from Michael Haardt.
4285 64. When Exim failed to open the panic log after failing to open the main log,
4286 the original message it was trying to log was written to stderr and debug
4287 output, but if they were not available (the usual case in production), it
4288 was lost. Now it is written to syslog before the two lines that record the
4289 failures to open the logs.
4291 65. Users' Exim filters run in subprocesses under the user's uid. It is
4292 possible for a "deliver" command or an alias in a "personal" command to
4293 provoke an address rewrite. If logging of address rewriting is configured,
4294 this fails because the process is not running as root or exim. There may be
4295 a better way of dealing with this, but for the moment (because 4.30 needs
4296 to be released), I have disabled address rewrite logging when running a
4297 filter in a non-root, non-exim process.
4303 1. The buildconfig auxiliary program wasn't quoting the value set for
4304 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4305 not defined. This bug was masked in 4.22 by the effect that was fixed in
4308 2. Some messages that were rejected after a message id was allocated were
4309 shown as "incomplete" by exigrep. It no longer does this for messages that
4310 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4312 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4313 have allow_domain_literals set, the ID did not get logged in the <= line.
4314 Domain literals are now always recognized in Message-ID: header lines.
4316 4. The first argument for a ${extract expansion item is the key name or field
4317 number. Leading and trailing spaces in this item were not being ignored,
4318 causing some misleading effects.
4320 5. When deliver_drop_privilege was set, single queue runner processes started
4321 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4322 same command in the process it spins off) were not dropping privilege.
4324 6. When the daemon running as "exim" started a queue runner, it always
4325 re-executed Exim in the spun-off process. This is a waste of effort when
4326 deliver_drop_privilege is set. The new process now just calls the
4327 queue-runner function directly.
4333 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4336 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4337 ignored. Though the use of -oP was forcing the writing of a pid file, it
4338 was always written to the default place.
4340 3. If the message "no IP address found for host xxxx" is generated during
4341 incoming verification, it is now followed by identification of the incoming
4342 connection (so you can more easily find what provoked it).
4344 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4346 5. Added some features to "harden" Exim a bit more against certain attacks:
4348 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4349 be put in Local/Makefile. This is like the never_users runtime option,
4350 but it cannot be overridden. The default setting is "root".
4352 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4353 prefix string with which any file named in a -C command line option
4356 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4357 is retained for -C and -D only if the caller of Exim is root. Without
4358 it, the exim user may also use -C and -D and retain privilege.
4360 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4361 command line option is disabled.
4363 6. Macro names set by the -D option must start with an upper case letter, just
4364 like macro names defined in the configuration file.
4366 7. Added "dereference=" facility to LDAP.
4368 8. Two instances of the typo "uknown" in the source files are fixed.
4370 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4371 the Configure-Makefile script screwed up while processing it.
4373 10. Incorporated PCRE 4.4.
4375 11. The SMTP synchronization check was not operating right at the start of an
4376 SMTP session. For example, it could not catch a HELO sent before the client
4377 waited for the greeting. There is now a check for outstanding input at the
4378 point when the greeting is written. Because of the duplex, asynchronous
4379 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4380 way, but not yet received, when the check is performed.
4382 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4383 on TCP/IP sockets, because this apparently causes some broken clients to
4386 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4387 unchanged) from the Cygwin maintainer.
4389 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4390 was supported instead of testing for mbx. Effectively a typo.
4392 15. The spa authenticator server code was not checking that the input it
4393 received was valid base64.
4395 16. The debug output line for the "set" modifier in ACLs was not showing the
4396 name of the variable that was being set.
4398 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4400 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4401 lookup on its own. Something else was needed to trigger the lookup. For
4402 example, a match in host_lookup or the need for a host name in a host list.
4403 Now, if $sender_host_name is referenced and the host name has not yet been
4404 looked up, a lookup is performed. If the lookup fails, the variable remains
4405 empty, and $host_lookup_failed is set to "1".
4407 19. Added "eqi" as a case-independent comparison operator.
4409 20. The saslauthd authentication condition could segfault if neither service
4410 nor realm was specified.
