1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.578 2009/10/20 12:45:49 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 622: 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/16 Bugzilla 851: Documentation example syntax fix.
104 NM/17 Changed NOTICE file to remove references to embedded PCRE.
106 NM/18 Bugzilla 894: Fix issue with very long lines including comments in lsearch
108 NM/19 Bugzilla 745: TLS version reporting
109 Patch provided by Phil Pennock
111 NM/20 Bugzilla 167: bool: condition support
112 Patch provided by Phil Pennock
114 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken clients
115 Patch provided by Phil Pennock
117 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date
118 Patch provided by Brad "anomie" Jorsch
120 NM/23 Bugzilla 687: Fix misparses in eximstats
121 Patch provided by Heiko Schlittermann
123 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid
124 Patch provided by Heiko Schlittermann
126 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file
128 NM/26 Bugzilla 799: Documentation correction for ratelimit
130 NM/27 Bugzilla 802: Improvements to local interface IP addr detection
131 Patch provided by David Brownlee
137 TK/01 Add preliminary DKIM support. Currently requires a forked version of
138 ALT-N's libdkim that I have put here:
139 http://duncanthrax.net/exim-experimental/
141 Note to Michael Haardt: I had to rename some vars in sieve.c. They
142 were called 'true' and it seems that C99 defines that as a reserved
143 keyword to be used with 'bool' variable types. That means you could
144 not include C99-style headers which use bools without triggering
145 build errors in sieve.c.
147 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
148 as mailq or other aliases. Changed the --help handling significantly
149 to do whats expected. exim_usage() emits usage/help information.
151 SC/01 Added the -bylocaldomain option to eximstats.
153 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
155 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
157 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
163 PH/01 Another patch from the Sieve maintainer.
165 PH/02 When an IPv6 address is converted to a string for single-key lookup
166 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
167 dots are used instead of colons so that keys in lsearch files need not
168 contain colons. This was done some time before quoting was made available
169 in lsearch files. However, iplsearch files do require colons in IPv6 keys
170 (notated using the quote facility) so as to distinguish them from IPv4
171 keys. This meant that lookups for IP addresses in host lists did not work
172 for iplsearch lookups.
174 This has been fixed by arranging for IPv6 addresses to be expressed with
175 colons if the lookup type is iplsearch. This is not incompatible, because
176 previously such lookups could never work.
178 The situation is now rather anomolous, since one *can* have colons in
179 ordinary lsearch keys. However, making the change in all cases is
180 incompatible and would probably break a number of configurations.
182 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
185 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
186 conversion specification without a maximum field width, thereby enabling
187 a rogue spamd server to cause a buffer overflow. While nobody in their
188 right mind would setup Exim to query an untrusted spamd server, an
189 attacker that gains access to a server running spamd could potentially
190 exploit this vulnerability to run arbitrary code as the Exim user.
192 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
193 $primary_hostname instead of what libspf2 thinks the hosts name is.
195 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
196 a directory entry by the name of the lookup key. Previously, if a
197 symlink pointed to a non-existing file or a file in a directory that
198 Exim lacked permissions to read, a lookup for a key matching that
199 symlink would fail. Now it is enough that a matching directory entry
200 exists, symlink or not. (Bugzilla 503.)
202 PH/03 The body_linecount and body_zerocount variables are now exported in the
205 PH/04 Added the $dnslist_matched variable.
207 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
208 This means they are set thereafter only if the connection becomes
211 PH/06 Added the client_condition to authenticators so that some can be skipped
212 by clients under certain conditions.
214 PH/07 The error message for a badly-placed control=no_multiline_responses left
215 "_responses" off the end of the name.
217 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
219 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
220 (without spaces) instead of just copying the configuration text.
222 PH/10 Added the /noupdate option to the ratelimit ACL condition.
224 PH/11 Added $max_received_linelength.
226 PH/12 Added +ignore_defer and +include_defer to host lists.
228 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
229 way in which PCRE > 7.0 is built.
231 PH/14 Implemented queue_only_load_latch.
233 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
234 MAIL command. The effect was to mangle the value on 64-bit systems.
236 PH/16 Another patch from the Sieve maintainer.
238 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
240 PH/18 If a system quota error occurred while trying to create the file for
241 a maildir delivery, the message "Mailbox is full" was not appended to the
242 bounce if the delivery eventually timed out. Change 4.67/27 below applied
243 only to a quota excession during the actual writing of the file.
245 PH/19 It seems that peer DN values may contain newlines (and other non-printing
246 characters?) which causes problems in log lines. The DN values are now
247 passed through string_printing() before being added to log lines.
249 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
250 and InterBase are left for another time.)
252 PH/21 Added message_body_newlines option.
254 PH/22 Guard against possible overflow in moan_check_errorcopy().
256 PH/23 POSIX allows open() to be a macro; guard against that.
258 PH/24 If the recipient of an error message contained an @ in the local part
259 (suitably quoted, of course), incorrect values were put in $domain and
260 $local_part during the evaluation of errors_copy.
266 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
267 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
270 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
271 issue a MAIL command.
273 PH/02 In an ACL statement such as
275 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
277 if a client was not listed at all, or was listed with a value other than
278 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
279 the condition was not true (as it should be), so access was not denied.
280 The bug was that the ! inversion was incorrectly passed on to the second
281 item. This has been fixed.
283 PH/03 Added additional dnslists conditions == and =& which are different from
284 = and & when the dns lookup returns more than one IP address.
286 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
287 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
289 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
290 FSYNC, which compiles an option called disable_fsync that allows for
291 bypassing fsync(). The documentation is heavily laced with warnings.
293 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
295 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
296 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
297 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
298 including adding "make clean"; (3) Added -fPIC when compiling the test
299 dynamically loaded module, to get rid of a warning.
301 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
302 message fails, move_frozen_messages = true and ignore_bounce_errors_after
303 = 0s. The bug is otherwise harmless.
305 PH/07 There was a bug in the dovecot authenticator such that the value of
306 $auth1 could be overwritten, and so not correctly preserved, after a
307 successful authentication. This usually meant that the value preserved by
308 the server_setid option was incorrect.
310 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
312 PH/09 Installed PCRE release 7.0.
314 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
315 run for batched SMTP input. It is now run at the start of every message
316 in the batch. While fixing this I discovered that the process information
317 (output by running exiwhat) was not always getting set for -bs and -bS
318 input. This is fixed, and it now also says "batched" for BSMTP.
320 PH/11 Added control=no_pipelining.
322 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
323 patch, slightly modified), and move the expansion of helo_data till after
324 the connection is made in the smtp transport (so it can use these
327 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
329 PH/14 Added log_selector = +pid.
331 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
333 PH/16 Add ${if forany and ${if forall.
335 PH/17 Added dsn_from option to vary the From: line in DSNs.
337 PH/18 Flush SMTP output before performing a callout, unless control =
338 no_callout_flush is set.
340 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
341 was true (the default) a successful delivery failed to delete the retry
342 item, thus causing premature timeout of the address. The bug is now
345 PH/20 Added hosts_avoid_pipelining to the smtp transport.
347 PH/21 Long custom messages for fakedefer and fakereject are now split up
348 into multiline reponses in the same way that messages for "deny" and
349 other ACL rejections are.
351 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
352 with slight modification.
354 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
355 draft, changing the syntax and factoring some duplicate code".
357 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
358 for deliveries of the second and subsequent messages over the same SMTP
361 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
362 ${reduce, with only minor "tidies".
364 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
366 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
367 expansion side effects.
369 PH/27 When a message times out after an over-quota error from an Exim-imposed
370 quota, the bounce message says "mailbox is full". This message was not
371 being given when it was a system quota that was exceeded. It now should
374 MH/03 Made $recipients available in local_scan(). local_scan() already has
375 better access to the recipient list through recipients_list[], but
376 $recipients can be useful in postmaster-provided expansion strings.
378 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
379 in the case of a MAIL command with additional options following the
380 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
381 were accidentally chopped off.
383 PH/29 SMTP synchronization checks are implemented when a command is read -
384 there is a check that no more input is waiting when there shouldn't be
385 any. However, for some commands, a delay in an ACL can mean that it is
386 some time before the response is written. In this time, more input might
387 arrive, invalidly. So now there are extra checks after an ACL has run for
388 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
389 pipelining has not been advertised.
391 PH/30 MH's patch to allow iscntrl() characters to be list separators.
393 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
394 returned in the SMTP response when smtp_return_error_details was false.
397 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
398 instead of the C I/O that was originally supplied, because problems were
401 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
402 Exim which did not show up earlier: it was assuming that a call to
403 SSL_CTX_set_info_callback() might give an error value. In fact, there is
404 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
405 was a macro that became an assignment, so it seemed to work. This has
406 changed to a proper function call with a void return, hence the compile
407 error. Exim's code has been fixed.
409 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
412 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
414 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
416 PH/37 If a message is not accepted after it has had an id assigned (e.g.
417 because it turns out to be too big or there is a timeout) there is no
418 "Completed" line in the log. When some messages of this type were
419 selected by exigrep, they were listed as "not completed". Others were
420 picked up by some special patterns. I have improved the selection
421 criteria to be more general.
423 PH/38 The host_find_failed option in the manualroute router can now be set
424 to "ignore", to completely ignore a host whose IP address cannot be
425 found. If all hosts are ignored, the behaviour is controlled by the new
426 host_all_ignored option.
428 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
429 homing or because of multiple MX records with /mx) generated more than
430 one IP address, and the following item turned out to be the local host,
431 all the secondary addresses of the first item were incorrectly removed
432 from the list, along with the local host and any following hosts (which
433 is what is supposed to happen).
435 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
436 whoever called Exim into the -H file. In the case of the daemon it was
437 behaving confusingly. When first started, it used values for whoever
438 started the daemon, but after a SIGHUP it used the Exim user (because it
439 calls itself on a restart). I have changed the code so that it now always
442 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
443 message are rejected with the same error (e.g. no authentication or bad
444 sender address), and a DATA command is nevertheless sent (as can happen
445 with PIPELINING or a stupid MUA), the error message that was given to the
446 RCPT commands is included in the rejection of the DATA command. This is
447 intended to be helpful for MUAs that show only the final error to their
450 PH/42 Another patch from the Sieve maintainer.
452 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
453 Eximstats - Fixed some broken HTML links and added missing column headers
455 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
456 columns for Rejects, Temp Rejects, Ham, and Spam rows.