4412 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4413 error message that was logged was misleading, and incoming SMTP
4414 connections were dropped. The message is now more accurate, and temporary
4415 errors are given to SMTP connections.
4417 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4418 (except RSET) with a 421 error, until QUIT is received. However, it was
4419 failing to send a response to QUIT.
4421 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4422 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4423 ACL is now run after the helo_try_verify_hosts code.
4425 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4426 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4427 case-independent, so other case variants are also recognized). Apparently
4428 some systems use these upper case variants.
4430 25. If more than two messages were waiting for the same host, and a transport
4431 filter was specified for the transport, Exim sent two messages over the
4432 same TCP/IP connection, and then failed with "socket operation on non-
4433 socket" when it tried to send the third.
4435 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4437 27. The extern definition of crypt16() in expand.c was not being excluded when
4438 the OS had its own crypt16() function.
4440 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4441 as a preferred synonym for return_size_limit, both as an option and as an
4444 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4446 30. Changed the default configuration ACL to relax the local part checking rule
4447 for addresses that are not in any local domains. For these addresses,
4448 slashes and pipe symbols are allowed within local parts, but the sequence
4449 /../ is explicitly forbidden.
4451 31. SPA server authentication was not clearing the challenge buffer before
4454 32. log_message in a "warn" ACL statement was writing to the reject log as
4455 well as to the main log, which contradicts the documentation and doesn't
4456 seem right (because no rejection is happening). So I have stopped it.
4458 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4459 However, I am unable to do any testing of this.
4461 34. Fixed an infelicity in the appendfile transport. When checking directories
4462 for a mailbox, to see if any needed to be created, it was accidentally
4463 using path names with one or more superfluous leading slashes; tracing
4464 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4466 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4467 contents are added to the log line that is written for every discarded
4468 recipient. (Previously a log_message setting was ignored.)
4470 36. The ${quote: operator now quotes the string if it is empty.
4472 37. The install script runs exim in order to find its version number. If for
4473 some reason other than non-existence or emptiness, which it checks, it
4474 could not run './exim', it was installing it with an empty version number,
4475 i.e. as "exim-". This error state is now caught, and the installation is
4478 38. An argument was missing from the function that creates an error message
4479 when Exim fails to connect to the socket for saslauthd authentication.
4480 This could cause Exim to crash, or give a corrupted message.
4482 39. Added isip, isip4, and isip6 to ${if conditions.
4484 40. The ACL variables $acl_xx are now saved with the message, and can be
4485 accessed later in routers, transports, and filters.
4487 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4488 strings in the file are not string-expanded.
4490 42. If a MAIL command specified a SIZE value that was too large to fit into an
4491 int variable, the check against message_size_limit failed. Such values are
4492 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4493 day this will have to be increased, but I don't think I want to be around
4494 when emails are that large.
4501 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4502 iconv() is not standard in FreeBSD.
4504 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4505 IPv6 enabled. The observed symptom was a segmentation fault on return from
4506 the function os_common_find_running_interfaces() in src/os.c.
4508 3. In the check_special_case() function in daemon.c I had used "errno" as an
4509 argument name, which causes warnings on some systems. This was basically a
4510 typo, since it was named "eno" in the comments!
4512 4. The code that waits for the clock to tick (at a resolution of some fraction
4513 of a second) so as to ensure message-id uniqueness was always waiting for
4514 at least one whole tick, when it could have waited for less. [This is
4515 almost certainly not relevant at current processor speeds, where it is
4516 unlikely to ever wait at all. But we try to future-proof.]
4518 5. The function that sleeps for a time interval that includes fractions of a
4519 second contained a race. It did not block SIGALRM between setting the
4520 timer, and suspending (a couple of lines later). If the interval was short
4521 and the sigsuspend() was delayed until after it had expired, the suspension
4522 never ended. On busy systems this could lead to processes getting stuck for
4525 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4526 process, before it forks any delivery processes. The open lookup caching
4527 mechanism meant that the open file or database connection was passed into
4528 the delivery process. The problem was that delivery processes always tidy
4529 up cached lookup data. This could cause a problem for the next delivery
4530 process started by the queue runner, because the external queue runner
4531 process does not know about the closure. So the next delivery process
4532 still has data in the lookup cache. In the case of a file lookup, there was
4533 no problem because closing a file descriptor in a subprocess doesn't affect
4534 the parent. However, if the lookup was caching a connection to a database,
4535 the connection was closed, and the second delivery process was likely to
4536 see errors such as "PGSQL: query failed: server closed the connection
4537 unexpectedly". The problem has been fixed by closing all cached lookups
4538 in a queue runner before running a delivery process.