458 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
460 PH/43 Yet another patch from the Sieve maintainer.
462 PH/44 I found a way to check for a TCP/IP connection going away before sending
463 the response to the final '.' that terminates a message, but only in the
464 case where the client has not sent further data following the '.'
465 (unfortunately, this is allowed). However, in many cases there won't be
466 any further data because there won't be any more messages to send. A call
467 to select() can be used: if it shows that the input is "ready", there is
468 either input waiting, or the socket has been closed. An attempt to read
469 the next input character can distinguish the two cases. Previously, Exim
470 would have sent an OK response which the client would never have see.
471 This could lead to message repetition. This fix should cure that, at
472 least in a lot of common cases.
474 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
475 advertised in response to EHLO.
481 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
482 fixed by 4.65/MH/01 (is this a record?) are fixed:
484 (i) An empty string was always treated as zero by the numeric comparison
485 operators. This behaviour has been restored.
487 (ii) It is documented that the numeric comparison operators always treat
488 their arguments as decimal numbers. This was broken in that numbers
489 starting with 0 were being interpreted as octal.
491 While fixing these problems I realized that there was another issue that
492 hadn't been noticed. Values of message_size_limit (both the global option
493 and the transport option) were treated as octal if they started with 0.
494 The documentation was vague. These values are now always treated as
495 decimal, and I will make that clear in the documentation.
501 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
502 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
505 MH/01 Don't check that the operands of numeric comparison operators are
506 integers when their expansion is in "skipping" mode (fixes bug
507 introduced by 4.64-PH/07).
509 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
510 child addresses, Exim now panics and dies. Previously, because the count
511 is held in a short int, deliveries were likely to be lost. As such a
512 large number of recipients for a single message is ridiculous
513 (performance will be very, very poor), I have chosen to impose a limit
514 rather than extend the field.
520 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
521 leftover -K file (the existence of which was triggered by #402).
522 While we were at it, introduced process PID as part of the -K
523 filename. This should rule out race conditions when creating
526 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
527 processing considerably. Previous code took too long for large mails,
528 triggering a timeout which in turn triggers #401.
530 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
531 in the DK code in transports.c. sendfile() is not really portable,
532 hence the _LINUX specificness.
534 TF/01 In the add_headers option to the mail command in an Exim filter,
535 there was a bug that Exim would claim a syntax error in any
536 header after the first one which had an odd number of characters
539 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
540 callout verification, Exim cached a "reject" for the entire domain. This
541 is correct for most verifications, but it is not correct for a recipient
542 verification with use_sender or use_postmaster set, because in that case
543 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
544 case of MAIL FROM:<> rejection from other early rejections (e.g.
545 rejection of HELO). When verifying a recipient using a non-null MAIL
546 address, the cache is ignored if it shows MAIL FROM:<> rejection.
547 Whatever the result of the callout, the value of the domain cache is
548 left unchanged (for any other kind of callout, getting as far as trying
549 RCPT means that the domain itself is ok).
551 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
554 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
555 manifest itself as EPIPE rather than ECONNECT. When tidying away a
556 session, the daemon ignores ECONNECT errors and logs others; it now
557 ignores EPIPE as well.
559 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
560 (quoted-printable decoding).
562 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
563 later the small subsequent patch to fix an introduced bug.
565 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
567 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
569 PH/08 An error is now given if message_size_limit is specified negative.
571 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
572 to be given (somewhat) arbitrary names.
574 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
577 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
578 miscellaneous code fixes
580 PH/10 Added the log_reject_target ACL modifier to specify where to log
583 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
584 hostname. This is wrong, because it relates to the incoming message (and
585 probably the interface on which it is arriving) and not to the outgoing
586 callout (which could be using a different interface). This has been
587 changed to use the value of the helo_data option from the smtp transport
588 instead - this is what is used when a message is actually being sent. If
589 there is no remote transport (possible with a router that sets up host
590 addresses), $smtp_active_hostname is used.
592 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
593 tweaks were necessary in order to get it to work (see also 21 below):
594 (a) The code assumed that strncpy() returns a negative number on buffer
595 overflow, which isn't the case. Replaced with Exim's string_format()
597 (b) There were several signed/unsigned issues. I just did the minimum
598 hacking in of casts. There is scope for a larger refactoring.
599 (c) The code used strcasecmp() which is not a standard C function.
600 Replaced with Exim's strcmpic() function.
601 (d) The code set only $1; it now sets $auth1 as well.
602 (e) A simple test gave the error "authentication client didn't specify
603 service in request". It would seem that Dovecot has changed its
604 interface. Fortunately there's a specification; I followed it and
605 changed what the client sends and it appears to be working now.
607 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
610 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
611 address A is aliased to B and C, where B exists and C does not. Without
612 -v the output is "A verified" because verification stops after a
613 successful redirection if more than one address is generated. However,
614 with -v the child addresses are also verified. Exim was outputting "A
615 failed to verify" and then showing the successful verification for C,
616 with its parentage. It now outputs "B failed to verify", showing B's
617 parentage before showing the successful verification of C.
619 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
620 look up a TXT record in a specific list after matching in a combined
623 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
624 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
625 they consult the DNS. I had assumed they would set it the way they
626 wanted; and indeed my experiments on Linux seem to show that in some
627 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
628 To be on the safe side, however, I have now made the interface to
629 host_find_byname() similar to host_find_bydns(), with an argument
630 containing the DNS resolver options. The host_find_byname() function now
631 sets these options at its start, just as host_find_bydns() does. The smtp
632 transport options dns_qualify_single and dns_search_parents are passed to
633 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
634 of host_find_byname() use the default settings of RES_DEFNAMES
635 (qualify_single) but not RES_DNSRCH (search_parents).
637 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
638 spool_read_header() do less string testing, by means of a preliminary
639 switch on the second character of optional "-foo" lines. (This is
640 overdue, caused by the large number of possibilities that now exist.
641 Originally there were few.) While I was there, I also converted the
642 str(n)cmp tests so they don't re-test the leading "-" and the first
643 character, in the hope this might squeeze out yet more improvement.
645 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
646 flag allowing group syntax was set by the header_syntax check but not
647 turned off, possible causing trouble later; (2) The flag was not being
648 set at all for the header_verify test, causing "group"-style headers to
649 be rejected. I have now set it in this case, and also caused header_
650 verify to ignore an empty address taken from a group. While doing this, I
651 came across some other cases where the code for allowing group syntax
652 while scanning a header line wasn't quite right (mostly, not resetting
653 the flag correctly in the right place). These bugs could have caused
654 trouble for malformed header lines. I hope it is now all correct.
656 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
657 with the "reply" argument non-NULL. The code, however (which originally
658 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
659 but it didn't always do it. This confused somebody who was copying the
660 code for some other use. I have removed all the tests.
662 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
663 feature that was used to support insecure browsers during the U.S. crypto
664 embargo. It requires special client support, and Exim is probably the
665 only MTA that supported it -- and would never use it because real RSA is
666 always available. This code has been removed, because it had the bad
667 effect of slowing Exim down by computing (never used) parameters for the
668 RSA_EXPORT functionality.
670 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
671 authenticator to fail if there's a tab character in the incoming data
672 (there should never be unless someone is messing about, as it's supposed
673 to be base64-encoded). Also added, on Timo's advice, the "secured" option
674 if the connection is using TLS or if the remote IP is the same as the
675 local IP, and the "valid-client-cert option" if a client certificate has
678 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
679 authenticators. This can be used for authorization after authentication
680 succeeds. (In the case of plaintext, it servers for both authentication
683 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
684 if any retry times were supplied.
686 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
687 connection, where there is no client IP address to check. In this
688 situation, the verify now always succeeds.
690 PH/25 Applied John Jetmore's -Mset patch.
692 PH/26 Added -bem to be like -Mset, but loading a message from a file.
694 PH/27 In a string expansion for a processed (not raw) header when multiple
695 headers of the same name were present, leading whitespace was being
696 removed from all of them, but trailing whitespace was being removed only
697 from the last one. Now trailing whitespace is removed from each header
698 before concatenation. Completely empty headers in a concatenation (as
701 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
702 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
704 PH/29 [Removed. This was a change that I later backed out, and forgot to
705 correct the ChangeLog entry (that I had efficiently created) before
706 committing the later change.]
708 PH/30 Exim was sometimes attempting to deliver messages that had suffered
709 address errors (4xx response to RCPT) over the same connection as other
710 messages routed to the same hosts. Such deliveries are always "forced",
711 so retry times are not inspected. This resulted in far too many retries
712 for the affected addresses. The effect occurred only when there were more
713 hosts than the hosts_max_try setting in the smtp transport when it had
714 the 4xx errors. Those hosts that it had tried were not added to the list
715 of hosts for which the message was waiting, so if all were tried, there
716 was no problem. Two fixes have been applied:
718 (i) If there are any address or message errors in an SMTP delivery, none
719 of the hosts (tried or untried) are now added to the list of hosts
720 for which the message is waiting, so the message should not be a
721 candidate for sending over the same connection that was used for a
722 successful delivery of some other message. This seems entirely
723 reasonable: after all the message is NOT "waiting for some host".
724 This is so "obvious" that I'm not sure why it wasn't done
725 previously. Hope I haven't missed anything, but it can't do any
726 harm, as the worst effect is to miss an optimization.
728 (ii) If, despite (i), such a delivery is accidentally attempted, the
729 routing retry time is respected, so at least it doesn't keep
730 hammering the server.
732 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
733 in ${readsocket because some servers need this prod.
735 PH/32 Added some extra debug output when updating a wait-xxx database.
737 PH/33 The hint "could be header name not terminated by colon", which has been
738 given for certain expansion errors for a long time, was not being given
739 for the ${if def:h_colon_omitted{... case.
741 PH/34 The spec says: "With one important exception, whenever a domain list is
742 being scanned, $domain contains the subject domain." There was at least
743 one case where this was not true.
745 PH/35 The error "getsockname() failed: connection reset by peer" was being
746 written to the panic log as well as the main log, but it isn't really
747 panic-worthy as it just means the connection died rather early on. I have
748 removed the panic log writing for the ECONNRESET error when getsockname()
751 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
752 runs only) independently of the message's sender address. This meant
753 that, if the 4xx error was in fact related to the sender, a different
754 message to the same recipient with a different sender could confuse
755 things. In particualar, this can happen when sending to a greylisting
756 server, but other circumstances could also provoke similar problems.
757 I have changed the default so that the retry time for these errors is now
758 based a combination of the sender and recipient addresses. This change
759 can be overridden by setting address_retry_include_sender=false in the
762 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
763 remote server are returned as part of bounce messages. This was not
764 happening for LMTP over a pipe (the lmtp transport), but now it is the
765 same for both kinds of LMTP.
767 PH/38 Despite being documented as not happening, Exim was rewriting addresses
768 in header lines that were in fact CNAMEs. This is no longer the case.
770 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
771 and queue runs started by the daemon processed all messages. This has
772 been fixed so that -R and -S can now usefully be given with -q<time>.
774 PH/40 Import PCRE release 6.7 (fixes some bugs).
776 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
778 PH/42 Give an error if -q is specified more than once.
780 PH/43 Renamed the variables $interface_address and $interface_port as
781 $received_ip_address and $received_port, to make it clear that these
782 values apply to message reception, and not to the outgoing interface when
783 a message is delivered. (The old names remain recognized, of course.)
785 PH/44 There was no timeout on the connect() call when using a Unix domain
786 socket in the ${readsocket expansion. There now is.
788 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
789 be meaningful with "accept".
791 SC/01 Eximstats V1.43
792 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
794 SC/02 Eximstats V1.44
795 Use a glob alias rather than an array ref in the generated
796 parser. This improves both readability and performance.
798 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
799 Collect SpamAssassin and rejection statistics.
800 Don't display local sender or destination tables unless
801 there is data to show.
802 Added average volumes into the top table text output.
804 SC/04 Eximstats V1.46
805 Collect data on the number of addresses (recipients)
806 as well as the number of messages.
808 SC/05 Eximstats V1.47
809 Added 'Message too big' to the list of mail rejection
810 reasons (thanks to Marco Gaiarin).
812 SC/06 Eximstats V1.48
813 Mainlog lines which have GMT offsets and are too short to
814 have a flag are now skipped.
816 SC/07 Eximstats V1.49 (Alain Williams)
817 Added the -emptyok flag.
819 SC/08 Eximstats V1.50
820 Fixes for obtaining the IP address from reject messages.
822 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
823 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
824 whitesspace changes from 4.64-PH/27
826 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
829 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
830 are found, allow negative numbers in numeric criteria)
832 JJ/06 exipick.20061117.2, added new $message_body_missing variable
834 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
835 to match changes made in 4.64-PH/43
837 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
839 PH/47 Put in an explicit test for a DNS lookup of an address record where the
840 "domain" is actually an IP address, and force a failure. This locks out
841 those revolvers/nameservers that support "A-for-A" lookups, in
842 contravention of the specifications.
844 PH/48 When a host name was looked up from an IP address, and the subsequent
845 forward lookup of the name timed out, the host name was left in
846 $sender_host_name, contrary to the specification.
848 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
849 restricted to single-key lookups, Exim was not diagnosing an error if
850 * or *@ was used with a query-style lookup.
852 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
854 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
855 long ago, but noone interested enough thought of it. Let's just say that
856 the "1.1" means that there are some new functions that weren't there at
857 some point in the past.
859 PH/51 Error processing for expansion failure of helo_data from an smtp
860 transport during callout processing was broken.
862 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
863 tested/used via the -bh/-bhc/-bs options.
865 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
866 bug, fixed in subsequent PCRE releases).
868 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
869 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
871 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
877 SC/01 Use a glob alias rather than an array ref in eximstats generated
878 parser. This improves both readability and performance.
880 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
881 Don't display local sender or destination tables in eximstats unless
882 there is data to show.
883 Added average volumes into the eximstats top table text output.
885 SC/03 Collect data on the number of addresses (recipients) as well
886 as the number of messages in eximstats.
888 TF/01 Correct an error in the documentation for the redirect router. Exim
889 does (usually) call initgroups() when daemonizing.
891 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
892 with consistent privilege compared to when running as a daemon.
894 TF/03 Note in the spec that $authenticated_id is not set for local
895 submissions from trusted users.
897 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
898 Thanks to Dean Brooks <dean@iglou.com> for the patch.
900 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
901 by adding some example configuration directives to the default
902 configuration file. A little bit of work is required to uncomment the
903 directives and define how usernames and passwords are checked, but
904 there is now a framework to start from.
906 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
907 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
908 without this. I don't know how relevant this is to other LDAP libraries.
910 PH/02 Add the verb name to the "unknown ACL verb" error.
912 PH/03 Magnus Holmgren's patch for filter_prepend_home.
914 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
916 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
917 directory not expanded when it should be if an expanded home directory
918 was set for the address (which is overridden by the transport).
920 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
923 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
924 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
925 because it is too late at that time, and has no effect.
927 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
928 security issue with \' (bugzilla #107). I could not use the
929 PQescapeStringConn() function, because it needs a PGconn value as one of
932 PH/08 When testing addresses using -bt, indicate those final addresses that
933 are duplicates that would not cause an additional delivery. At least one
934 person was confused, thinking that -bt output corresponded to deliveries.
935 (Suppressing duplicates isn't a good idea as you lose the information
936 about possibly different redirections that led to the duplicates.)
938 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
939 systems where poll() doesn't work, in particular OS X.
941 PH/10 Added more information to debugging output for retry time not reached.
943 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
944 operations in malware.c.
946 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
949 PH/13 If write_rejectlog was set false when logging was sent to syslog with
950 syslog_duplication set false, log lines that would normally be written
951 both the the main log and to the reject log were not written to syslog at
954 PH/14 In the default configuration, change the use of "message" in ACL warn
955 statements to "add_header".
957 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
958 not followed by a command (e.g. "seen endif").
960 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
961 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
964 PH/17 Added extra conditions to the default value of delay_warning_condition
968 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
969 { match{$h_precedence:}{(?i)bulk|list|junk} } \
970 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
973 The Auto-Submitted: and various List- headers are standardised, whereas I
974 don't think Precedence: ever was.
976 PH/18 Refactored debugging code in route_finduser() to show more information,
977 in particular, the error code if getpwnam() issues one.
979 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
980 This is apparently needed in addition to the PH/07 change above to avoid
981 any possible encoding problems.
983 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
984 but not after initializing Perl.
986 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
987 output them only if debugging. By default they are written stderr,
988 apparently, which is not desirable.
990 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
993 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
996 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
998 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
999 authenticated or an ident call has been made. Suppress the default
1000 values for $authenticated_id and $authenticated_sender (but permit -oMai
1001 and -oMas) when testing with -bh.
1003 PH/24 Re-jigged the order of the tests in the default configuration so that the
1004 tests for valid domains and recipients precede the DNS black list and CSA
1005 tests, on the grounds that those ones are more expensive.
1007 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1008 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1009 This bug exists in every version of Exim that I still have, right back to
1012 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1013 However, an attempt to turn on case-sensitivity in a regex key by
1014 including (?-i) didn't work because the subject string was already
1015 lowercased, and the effects were non-intuitive. It turns out that a
1016 one-line patch can be used to allow (?-i) to work as expected.
1022 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1023 other effects) broke the use of negated acl sub-conditions.
1025 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1028 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1029 "Deny" causes Exim to reject the incoming connection with a 554 error.
1030 Unfortunately, if there is a major crisis, such as a disk failure,
1031 tcp-wrappers gives "deny", whereas what one would like would be some
1032 kind of temporary error. A kludge has been added to help with this.
1033 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1034 554 error is used if errno is still zero or contains ENOENT (which occurs
1035 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1038 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1040 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1041 errors. Otherwise a message that provokes a temporary error (when other
1042 messages do not) can cause a whole host to time out.
1044 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1045 addresses were routed directly to files or pipes from a redirect router.
1046 File deliveries just didn't batch; pipe deliveries might have suffered
1049 PH/06 A failure to get a lock for a hints database would erroneously always say
1050 "Failed to get write lock", even when it was really a read lock.
1052 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1053 of 0600. This has been changed to use the value of the lockfile_mode
1054 option (which defaults to 0600).
1056 PH/08 Applied small patch from the Sieve maintainer.
1058 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1059 folder from quota calculations, a direct delivery into this folder messed
1060 up the contents of the maildirsize file. This was because the regex was
1061 used only to exclude .Trash (or whatever) when the size of the mailbox
1062 was calculated. There was no check that a delivery was happening into an
1063 excluded directory. This bug has been fixed by ignoring all quota
1064 processing for deliveries into excluded directories.
1066 PH/10 Added the maildirfolder_create_regex option to appendfile.
1072 PH/01 The code for finding all the local interface addresses on a FreeBSD
1073 system running IPv6 was broken. This may well have applied to all BSD
1074 systems, as well as to others that have similar system calls. The broken
1075 code found IPv4 interfaces correctly, but gave incorrect values for the
1076 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1077 that it would not match correctly against @[] and not recognize the IPv6
1080 PH/02 The ipliteral router was not recognizing addresses of the form user@
1081 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1083 PH/03 Added disable_ipv6.
1085 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1086 decoded form, because it is most often used to construct To: headers
1087 lines in autoreplies, and the decoded form may well be syntactically
1088 invalid. However, $reply_address has leading white space removed, and all
1089 newlines turned into spaces so that the autoreply transport does not
1092 PH/05 If group was specified without a user on a router, and no group or user
1093 was specified on a transport, the group from the router was ignored.
1095 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1096 for visible compile-time settings that can be used to change these
1097 numbers, for those that want even more. Backwards compatibility with old
1098 spool files has been maintained. However, going back to a previous Exim
1099 release will lost any variables that are in spool files.
1101 PH/07 Two small changes when running in the test harness: increase delay when
1102 passing a TCP/IP connection to a new process, in case the original
1103 process has to generate a bounce, and remove special handling of
1104 127.0.0.2 (sic), which is no longer necessary.
1106 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1107 be the same on different OS.