4540 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4541 seem to have the "const" qualifier which it has on other OS. I've
4544 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4545 *of the same type* (client or server) with the same public name that an
4546 error should be diagnosed.
4548 9. When Exim looked up a host name for an IP address, but failed to find the
4549 original IP address when looking up the host name (a safety check), it
4550 output the message "<ip address> does not match any IP for NULL", which was
4551 confusing, to say the least. The bug was that the host name should have
4552 appeared instead of "NULL".
4554 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4555 user that is built into the binary, and the -C or -D options is used, root
4556 privilege is dropped before the configuration file is read. In addition,
4557 logging is switched to stderr instead of the normal log files. If the
4558 configuration then re-defines the Exim user, the unprivileged environment
4559 is probably not what is expected, so Exim logs a panic warning message (but
4562 However, if deliver_drop_privilege is set, the unprivileged state may well
4563 be exactly what is intended, so the warning has been cut out in that case,
4564 and Exim is allowed to try to write to its normal log files.
4570 1. smtp_return_error_details was not giving details for temporary sender
4571 or receiver verification errors.
4573 2. Diagnose a configuration error if two authenticators have the same public
4576 3. Exim used not to create the message log file for a message until the first
4577 delivery attempt. This could be confusing when incoming messages were held
4578 for policy or load reasons. The message log file is now created at the time
4579 the message is received, and an initial "Received" line is written to it.
4581 4. The automatically generated man page for command line options had a minor
4582 bug that caused no ill effects; however, a more serious problem was that
4583 the procedure for building the man page automatically didn't always
4584 operate. Consequently, release 4.20 contains an out-of-date version. This
4585 shouldn't happen again.
4587 5. When building Exim with embedded Perl support, the script that builds the
4588 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4589 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4591 6. The freeze_tell option was not being used for messages that were frozen on
4592 arrival, either by an ACL or by local_scan().
4594 7. Added the smtp_incomplete_transaction log selector.
4596 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4597 was accepting AUTH without a new EHLO.
4599 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4600 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4603 10. Logging of TCP/IP connections (when configured) now happens in the main
4604 daemon process instead of the child process, so that the TCP/IP connection
4605 count is more accurate (but it can never be perfect).
4607 11. The use of "drop" in a nested ACL was not being handled correctly in the
4608 outer ACL. Now, if condition failure induced by the nested "drop" causes
4609 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4610 or "require"), the connection is dropped.
4612 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4613 that yield "discard" can now be used with an "accept" or a "discard" verb,
4614 but an error is generated for any others (because I can't see a useful way
4615 to define what should happen).
4617 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4618 are now processed in the same way as lines in the Exim configuration file.
4619 In particular, continuation lines are supported.
4621 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4623 15. Added -ti meaning -t -i.
4625 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4626 domains, and warn by logging if others are found.
4628 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4629 returned from GIFCONF when Exim is trying to find the list of interfaces on
4630 a host. The code in os.c has been modified to copy each ifreq to an aligned
4631 structure in all cases.
4633 Also, in some cases, the returned ifreq's were being copied to a 'struct
4634 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4635 means the last couple of bytes of an IPv6 address could be chopped if the
4636 ifreq contained only a normal sockaddr (14 bytes storage).
4638 18. Named domain lists were not supported in the hosts_treat_as_local option.
4639 An entry such as +xxxx was not recognized, and was treated as a literal
4642 19. Ensure that header lines added by a DATA ACL are included in the reject log
4643 if the ACL subsequently rejects the message.
4645 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4646 MD5 (which is deprecated).