1109 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1112 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1113 whether --show-vars was specified or not
1115 JJ/02 exipick: Added support for new ACL variable spool format introduced
1118 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1119 syntactically invalid From: or Reply-to: line, and a filter used this to
1120 generate an autoreply, and therefore failed to obtain an address for the
1121 autoreply, Exim could try to deliver to a non-existent relative file
1122 name, causing unrelated and misleading errors. What now happens is that
1123 it logs this as a hard delivery error, but does not attempt to create a
1126 PH/11 The exinext utility has a -C option for testing purposes, but although
1127 the given file was scanned by exinext itself; it wasn't being passed on
1128 when Exim was called.
1130 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1131 an end-of-file indication when reading a command response.
1133 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1134 compiled. In many other places in Exim, IPv6 addresses are always
1135 recognized, so I have changed this. It also means that IPv4 domain
1136 literals of the form [IPV4:n.n.n.n] are now always recognized.
1138 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1139 used if the router is not running as root, for example, when verifying at
1140 ACL time, or when using -bh. The debugging output from this situation was
1141 non-existent - all you got was a failure to exec. I have made two
1144 (a) Failures to set uid/gid, the current directory, or a process leader
1145 in a subprocess such as that created by queryprogram now generate
1146 suitable debugging ouput when -d is set.
1148 (b) The queryprogram router detects when it is not running as root,
1149 outputs suitable debugging information if -d is set, and then runs
1150 the subprocess without attempting to change uid/gid.
1152 PH/15 Minor change to Makefile for building test_host (undocumented testing
1155 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1156 additional section of a DNS packet that returns MX or SRV records.
1157 Instead, it always explicitly searches for A/AAAA records. This avoids
1158 major problems that occur when a DNS server includes only records of one
1159 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1160 fixed another bug: if SRV records were looked up and the corresponding
1161 address records were *not* found in the additional section, the port
1162 values from the SRV records were lost.
1164 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1165 using the correct key (the original address) when searching the retry
1166 rules in order to find which one to use for generating the retry hint.
1168 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1169 adding the default one. Similarly, if it contains a Reply-To: header, the
1170 errors_reply_to option, if set, is not used.
1172 PH/19 When calculating a retry time, Exim used to measure the "time since
1173 failure" by looking at the "first failed" field in the retry record. Now
1174 it does not use this if it is later than than the arrival time of the
1175 message. Instead it uses the arrival time. This makes for better
1176 behaviour in cases where some deliveries succeed, thus re-setting the
1177 "first failed" field. An example is a quota failure for a huge message
1178 when small messages continue to be delivered. Without this change, the
1179 "time since failure" will always be short, possible causing more frequent
1180 delivery attempts for the huge message than are intended.
1181 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1183 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1184 $1, $2, $3) because the numerical variables can be reset during some
1185 expansion items (e.g. "match"), thereby losing the authentication data.
1187 PH/21 Make -bV show the size of off_t variables so that the test suite can
1188 decide whether to run tests for quotas > 2G.
1190 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1191 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1192 filecount value is greater than 2G or if a quota value is greater than 2G
1193 on a system where the size of off_t is not greater than 4, a panic error
1196 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1197 never match. The debug and -bh output now contains an explicit error
1198 message indicating a malformed IPv4 address or mask.
1200 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1201 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1202 PH/23 above applies.
1204 PH/25 Do not write to syslog when running in the test harness. The only
1205 occasion when this arises is a failure to open the main or panic logs
1206 (for which there is an explicit test).
1208 PH/26 Added the /no_tell option to "control=freeze".
1210 PH/27 If a host name lookup failed very early in a connection, for example, if
1211 the IP address matched host_lookup and the reverse lookup yielded a name
1212 that did not have a forward lookup, an error message of the form "no IP
1213 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1214 could be logged. Now it outputs the IP address instead of "NULL".
1216 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1217 allows the sender and the authenticated sender to be set when
1218 submitting a message from within Exim. Since child_open_exim() is
1219 documented for local_scan(), the new function should be too.
1221 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1222 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1223 results in an empty string is now treated as unset.
1225 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1227 PH/31 Added sender_verify_fail logging option.
1229 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1230 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1231 tidied the source and removed it altogether.
1233 PH/33 When a queue run was abandoned because the load average was too high, a
1234 log line was always written; now it is written only if the queue_run log
1235 selector is set. In addition, the log line for abandonment now contains
1236 information about the queue run such as the pid. This is always present
1237 in "start" and "stop" lines but was omitted from the "abandon" line.
1239 PH/34 Omit spaces between a header name and the colon in the error message that
1240 is given when verify = headers_syntax fails (if there are lots of them,
1241 the message gets confusing).
1243 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1244 names, as there are now some PTR records that contain slashes. This check
1245 is only to protect against broken name servers that fall over on strange
1246 characters, so the fact that it applies to all lookups doesn't matter.
1248 PH/36 Now that the new test suite is complete, we can remove some of the
1249 special code in Exim that was needed for the old test suite. For example,
1250 sorting DNS records because real resolvers return them in an arbitrary
1251 order. The new test suite's fake resolver always returns records in the
1254 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1255 bounces) except when queue_only is set, to avoid logging races between
1256 the different processes.
1258 PH/38 Panic-die if .include specifies a non-absolute path.
1260 PH/39 A tweak to the "H" retry rule from its user.
1262 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1263 a label. They prevented compilation on older perls.
1265 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1266 a warning to be raised on newish perls.
1268 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1269 on queue. Changes to match documented behaviour of showing count of
1270 messages matching specified criteria.
1272 PH/40 Changed the default ident timeout from 30s to 5s.
1274 PH/41 Added support for the use of login_cap features, on those BSD systems
1275 that have them, for controlling the resources used by pipe deliveries.
1277 PH/42 The content-scanning code uses fopen() to create files in which to put
1278 message data. Previously it was not paying any attention to the mode of
1279 the files. Exim runs with umask(0) because the rest of the code creates
1280 files with open(), and sets the required mode explicitly. Thus, these
1281 files were ending up world-writeable. This was not a big issue, because,
1282 being within the spool directory, they were not world-accessible. I have
1283 created a function called modefopen, which takes an additional mode
1284 argument. It sets umask(777), creates the file, chmods it to the required
1285 mode, then resets the umask. All the relevant calls to fopen() in the
1286 content scanning code have been changed to use this function.
1288 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1289 to 24 hours. This avoids potential overflow problems when processing G
1290 and H retry rules. I suspect nobody ever tinkers with this value.
1292 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1294 PH/45 When the plaintext authenticator is running as a client, the server's
1295 challenges are checked to ensure they are valid base64 strings. By
1296 default, the authentication attempt is cancelled if an invalid string is
1297 received. Setting client_ignore_invalid_base64 true ignores these errors.
1298 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1299 they are received. Thus, the responses can be made to depend on the
1300 challenges. If an invalid string is ignored, an empty string is placed in
1303 PH/46 Messages that are created by the autoreply transport now contains a
1304 References: header, in accordance with RFCs 2822 and 3834.
1306 PH/47 Added authenticated_sender_force to the smtp transport.
1308 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1310 PH/49 Installed latest patch from the Sieve maintainer.
1312 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1313 was also set, the appendfile transport was unnecessarily scanning a
1314 directory of message files (e.g. for maildir delivery) to find the count
1315 of files (along with the size), even though it did not need this
1316 information. It now does the scan only if it needs to find either the
1317 size of the count of files.
1319 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1321 PH/52 Two bugs concerned with error handling when the smtp transport is
1324 (i) Exim was not creating retry information for temporary errors given
1325 for individual recipients after the DATA command when the smtp transport
1326 was used in LMTP mode. This meant that they could be retried too
1327 frequently, and not timed out correctly.
1329 (ii) Exim was setting the flag that allows error details to be returned
1330 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1331 recipients that were returned after the DATA command.
1333 PH/53 This is related to PH/52, but is more general: for any failing address,
1334 when detailed error information was permitted to be returned to the
1335 sender, but the error was temporary, then after the final timeout, only
1336 "retry timeout exceeded" was returned. Now it returns the full error as
1337 well as "retry timeout exceeded".
1339 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1340 do this, and (what is worse) MTAs that accept it.
1342 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1343 will now be deprecated.
1345 PH/56 New os.c-cygwin from the Cygwin maintainer.
1347 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1348 formats (previously only available in exim formats via -bpr, -bpru,
1349 and -bpra. Now also available in native and exiqgrep formats)
1351 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1352 with very large, slow to parse queues
1354 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1356 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1358 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1359 responses to authentication challenges, though it was showing the
1360 challenges; (ii) I've removed the CR characters from the debug output for
1363 PH/58 Allow for the insertion of a newline as well as a space when a string
1364 is turned into more than one encoded-word during RFC 2047 encoding. The
1365 Sieve code now uses this.
1367 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1368 data_4xx, lost_connection, tls_required.
1370 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1371 message was being sent as an SMTP response.
1373 PH/61 Add -l and -k options to exicyclog.
1375 PH/62 When verifying, if an address was redirected to one new address, so that
1376 verification continued, and the new address failed or deferred after
1377 having set something in $address_data, the value of $address_data was not
1378 passed back to the ACL. This was different to the case when no
1379 redirection occurred. The value is now passed back in both cases.
1381 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1382 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1383 use login_cap.h, so on its own it isn't the distinguishing feature. The
1384 new name refers directly to the setclassresources() function.
1386 PH/65 Added configuration files for NetBSD3.
1388 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1390 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1391 is preferred over IPv4.
1393 PH/68 The bounce_return_message and bounce_return_body options were not being
1394 honoured for bounces generated during the reception of non-SMTP messages.
1395 In particular, this applied to messages rejected by the ACL. This bug has
1396 been fixed. However, if bounce_return_message is true and bounce_return_
1397 body is false, the headers that are returned for a non-SMTP message
1398 include only those that have been read before the error was detected.
1399 (In the case of an ACL rejection, they have all been read.)
1401 PH/69 The HTML version of the specification is now built in a directory called
1402 spec_html instead of spec.html, because the latter looks like a path with
1403 a MIME-type, and this confuses some software.
1405 PH/70 Catch two compiler warnings in sieve.c.
1407 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1408 function verify_get_ident() calls ip_connect() to connect a socket, but
1409 if the "connect()" function timed out, ip_connect() used to close the
1410 socket. However, verify_get_ident() also closes the socket later, and in
1411 between Exim writes to the log, which may get opened at this point. When
1412 the socket was closed in ip_connect(), the log could get the same file
1413 descriptor number as the socket. This naturally causes chaos. The fix is
1414 not to close the socket in ip_connect(); the socket should be closed by
1415 the function that creates it. There was only one place in the code where
1416 this was missing, in the iplookup router, which I don't think anybody now
1417 uses, but I've fixed it anyway.
1419 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1420 well as to direct DNS lookups. Otherwise the handling of names in host
1421 lists is inconsistent and therefore confusing.
1427 PH/01 Two changes to the default runtime configuration:
1429 (1) Move the checks for relay_from_hosts and authenticated clients from
1430 after to before the (commented out) DNS black list checks.
1432 (2) Add control=submission to the relay_from_hosts and authenticated
1433 clients checks, on the grounds that messages accepted by these
1434 statements are most likely to be submissions.
1436 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1438 (1) Generate an error if the third argument for the ${prvs expansion is
1441 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1444 (3) Reset the variables that are obtained from the first argument of
1445 ${prvscheck and used in the second argument before leaving the code,
1446 because their memory is reclaimed, so using them afterwards may do
1449 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1450 one (it's much easier than Tom thought :-).
1452 (5) Because of (4), we can now allow for the use of $prvscheck_result
1453 inside the third argument.
1455 PH/03 For some reason, the default setting of PATH when running a command from
1456 a pipe transport was just "/usr/bin". I have changed it to
1459 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1460 anything to be listed in the output from -bV.
1462 PH/05 When a filter generated an autoreply, the entire To: header line was
1463 quoted in the delivery log line, like this:
1465 => >A.N.Other <ano@some.domain> <original@ddress> ...
1467 This has been changed so that it extracts the operative address. There
1468 may be more than one such address. If so, they are comma-separated, like
1471 => >ano@some.domain,ona@other.domain <original@ddress> ...
1473 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1474 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1475 not being looked up in the rDNS to get a host name, Exim was showing the
1476 IP address twice in Received: lines, even though the IP addresses were
1477 identical. For example:
1479 Received: from [1.2.3.4] (helo=[1.2.3.4])
1481 However, if the real host name was known, it was omitting the HELO data
1482 if it matched the actual IP address. This has been tidied up so that it
1483 doesn't show the same IP address twice.
1485 PH/07 When both +timestamp and +memory debugging was on, the value given by
1486 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1487 called by the debug printing, thereby overwriting the timestamp buffer.
1488 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1490 PH/08 When the original message was included in an autoreply transport, it
1491 always said "this is a copy of the message, including all the headers",
1492 even if body_only or headers_only was set. It now gives an appropriate
1495 PH/09 Applied a patch from the Sieve maintainer which:
1497 o fixes some comments
1498 o adds the (disabled) notify extension core
1499 o adds some debug output for the result of if/elsif tests
1500 o points to the current vacation draft in the documentation
1501 and documents the missing references header update
1505 o fixes a bug in processing the envelope test (when testing
1506 multiple envelope elements, the last element determinted the
1509 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1510 Electronic Mail") by including:
1512 Auto-submitted: auto-generated
1514 in the messages that it generates (bounce messages and others, such as
1515 warnings). In the case of bounce messages for non-SMTP mesages, there was
1516 also a typo: it was using "Auto_submitted" (underscore instead of
1517 hyphen). Since every message generated by Exim is necessarily in response
1518 to another message, thes have all been changed to:
1520 Auto-Submitted: auto-replied
1522 in accordance with these statements in the RFC:
1524 The auto-replied keyword:
1526 - SHOULD be used on messages sent in direct response to another
1527 message by an automatic process,
1529 - MUST NOT be used on manually-generated messages,
1531 - MAY be used on Delivery Status Notifications (DSNs) and Message
1532 Disposition Notifications (MDNs),
1534 - MUST NOT be used on messages generated by automatic or periodic
1535 processes, except for messages which are automatic responses to
1538 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1539 to the default Received: header definition.
1541 PH/12 Added log selector acl_warn_skipped (default on).
1543 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1544 variables because (a) they are in the wrong storage pool and (b) even if
1545 they were copied, it wouldn't work properly because of the caching.
1547 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1548 checking when decoding. Apparently there are clients that generate
1549 overlong encoded strings. Why am I not surprised?
1551 PH/15 If the first argument of "${if match_address" was not empty, but did not
1552 contain an "@" character, Exim crashed. Now it writes a panic log message
1553 and treats the condition as false.
1555 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1557 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1558 extension "envelope-auth". The code is finished and in agreement with
1559 other implementations, but there is no documentation so far and in fact,
1560 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1561 not changing the active code.
1563 Print executed "if" and "elsif" statements when debugging is used. This
1564 helps a great deal to understand what a filter does.
1566 Document more things not specified clearly in RFC3028. I had all this
1567 sorted out, when out of a sudden new issues came to my mind. Oops."
1569 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1572 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1573 canonical form (as documented). However, after a host name lookup from
1574 the IP address, check_host() was doing a simple string comparison with
1575 addresses acquired from the DNS when checking that the found name did
1576 have the original IP as one of its addresses. Since any found IPv6
1577 addresses are likely to be in abbreviated form, the comparison could
1578 fail. Luckily, there already exists a function for doing the comparison
1579 by converting both addresses to binary, so now that is used instead of
1580 the text comparison.
1582 PH/20 There was another similar case to PH/19, when a complete host name was
1583 given in a host list; looking up its IP address could give an abbreviated
1584 form, whereas the current host's name might or might not be abbreviated.
1585 The same fix has been applied.
1591 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1592 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1595 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1596 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1598 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1600 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1601 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1602 Also, for "Auto-Submitted", treat anything other than "no" as
1603 non-personal, in accordance with RFC 3834. (Previously it treated
1604 anything starting "auto-" as non-personal.)
1606 TF/01 The control=submission/name=... option had a problem with syntax
1607 errors if the name included a slash character. The /name= option
1608 now slurps the rest of the string, so it can include any characters
1609 but it must come last in the list of options (after /sender_retain
1612 PH/05 Some modifications to the interface to the fake nameserver for the new
1620 TK/01 Added the "success_on_redirect" address verification option. See
1621 NewStuff for rationale and an example.
1623 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1625 PH/02 Patch to exigrep to allow it to work on syslog lines.
1627 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1628 fread() to skip over the body file's header line, because in Cygwin the
1629 header line is locked and is inaccessible.
1631 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1632 co-exist for some time) to make it clear that it is the Exim ID that is
1633 referenced, not the Message-ID: header line.
1635 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1636 string_format() function, because snprintf() does not exist on all
1639 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1640 ${stat: expansion item.
1642 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1643 protocol synchronization error", to keep the pedants happy.
1645 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1646 well as for IRIX systems, when gcc is being used. See the host.c source
1649 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1651 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1654 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1655 transport and to the smtp transport in LMTP mode.
1657 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1659 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1660 run a filter in a subprocess. This could lead to confusion in subsequent
1661 lookups in the parent process. There should also be a search_tidyup() at
1662 the end of the subprocess.
1664 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1665 only if the host matched helo_try_verify_hosts, which caused the
1666 verification to occur when the EHLO/HELO command was issued. The ACL just
1667 tested the remembered result. Now, if a previous verification attempt has
1668 not happened, "verify = helo" does it there and then.
1670 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1672 TK/03 Fix log output including CR from clamd.
1674 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1675 exist provoked a memory error which could cause a segfault.
1677 PH/15 Installed PCRE 6.2
1679 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1681 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1682 of the problem. Specifically, suggested +O2 rather than +O1 for the
1685 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1687 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1688 record, the port was not correctly logged when the outgoing_port log
1689 selector was set (it logged the transort's default port).
1691 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1692 fallback_hosts, and "hosts" in the smtp transport.
1694 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1695 host errors such as "Connection refused".
1697 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1698 authentication with radiusclient 0.4.9:
1700 - Error returned from rc_read_config was caught wrongly
1701 - Username/password not passed on to radius server due to wrong length.
1703 The presumption is that some radiusclient API changes for 4.51/PH/17
1704 were not taken care of correctly. The code is still untested by me (my
1705 Linux distribution still has 0.3.2 of radiusclient), but it was
1706 contributed by a Radius user.
1708 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1709 expanding the "port" option of the smtp transport.
1711 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1712 while reading a MIME header. Thanks to Tom Hughes for a patch.
1714 PH/24 Include config.h inside local_scan.h so that configuration settings are
1717 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1718 that in an ACL for RCPT (for example), you can examine exactly what was
1721 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1722 commands, but it was not correctly comparing the address with the actual
1723 client host address. Thus, it would show the EHLO address in Received:
1724 header lines when this was not necessary.
1726 PH/27 Added the % operator to ${eval:}.
1728 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1729 it should be ignoring failures (because with -C, for example, it has lost
1730 privilege). It wasn't ignoring creation failures other than "already
1733 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1736 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1737 because an input line was too long, either on its own, or by virtue of
1738 too many continuations, the temporary file was not being removed, and the
1739 return code was incorrect.
1741 PH/31 Missing "BOOL" in function definition in filtertest.c.
1743 PH/32 Applied Sieve patches from the maintainer.
1745 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1747 PH/33 Added "verify = not_blind".
1749 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1750 Local/Makefile (with some defaults set). These are used in built scripts
1751 such as exicyclog, but they have never been used in the exim_install
1752 script (though there are many overriding facilities there). I have
1753 arranged that the exim_install script now takes note of these two
1756 PH/35 Installed configuration files for Dragonfly.
1758 PH/36 When a locally submitted message by a trusted user did not contain a
1759 From: header, and the sender address was obtained from -f or from an SMTP
1760 MAIL command, and the trusted user did not use -F to supply a sender
1761 name, $originator_name was incorrectly used when constructing a From:
1762 header. Furthermore, $originator_name was used for submission mode
1763 messages from external hosts without From: headers in a similar way,
1764 which is clearly wrong.
1766 PH/37 Added control=suppress_local_fixups.
1768 PH/38 When log_selector = +received_sender was set, and the addition of the
1769 sender made the log line's construction buffer exactly full, or one byte
1770 less than full, an overflow happened when the terminating "\n" was
1773 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1774 when the result of a list match is failure because a DNS lookup failed.
1776 PH/40 RM_COMMAND is now used in the building process.
1778 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1779 the "build-* directories that it finds.
1781 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1782 address in a domain literal was a prefix of an interface address.
1784 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1785 when verifying a sender address, unless rewrite_headers is false.