4648 21. When testing a filter file using -bf, Exim was writing a message when it
4649 took the sender from a "From " line in the message, but it was not doing so
4650 when it took $return_path from a Return-Path: header line. It now does.
4652 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4653 with a valid header line field name (a series of printing characters
4654 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4656 23. Changed "disc" in the source to "disk" to conform to the documentation and
4657 the book and for uniformity.
4659 24. Ignore Sendmail's -Ooption=value command line item.
4661 25. When execve() failed while trying to run a command in a pipe transport,
4662 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4663 could be confused with a return value of 69 from the command itself. This
4664 has been changed to 127, the value the shell returns if it is asked to run
4665 a non-existent command. The wording for the related log line suggests a
4666 non-existent command as the problem.
4668 26. If received_header_text expands to an empty string, do not add a Received:
4669 header line to the message. (Well, it adds a token one on the spool, but
4670 marks it "old" so that it doesn't get used or transmitted.)
4672 27. Installed eximstats 1.28 (addition of -nt option).
4674 28. There was no check for failure on the call to getsockname() in the daemon
4675 code. This can fail if there is a shortage of resources on the system, with
4676 ENOMEM, for example. A temporary error is now given on failure.
4678 29. Contrary to the C standard, it seems that in some environments, the
4679 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4680 program. Exim now does this explicitly; it affects the formatting of
4681 timestamps using strftime().
4683 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4684 complaints. I've now initialized all the variables, to avoid this.
4686 32. Header lines added by a system filter were not being "seen" during
4687 transport-time rewrites.
4689 33. The info_callback() function passed to OpenSSL is set up with type void
4690 (*)(SSL *, int, int), as described somewhere. However, when calling the
4691 function (actually a macro) that sets it up, the type void(*)() is
4692 expected. I've put in a cast to prevent warnings from picky compilers.
4694 34. If a DNS black list lookup found a CNAME record, but there were no A
4695 records associated with the domain it pointed at, Exim crashed.
4697 35. If a DNS black list lookup returned more than one A record, Exim ignored
4698 all but the first. It now scans all returned addresses if a particular IP
4699 value is being sought. In this situation, the contents of the
4700 $dnslist_value variable are a list of all the addresses, separated by a
4703 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4704 to accept a host name and all its aliases if the forward lookup for any of
4705 them yielded the IP address of the incoming connection. Now it accepts only
4706 those names whose forward lookup yields the correct IP address. Any other
4707 names are discarded. This closes a loophole whereby a rogue DNS
4708 administrator could create reverse DNS records to break through a
4709 wildcarded host restriction in an ACL.
4711 37. If a user filter or a system filter that ran in a subprocess used any of
4712 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4713 the wrong values were passed to the pipe command ($thisaddress had the
4714 value of $0, $0 had the value of $1, etc). This bug was introduced by
4715 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4717 38. Improved the line breaking for long SMTP error messages from ACLs.
4718 Previously, if there was no break point between 40 and 75 characters, Exim
4719 left the rest of the message alone. Two changes have been made: (a) I've
4720 reduced the minimum length to 35 characters; (b) if it can't find a break
4721 point between 35 and 75 characters, it looks ahead and uses the first one
4722 that it finds. This may give the occasional overlong line, but at least the
4723 remaining text gets split now.
4725 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4726 file descriptors might be low, and that setting 1000 would always raise it.
4727 It turns out that in some environments, the limit is already over 1000 and
4728 that lowering it causes trouble. So now Exim takes care not to decrease it.
4730 40. When delivering a message, the value of $return_path is set to $sender_
4731 address at the start of routing (routers may change the value). By an
4732 oversight, this default was not being set up when an address was tested by
4733 -bt or -bv, which affected the outcome if any router or filter referred to
4736 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4737 log only when "message" or "log_message" are set. However, if one of the
4738 conditions was an address verification, or a call to a nested ACL, the
4739 messages generated by the underlying test were being passed through. This
4740 no longer happens. The underlying message is available in $acl_verify_
4741 message for both "message" and "log_message" expansions, so it can be
4742 passed through if needed.