1787 PH/44 Wrote a long comment about why errors_to addresses are verified as
1788 recipients, not senders.
1790 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1791 the ratelimit ACL was added.
1793 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1795 PH/46 Added extra information about PostgreSQL errors to the error string.
1797 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1798 avoiding the need to install special zones in a real server. This is
1799 backwards compatible; if it can't find the fake resolver, it drops back.
1800 Thus, both old and new test suites can be run.
1802 TF/02 Added util/ratelimit.pl
1804 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1805 clock is set back in time.
1807 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1808 Candler <B.Candler@pobox.com>.
1810 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1811 OK for addresses that are the result of redirections.
1813 PH/48 A number of further additions for the benefit of the new test suite,
1814 including a fake gethostbyname() that interfaces to the fake DNS resolver
1817 TF/06 The fix for widen_domains has also been applied to qualify_single and
1818 search_parents which are the other dnslookup options that can cause
1821 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1824 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1826 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1827 service" if it encountered an MX record with an empty target hostname.
1828 The message is now "an MX or SRV record indicated no SMTP service".
1830 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1831 if the DNS of the sending site is misconfigured. This is quite a
1832 common situation. This change restores the behaviour of treating a
1833 helo verification defer as a failure.
1835 PH/51 If self=fail was set on a router, the bounce message did not include the
1836 actual error message.
1842 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1844 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1845 command itself ended in error, the underlying message about the transport
1846 filter timeout was being overwritten with the pipe command error. Now the
1847 underlying error message should be appended to the second error message.
1849 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1851 PH/02 Reduce the amount of output that "make" produces by default. Full output
1852 can still be requested.
1854 PH/03 The warning log line about a condition test deferring for a "warn" verb
1855 was being output only once per connection, rather than after each
1856 occurrence (because it was using the same function as for successful
1857 "warn" verbs). This seems wrong, so I have changed it.
1859 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1860 it should not have, which might have caused a crash in the right
1861 circumstances, but probably never did.
1863 PH/04 Installed a modified version of Tony Finch's patch to make submission
1864 mode fix the return path as well as the Sender: header line, and to
1865 add a /name= option so that you can make the user's friendly name appear
1868 TF/03 Added the control = fakedefer ACL modifier.
1870 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1871 Mark Lowes for thorough testing.
1873 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1875 TK/03 Merged latest SRS patch from Miles Wilton.
1877 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1878 with the definition in sysexits.h (which is #included earlier).
1879 Fortunately, Exim does not actually use EX_OK. The code used to try to
1880 preserve the sysexits.h value, by assumimg that macro definitions were
1881 scanned for macro replacements. I have been disabused of this notion,
1882 so now the code just undefines EX_OK before #including unistd.h.
1884 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1885 in the smtp transport. When a block could not be written in a single
1886 write() function, the timeout was being re-applied to each part-write.
1887 This seems wrong - if the receiver was accepting one byte at a time it
1888 would take for ever. The timeout is now adjusted when this happens. It
1889 doesn't have to be particularly precise.
1891 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1892 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1894 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1895 without a domain if the check to <postmaster@domain> fails.
1897 SC/01 Eximstats: added -xls and the ability to specify output files
1898 (patch written by Frank Heydlauf).
1900 SC/02 Eximstats: use FileHandles for outputing results.
1902 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1904 SC/04 Eximstats: fixed display of large numbers with -nvr option
1906 SC/05 Eximstats: fixed merging of reports with empty tables.
1908 SC/06 Eximstats: added the -include_original_destination flag
1910 SC/07 Eximstats: removed tabs and trailing whitespace.
1912 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1914 TK/06 MBOX spool code: Add real "From " MBOX separator line
1915 so the .eml file is really in mbox format (even though
1916 most programs do not really care). Patch from Alex Miller.
1918 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1919 The latter is generated from $received_to and is only set if the
1920 message has one envelope recipient. SA can use these headers,
1921 obviously out-of-the-box. Patch from Alex Miller.
1923 PH/08 The ${def test on a variable was returning false if the variable's
1924 value was "0", contrary to what the specification has always said!
1925 The result should be true unless the variable is empty.
1927 PH/09 The syntax error of a character other than { following "${if
1928 def:variable_name" (after optional whitespace) was not being diagnosed.
1929 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1930 accidental colon was present, for example, could give incorrect results.
1932 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1933 result is used (not including appendfile, where other changes are about
1936 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1937 This involved changing a lot of size variables from int to off_t. It
1938 should work with maildirs and everything.
1940 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1941 spamd dying while we are connected to it.
1943 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1946 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1947 function for BDB 4.3.
1949 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1951 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1952 the sources for PCRE have been re-arranged and more files are now
1955 PH/15 The code I had for printing potentially long long variables in PH/11
1956 above was not the best (it lost precision). The length of off_t variables
1957 is now inspected at build time, and an appropriate printing format (%ld
1958 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1959 to be "long long int" or "long int". This is needed for the internal
1960 formatting function string_vformat().
1962 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1963 the configuration file to be ":syslog", then the script "guesses" where
1964 the logs files are, rather than using the compiled in default. In our
1965 case the guess is not the same as the compiled default, so the script
1966 suddenly stopped working when I started to use syslog. The patch checks
1967 to see if log_file_path is "". If so, it attempts to read it from exim
1968 with no configuration file to get the compiled in version, before it
1969 falls back to the previous guessing code."
1971 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1972 implementing BATV in an Exim configuration. See NewStuff for the gory
1975 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1976 Makefile that are specific to HP-UX.
1978 PH/18 If the "use_postmaster" option was set for a recipient callout together
1979 with the "random" option, the postmaster address was used as the MAIL
1980 FROM address for the random test, but not for the subsequent recipient
1981 test. It is now used for both.
1983 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1984 patch removes a few documentation additions to RFC 3028, because the
1985 latest draft now contains them. It adds the new en;ascii-case comparator
1986 and a new error check for 8bit text in MIME parts. Comparator and
1987 require names are now matched exactly. I enabled the subaddress
1988 extension, but it is not well tested yet (read: it works for me)."
1990 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1991 rework some of the code of TK/09 above to avoid the hardwired use of
1992 "%lld" and "long long". Replaced the call to snprintf() with a call to
1995 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1996 records point to non-existent hosts", "retry timeout exceeded", and
1997 "retry time not reached for any host after a long failure period".
1999 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2000 experimental DomainKeys support:
2002 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2003 (2) On an error such as an illegally used "control", the wrong name for
2004 the control was given.
2006 These problems did NOT occur unless DomainKeys support was compiled.
2008 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2010 PH/24 Added ${if match_ip condition.
2012 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2013 databases so that it will be absolutely obvious if a crash occurs in the
2014 DB library. This is a regular occurrence (often caused by mis-matched
2017 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2018 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2019 picked up on a user's system that detects such things. There doesn't seem
2020 to be a gcc warning option for this - only an attribute that has to be
2021 put on the function's prototype. It seems that in Fedora Core 4 they have
2022 set this on a number of new functions. No doubt there will be more in due
2025 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2026 specify a transport. However, if an address that was verified by such a
2027 router was the subject of a callout, Exim crashed because it tried to
2028 read the rcpt_include_affixes from the non-existent transport. Now it
2029 just assumes that the setting of that option is false. This bug was
2030 introduced by 4.51/PH/31.
2032 PH/28 Changed -d+all to exclude +memory, because that information is very
2033 rarely of interest, but it makes the output a lot bigger. People tend to
2034 do -d+all out of habit.
2036 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2037 code in os-type was giving problems when libc.so lives in lib64, like on
2040 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2041 aren't the modern standard, and it seems that some systems' include files
2042 don't always have them. Exim was already checking for some of the newer
2043 ones like T_AAAA, and defining it itself. I've added checks for all the
2044 record types that Exim uses.
2046 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2047 not automatically generating a new one, as it is supposed to. This
2048 prevented TLS from working. If the file did exist, but contained invalid
2049 data, a new version was generated, as expected. It was only the case of a
2050 non-existent file that was broken.
2052 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2053 with a change in libdomainkeys > 0.64.
2055 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2056 from DNS. If the selector record carries the flag, it now has
2057 precedence over the domain-wide flag.
2059 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2061 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2062 the use of a port name that isn't defined in /etc/services) occurred, the
2063 message was deferred as in a normal delivery, and thus remained on the
2064 spool, instead of being failed because of the mua_wrapper setting. This
2065 is now fixed, and I tidied up some of the mua_wrapper messages at the
2068 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2069 the messages in a hash of arrays rather than using individual hashes.
2070 This is a bit cleaner and results in dramatic memory savings, albeit
2071 at a slight CPU cost.
2073 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2074 as requested by Marc Sherman.
2076 SC/10 Eximstats: added histograms for user specified patterns as requested
2079 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2081 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2082 fopen() in the content-scanning modules that did not already have it.
2088 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2089 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2091 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2093 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2095 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2096 to test Sieve filters that use "vacation".
2098 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2099 that changes the way the GnuTLS parameters are stored in the cache file.
2100 The new format can be generated externally. For backward compatibility,
2101 if the data in the cache doesn't make sense, Exim assumes it has read an
2102 old-format file, and it generates new data and writes a new file. This
2103 means that you can't go back to an older release without removing the
2106 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2107 work if there are any delivery delays because "one_time" forces the
2108 parent to be marked "delivered", so its unseen clone is never tried
2109 again. For this reason, Exim now forbids the simultaneous setting of
2112 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2113 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2114 entry if you want to know the details. The fix, however, neglected to
2115 consider the case when local delivery batching is involved. The test for
2116 "previously delivered" was not happening when checking to see if an
2117 address could be batched with a previous (undelivered) one; under
2118 certain circumstances this could lead to multiple deliveries to the same
2121 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2122 in its include files, and this causes problems building Exim.
2124 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2125 header_syntax) but Exim was just ignoring anything given after a slash.
2126 In particular, this caused confusion with an attempt to use "verify =
2127 reverse_host_lookup/defer_ok". An error is now given when options are
2128 supplied for verify items that do not have them. (Maybe reverse_host_
2129 lookup should have a defer_ok option, but that's a different point.)
2131 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2132 defined by RFC 821) to 2048, because there were problems with some AUTH
2133 commands, and RFC 1869 says the size should be increased for extended
2134 SMTP commands that take arguments.
2136 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2139 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2140 "unknown" error; now it says that the functionality isn't in the binary.
2142 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2143 an address' error message when a string expansion fails (syntax or
2144 whatever). Otherwise the password may appear in the log. Following change
2145 PH/42 below, there is no longer a chance of it appearing in a bounce
2148 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2150 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2151 of its addresses was ever tried. (Bugzilla bug #2.)
2153 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2154 the result incorrectly in the debug output. (It correctly added a newline
2155 to what was transported.)
2157 TF/01 Added $received_time.
2159 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2160 commented out examples of how to interface to a virus scanner and to
2161 SpamAssassin. Also added commented examples of av_scanner and
2162 spamd_address settings.
2164 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2165 and controls are allowed in which ACLs. There were a couple of minor
2166 errors. Some of the entries in the conditions table (which is a table of
2167 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2168 negation of where the condition IS allowed.
2170 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2172 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2173 header file does not have a version number, so I've had to invent a new
2174 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2175 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2176 radiusclient), but it was contributed by a Radius user.
2178 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2179 files or directories, for OpenSSL.
2181 PH/19 When an Exim process that is running as root has to create an Exim log
2182 file, it does so in a subprocess that runs as exim:exim so as to get the
2183 ownership right at creation (otherwise, other Exim processes might see
2184 the file with the wrong ownership). There was no test for failure of this
2185 fork() call, which would lead to the process getting stuck as it waited
2186 for a non-existent subprocess. Forks do occasionally fail when resources
2187 run out. I reviewed all the other calls to fork(); they all seem to check
2190 PH/20 When checking for unexpected SMTP input at connect time (before writing
2191 the banner), Exim was not dealing correctly with a non-positive return
2192 from the read() function. If the client had disconnected by this time,
2193 the result was a log entry for a synchronization error with an empty
2194 string after "input=" when read() returned zero. If read() returned -1
2195 (an event I could not check), uninitialized data bytes were printed.
2196 There were reports of junk text (parts of files, etc) appearing after
2199 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2201 PH/22 Added support for macro redefinition, and (re)definition in between
2202 driver and ACL definitions.
2204 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2205 forgetting to use the resulting value; it was using the unexpanded value.
2207 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2208 hadn't been configured. The fix is from Juergen Kreileder, who
2209 understands it better than I do:
2211 "Here's what I see happening with three configured cyrus_sasl
2212 authenticators configured (plain, login, cram-md5):
2214 On startup auth_cyrus_sasl_init() gets called for each of these.
2215 This means three calls to sasl_listmech() without a specified mech_list.
2216 => SASL tests which mechs of all available mechs actually work
2217 => three warnings about OTP not working
2218 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2220 With the patch, sasl_listmech() also gets called three times. But now
2221 SASL's mech_list option is set to the server_mech specified in the the
2222 authenticator. Or in other words, the answer from sasl_listmech()
2223 gets limited to just the mech you're testing for (which is different
2225 => the return list contains just 'plain' or 'login', 'cram-md5' or
2226 nothing depending on the value of ob->server_mech.
2228 I've just tested the patch: Authentication still works fine,
2229 unavailable mechs specified in the exim configuration are still
2230 caught, and the auth.log warnings about OTP are gone."
2232 PH/25 When debugging is enabled, the contents of the command line are added
2233 to the debugging output, even when log_selector=+arguments is not
2236 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2237 answer is "GNU", and only if the return is "GNU/something" is the answer
2240 PH/27 $acl_verify_message is now set immediately after the failure of a
2241 verification in an ACL, and so is available in subsequent modifiers. In
2242 particular, the message can be preserved by coding like this:
2244 warn !verify = sender
2245 set acl_m0 = $acl_verify_message
2247 Previously, $acl_verify_message was set only while expanding "message"
2248 and "log_message" when a very denied access.
2250 PH/28 Modified OS/os.c-Linux with
2252 -#ifndef OS_LOAD_AVERAGE
2253 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2255 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2256 nomenclature these days.)
2258 PH/29 Installed patch from the Sieve maintainer that adds the options
2259 sieve_useraddress and sieve_subaddress to the redirect router.
2261 PH/30 In these circumstances:
2262 . Two addresses routed to the same list of hosts;
2263 . First host does not offer TLS;
2264 . First host accepts first address;
2265 . First host gives temporary error to second address;
2266 . Second host offers TLS and a TLS session is established;
2267 . Second host accepts second address.
2268 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2269 and peerdn, if requested) that were in fact used only for the second
2272 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2273 attention to any local part prefix or suffix that was matched by the
2274 router that accepted the address. It now behaves in the same way as it
2275 does for delivery: the affixes are removed from the local part unless
2276 rcpt_include_affixes is set on the transport.
2278 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2279 timeout during the DATA phase of an incoming message.
2281 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2282 applied a patch sanctioned by the Sieve maintainer.
2284 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2285 the uid or gid is negative. A case of a negative gid caused this to be
2286 noticed. The fix allows for either to be negative.
2288 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2289 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2292 PH/36 The change PH/12 above was broken. Fixed it.
2294 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2295 the grounds that routing the same address twice would always produce the
2296 same answer. This might have been true once, but it is certainly no
2297 longer true now. Routing a child address may depend on the previous
2298 routing that produced that child. Some complicated redirection strategies
2299 went wrong when messages had multiple recipients, and made Exim's
2300 behaviour dependent on the order in which the addresses were given.
2302 I have moved the duplicate checking until after the routing is complete.
2303 Exim scans the addresses that are assigned to local and remote
2304 transports, and removes any duplicates. This means that more work will be
2305 done, as duplicates will always all be routed, but duplicates are
2306 presumably rare, so I don't expect this is of any significance.
2308 For deliveries to pipes, files, and autoreplies, the duplicate checking
2309 still happens during the routing process, since they are not going to be
2312 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2313 It corrects a timeout issue with spamd. This is Ian's comment: "The
2314 background is that sometimes spamd either never reads data from a
2315 connection it has accepted, or it never writes response data. The exiscan
2316 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2317 blindly assumes that writes won't block so it may never time out."
2319 PH/39 Allow G after quota size as well as K and M.
2321 PH/40 The value set for $authenticated_id in an authenticator may not contain
2322 binary zeroes or newlines because the value is written to log lines and
2323 to spool files. There was no check on this. Now the value is run through
2324 the string_printing() function so that such characters are converted to
2325 printable escape sequences.
2327 PH/41 $message_linecount is a new variable that contains the total number of
2328 lines in the message. Compare $body_linecount, which is the count for the
2331 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2332 bounce and delay warning messages, except in certain special cases, which
2335 (a) An SMTP error message from a remote host;
2336 (b) A message specified in a :fail: redirection;
2337 (c) A message specified in a "fail" command in a system filter;
2338 (d) A message specified in a FAIL return from the queryprogram router;
2339 (e) A message specified by the cannot_route_message router option.
2341 In these cases only, Exim does include the error details in bounce and
2342 warning messages. There are also a few cases where bland messages such
2343 as "unrouteable address" or "local delivery error" are given.
2345 PH/43 $value is now also set for the "else" part of a ${run expansion.
2347 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2348 being worked on, but at least Exim now implements the latest version to
2351 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2352 process to complete was treated as a delivery failure, a timeout while
2353 writing the message to the pipe was logged, but erroneously treated as a
2354 successful delivery. Such timeouts include transport filter timeouts. For
2355 consistency with the overall process timeout, these timeouts are now
2356 treated as errors, giving rise to delivery failures by default. However,
2357 there is now a new Boolean option for the pipe transport called
2358 timeout_defer, which, if set TRUE, converts the failures into defers for
2359 both kinds of timeout. A transport filter timeout is now identified in
2362 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2363 systems where "make" and "gmake" are different, calling "gmake" at top
2364 level broke things. I've arranged for the value of $(MAKE) to be passed
2365 from the Makefile to this script so that it can call the same version of
2369 A note about Exim versions 4.44 and 4.50
2370 ----------------------------------------
2372 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2373 changes of various kinds. As a consequence, a big documentation update was
2374 needed. This delayed the release for rather longer than seemed good, especially
2375 in the light of a couple of (minor) security issues. Therefore, the changes
2376 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2377 release. So 4.44 and 4.50 are in effect two different branches that both start
2380 I have left the 4.50 change log unchanged; it contains all the changes since
2381 4.43. The change log for 4.44 is below; many of its items are identical to
2382 those for 4.50. This seems to be the most sensible way to preserve the
2383 historical information.
2389 1. Minor wording change to the doc/README.SIEVE file.
2391 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2392 computation of the current number of files was incorrect.
2394 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2395 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2398 4. Give more explanation in the error message when the command for a transport
2399 filter fails to execute.
2401 5. There are several places where Exim runs a non-Exim command in a
2402 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2403 was being done only for the command run by the queryprogram router. It is
2404 now done for all such subprocesses. The other cases are: ${run, transport
2405 filters, and the commands run by the lmtp and pipe transports.
2407 6. Added CONFIGURE_GROUP build-time option.
2409 7. Some older OS have a limit of 256 on the maximum number of file
2410 descriptors. Exim was using setrlimit() to set 1000 as a large value
2411 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2412 systems. I've change it so that if it can't get 1000, it tries for 256.
2414 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2415 was an oversight, and furthermore, ever since the addition of extra
2416 controls (e.g. 4.43/32), the checks on when to allow different forms of
2417 "control" were broken. There should now be diagnostics for all cases when a
2418 control that does not make sense is encountered.
2420 9. Added the /retain_sender option to "control=submission".
2422 10. $recipients is now available in the predata ACL (oversight).
2424 11. Tidy the search cache before the fork to do a delivery from a message
2425 received from the command line. Otherwise the child will trigger a lookup
2426 failure and thereby defer the delivery if it tries to use (for example) a
2427 cached ldap connection that the parent has called unbind on.
2429 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2430 of $address_data from the recipient verification was clobbered by the
2431 sender verification.
2433 13. The value of address_data from a sender verification is now available in
2434 $sender_address_data in subsequent conditions in the ACL statement.
2436 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2438 15. Added a new option "connect=<time>" to callout options, to set a different
2441 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2442 was its contents. (It was OK if the option was not defined at all.)
2444 17. A "Completed" log line is now written for messages that are removed from
2445 the spool by the -Mrm option.
2447 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2448 information about exactly what failed.