4744 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4745 new expansion $bh_ to give the encoded byte string without charset
4746 translation. Translation happens only if iconv() is available; HAVE_ICONV
4747 indicates this at build time. HEADERS_CHARSET gives the charset to
4748 translate to; headers_charset can change it in the configuration, and
4749 "headers charset" can change it in an individual filter file.
4751 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4752 that creates RFC 2047 encoded "words" labels them as that charset instead
4753 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4754 expansion operator; (ii) when Exim creates a From: line for a local
4755 message; (iii) when a header line is rewritten to include a "phrase" part.
4757 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4758 buggy, causing it to skip the first lines of messages whose message ID
4759 ended in 'D'. This would not have bitten before Exim release 4.14, because
4760 message IDs were unlikely to end in 'D' before then. The effect was to have
4761 incorrect size information for certain domains.
4763 45. #include "config.h" was missing at the start of the crypt16.c module. This
4764 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4767 46. If there was a timeout during a "random" callout check, Exim treated it as
4768 a failure of the random address, and carried on sending RSET and the real
4769 address. If the delay was just some slowness somewhere, the response to the
4770 original RCPT would be taken as a response to RSET and so on, causing
4771 mayhem of various kinds.
4773 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4774 when I implemented it. It didn't allow for the fact that some option values
4775 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4776 the right test for positive values.
4778 48. Domain names in DNS records are case-independent. Exim always looks them up
4779 in lower case. Some resolvers return domain names in exactly the case they
4780 appear in the zone file, that is, they may contain uppercase letters. Not
4781 all resolvers do this - some return always lower case. Exim was treating a
4782 change of case by a resolver as a change of domain, similar to a widening
4783 of a domain abbreviation. This triggered its re-routing code and so it was
4784 trying to route what was effectively the same domain again. This normally
4785 caused routing to fail (because the router wouldn't handle the domain
4786 twice). Now Exim checks for this case specially, and just changes the
4787 casing of the domain that it ultimately uses when it transmits the message
4790 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4793 50. If a filter generated a file delivery with a non-absolute name (possible if
4794 no home directory exists for the router), the forbid_file option was not
4797 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4798 the existing equality matching.
4800 52. Exim was using ints instead of ino_t variables in some places where it was
4801 dealing with inode numbers.
4803 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4804 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4805 called TMPDIR, and if it finds it is different, it changes its value.
4807 54. The smtp_printf() function is now made available to local_scan() so
4808 additional output lines can be written before returning. There is also an
4809 smtp_fflush() function to enable the detection of a dropped connection.
4810 The variables smtp_input and smtp_batched_input are exported to
4813 55. Changed the default runtime configuration: the message "Unknown user"
4814 has been removed from the ACL, and instead placed on the localuser router,
4815 using the cannot_route_message feature. This means that any verification
4816 failures that generate their own messages won't get overridden. Similarly,
4817 the "Unrouteable address" message that was in the ACL for unverifiable
4818 relay addresses has also been removed.
4820 56. Added hosts_avoid_esmtp to the smtp transport.
4822 57. The exicyclog script was not checking for the esoteric option
4823 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4824 will work only if exicyclog is run under the appropriate euid.
4826 58. Following a discussion on the list, the rules by which Exim recognises line
4827 endings on incoming messages have been changed. The -dropcr and drop_cr
4828 options are now no-ops, retained only for backwards compatibility. The
4829 following line terminators are recognized: LF CRLF CR. However, special
4830 processing applies to CR:
4832 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4833 nor a local message in the state where . is a terminator.
4835 (ii) If a bare CR is encountered in a header line, an extra space is added
4836 after the line terminator so as not to end the header. The reasoning
4837 behind this is that bare CRs in header lines are most likely either
4838 to be mistakes, or people trying to play silly games.
4840 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4841 was being incorrectly given as 18 bytes larger than it should have been.
4842 This is a VOB (very old bug).
4844 60. This may never have affected anything current, but just in case it has:
4845 When the local host is found other than at the start of a list of hosts,
4846 the local host, those with the same MX, and any that follow, are discarded.