2450 19. Added -dd to debug only the daemon process.
2452 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2453 handles timeouts, both on the server side and network timeouts. Renamed the
2454 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2456 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2457 It is now set to "smtps".
2459 22. $host_address is now set to the target address during the checking of
2460 ignore_target_hosts.
2462 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2463 being passed; this would have caused $sender_host_name to have been used if
2464 matching the list had actually called for a host name (not very likely,
2465 since this list is usually IP addresses). A host name is now passed as
2468 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2469 code that specifies a non-restarting handler (typically sigaction() in
2470 modern systems) in an attempt to fix a rare and obscure crash bug.
2472 25. Narrowed the window for a race in the daemon that could cause it to ignore
2473 SIGCHLD signals. This is not a major problem, because they are used only to
2474 wake it up if nothing else does.
2476 26. A malformed maildirsize file could cause Exim to calculate negative values
2477 for the mailbox size or file count. Odd effects could occur as a result.
2478 The maildirsize information is now recalculated if the size or filecount
2481 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2482 support for a long time. Removed HAVE_SYS_VFS_H.
2484 28. Installed the latest version of exipick from John Jetmore.
2486 29. In an address list, if the pattern was not a regular expression, an empty
2487 subject address (from a bounce message) matched only if the pattern was an
2488 empty string. Non-empty patterns were not even tested. This was the wrong
2489 because it is perfectly reasonable to use an empty address as part of a
2490 database query. An empty address is now tested by patterns that are
2491 lookups. However, all the other forms of pattern expect the subject to
2492 contain a local part and a domain, and therefore, for them, an empty
2493 address still always fails if the pattern is not itself empty.
2495 30. Exim went into a mad DNS loop when attempting to do a callout where the
2496 host was specified on an smtp transport, and looking it up yielded more
2497 than one IP address.
2499 31. Re-factored the code for checking spool and log partition space into a
2500 function that finds that data and another that does the check. The former
2501 is then used to implement four new variables: $spool_space, $log_space,
2502 $spool_inodes, and $log_inodes.
2504 32. The RFC2047 encoding function was originally intended for short strings
2505 such as real names; it was not keeping to the 75-character limit for
2506 encoded words that the RFC imposes. It now respects the limit, and
2507 generates multiple encoded words if necessary. To be on the safe side, I
2508 have increased the buffer size for the ${rfc2047: expansion operator from
2511 33. It is now permitted to omit both strings after an "if" condition; if the
2512 condition is true, the result is "true". As before, when the second string
2513 is omitted, a false condition yields an empty string. This makes it less
2514 cumbersome to write custom ACL and router conditions.
2516 34. Failure to deliver a bounce message always caused it to be frozen, even if
2517 there was an errors_to setting on the router. The errors_to setting is now
2520 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2521 canonical form (fully expanded) before being placed in
2522 $sender_host_address.
2524 36. The table in the code that translates DNS record types into text (T_A to
2525 "A" for instance) was missing entries for NS and CNAME. It is just possible
2526 that this could have caused confusion if both these types were looked up
2527 for the same domain, because the text type is used as part of Exim's
2528 per-process caching. But the chance of anyone hitting this buglet seems
2531 37. The dnsdb lookup has been extended in a number of ways.
2533 (1) There is a new type, "zns", which walks up the domain tree until it
2534 finds some nameserver records. It should be used with care.
2536 (2) There is a new type, "mxh", which is like "mx" except that it returns
2537 just the host names, not the priorities.
2539 (3) It is now possible to give a list of domains (or IP addresses) to be
2540 looked up. The behaviour when one of the lookups defers can be
2541 controlled by a keyword.
2543 (4) It is now possible to specify the separator character for use when
2544 multiple records are returned.
2546 38. The dnslists ACL condition has been extended: it is now possible to supply
2547 a list of IP addresses and/or domains to be looked up in a particular DNS
2550 39. Added log_selector=+queue_time_overall.
2552 40. When running the queue in the test harness, wait just a tad after forking a
2553 delivery process, to get repeatability of debugging output.
2555 41. Include certificate and key file names in error message when GnuTLS fails
2556 to set them up, because the GnuTLS error message doesn't include the name
2557 of the failing file when there is a problem reading it.
2559 42. Allow both -bf and -bF in the same test run.
2561 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2563 44. The "Exiscan patch" is now merged into the mainline Exim source.
2565 45. Sometimes the final signoff response after QUIT could fail to get
2566 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2567 < 0 before doing a fflush(). This bug looks as though it goes back to the
2568 introduction of TLS in release 3.20, but "sometimes" must have been rare
2569 because the tests only now provoked it.
2571 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2572 (this can affect the format of dates).
2574 47. exim_tidydb, when checking for the continued existence of a message for
2575 which it has found a message-specific retry record, was not finding
2576 messages that were in split spool directories. Consequently, it was
2577 deleting retry records that should have stayed in existence.
2579 48. Steve fixed some bugs in eximstats.
2581 49. The SPA authentication driver was not abandoning authentication and moving
2582 on to the next authenticator when an expansion was forced to fail,
2583 contradicting the general specification for all authenticators. Instead it
2584 was generating a temporary error. It now behaves as specified.
2586 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2587 (the order specifies the preference for clients). The order is now AES256,
2588 AES128, 3DES, ARCFOUR128.
2590 51. Small patch to Sieve code - explicitly set From: when generating an
2593 52. Exim crashed if a remote delivery caused a very long error message to be
2594 recorded - for instance if somebody sent an entire SpamAssassin report back
2595 as a large number of 550 error lines. This bug was coincidentally fixed by
2596 increasing the size of one of Exim's internal buffers (big_buffer) that
2597 happened as part of the Exiscan merge. However, to be on the safe side, I
2598 have made the code more robust (and fixed the comments that describe what
2601 53. Now that there can be additional text after "Completed" in log lines (if
2602 the queue_time_overall log selector is set), a one-byte patch to exigrep
2603 was needed to allow it to recognize "Completed" as not the last thing in
2606 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2607 patch that reportedly fixes this has been added. I am not expert enough to
2608 create a test for it. This is what the patch creator wrote:
2610 "I found a little strange behaviour of ldap code when working with
2611 Windows 2003 AD Domain, where users was placed in more than one
2612 Organization Units. When I tried to give exim partial DN, the exit code
2613 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2614 But simultaneously result of request was absolutely normal ldap result,
2615 so I produce this patch..."
2617 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2618 REFERENCE, so I have modified the code to exclude the patch when that macro
2621 55. Some experimental protocols are using DNS PTR records for new purposes. The
2622 keys for these records are domain names, not reversed IP addresses. The
2623 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2624 leaves it alone. Component reversal etc. now happens only for IP addresses.
2627 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2629 57. Double the size of the debug message buffer (to 2048) so that more of very
2630 long debug lines gets shown.
2632 58. The exicyclog utility now does better if the number of log files to keep
2633 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2635 59. Two changes related to the smtp_active_hostname option:
2637 (1) $smtp_active_hostname is now available as a variable.
2638 (2) The default for smtp_banner uses $smtp_active_hostname instead
2639 of $primary_hostname.
2641 60. The host_aton() function is supposed to be passed a string that is known
2642 to be a valid IP address. However, in the case of IPv6 addresses, it was
2643 not checking this. This is a hostage to fortune. Exim now panics and dies
2644 if the condition is not met. A case was found where this could be provoked
2645 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2646 components; fortuitously, this particular loophole had already been fixed
2647 by change 4.50/55 above.
2649 If there are any other similar loopholes, the new check in host_aton()
2650 itself should stop them being exploited. The report I received stated that
2651 data on the command line could provoke the exploit when Exim was running as
2652 exim, but did not say which command line option was involved. All I could
2653 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2654 running as the user.
2657 61. There was a buffer overflow vulnerability in the SPA authentication code
2658 (which came originally from the Samba project). I have added a test to the
2659 spa_base64_to_bits() function which I hope fixes it.
2662 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2663 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2665 63. The daemon start-up calls getloadavg() while still root for those OS that
2666 need the first call to be done as root, but it missed one case: when
2667 deliver_queue_load_max is set with deliver_drop_privilege. This is
2668 necessary for the benefit of the queue runner, because there is no re-exec
2669 when deliver_drop_privilege is set.
2671 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2672 This has been fixed.
2674 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2675 was in use, was not putting the data itself into the right store pool;
2676 consequently, it could be overwritten for a subsequent message in the same
2677 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2680 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2682 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2683 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2684 the function were treating the return as a boolean value, which happened to
2685 work because 0=false and not-0=true, but is not correct code.
2687 68. The host_aton() function was not handling scoped IPv6 addresses (those
2688 with, for example, "%eth0" on the end) correctly.
2690 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2691 negated items (that is, ~something) in unsigned ints. Some compilers
2692 apparently mutter when there is no cast.
2694 70. If an address verification called from an ACL failed, and did not produce a
2695 user-specific message (i.e. there was only a "system" message), nothing was
2696 put in $acl_verify_message. In this situation, it now puts the system
2699 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2700 session; change 4.31/43 added the unwanted input to the log line - except
2701 that it did not do this in the start of session case. It now does.
2703 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2704 This is wrong and can cause the other end to generate a synchronization
2705 error if it is another Exim or anything else that does the synchronization
2706 check. A QUIT command is no longer sent after a timeout.
2708 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2709 during host lookups.
2711 74. The defer_ok option of callout verification was not working if it was used
2712 when verifying addresses in header lines, that is, for this case:
2714 verify = header_sender/callout=defer_ok
2716 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2717 those file descriptors could be used for SMTP connections. If anything
2718 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2719 could be sent to the SMTP client, causing chaos. The daemon now opens
2720 stdin, stdout, and stderr to /dev/null when it puts itself into the
2723 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2724 log by default. The user can override this with suitable Perl magic.
2726 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2727 the log message when discard triggers, was not working for the DATA ACL or
2728 for the non-SMTP ACL.
2730 78. Error message wording change in sieve.c.
2732 79. If smtp_accept_max_per_host was set, the number of connections could be
2733 restricted to fewer than expected, because the daemon was trying to set up
2734 a new connection before checking whether the processes handling previous
2735 connections had finished. The check for completed processes is now done
2736 earlier. On busy systems, this bug wouldn't be noticed because something
2737 else would have woken the daemon, and it would have reaped the completed