4847 When the list in question was part of a longer list of hosts, the following
4848 hosts (not currently being processed) were also being discarded. This no
4849 longer happens. I'm not sure if this situation could ever has previously
4852 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4855 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4856 header. This is something that is recommended in a new Internet Draft, and
4857 is something that is documented as being done by Sendmail. There are two
4858 possible values. For messages generated by the autoreply transport, Exim
4861 Auto-Submitted: auto-replied
4863 whereas for all other generated messages (e.g. bounces) it adds
4865 Auto-Submitted: auto-generated
4867 63. The "personal" condition in filters now includes a test for the
4868 Auto-Submitted: header. If it contains the string "auto-" the message it
4869 not considered personal.
4871 64. Added rcpt_include_affixes as a generic transport option.
4873 65. Added queue_only_override (default true).
4875 66. Added the syslog_duplication option.
4877 67. If what should have been the first header line of a message consisted of
4878 a space followed by a colon, Exim was mis-interpreting it as a header line.
4879 It isn't of course - it is syntactically invalid and should therefore be
4880 treated as the start of the message body. The misbehaviour could have
4881 caused a number of strange effects, including loss of data in subsequent
4882 header lines, and spool format errors.
4884 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4885 client host had authenticated. This control can now be exercised by an ACL
4886 for more flexibility.
4888 69. By default, callouts do not happen when testing with -bh. There is now a
4889 variant, -bhc, which does actually run the callout code, including
4890 consulting and updating the callout cache.
4892 70. Added support for saslauthd authentication, courtesy of Alexander
4895 71. If statvfs() failed on the spool or log directories while checking their
4896 size for availability, Exim confusingly gave the error "space shortage".
4897 Furthermore, in debugging mode it crashed with a floating point exception.
4898 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4899 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4900 serious problem, Exim now writes to the main and panic logs when this
4901 happens, with details of the failure. It then refuses to accept the
4902 incoming message, giving the message "spool directory problem" or "log
4903 directory problem" with a 421 code for SMTP messages.
4905 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4906 0, 1, and 2 exist, because some OS complain for execs without them (see
4907 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4908 descriptors. However, the code omitted to check that the open succeeded,
4909 causing mysterious errors if for some reason the permissions on /dev/null
4910 got screwed. Now Exim writes a message to the main and panic logs, and
4911 bombs out if it can't open /dev/null.
4913 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4914 interact so that it is all more flexible. It is supposed to remain
4915 backwards compatible. Also added extra_local_interfaces.
4917 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4918 to bomb out with an assertion failure - to the client this appears as a
4919 connection drop. This problem occurs in the part of the code that was taken
4920 from the Samba project. Fortunately, the assertion is in a very simple
4921 function, so I have fixed this by reproducing the function inline in the
4922 one place where it is called, and arranging for authentication to fail
4923 instead of killing the process with assert().
4925 75. The SPA client code was not working when the server requested OEM rather
4926 than Unicode encoding.
4928 76. Added code to make require_files with a specific uid setting more usable in
4929 the case where statting the file as root fails - usually a non-root-mounted
4930 NFS file system. When this happens and the failure is EACCES, Exim now
4931 forks a subprocess and does the per-uid checking as the relevant uid.
4933 77. Added process_log_path.
4935 78. If log_file_path was not explicitly set, a setting of check_log_space or
4936 check_log_inodes was ignored.
4938 79. If a space check for the spool or log partitions fails, the incident is now
4939 logged. Of course, in the latter case the data may get lost...
4941 80. Added the %p formatting code to string_format() so that it can be used to
4942 print addresses in debug_print(). Adjusted all the address printing in the
4943 debugging in store.c to use %p rather than %d.
4945 81. There was a concern that a line of code in smtp_in.c could overflow a
4946 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4947 initially expressed, the concern was not well-founded, because trailing
4948 spaces are removed early. However, if the trailing spaces were followed by
4949 a NULL, they did not get removed, so the overflow was possible. Two fixes
4952 (a) I re-wrote the offending code in a cleaner fashion.
4953 (b) If an incoming SMTP command contains a NULL character, it is rejected
4956 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4957 runs initgroups(), so that if the Exim user is in any additional groups,
4958 they will be used during message reception.
4964 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob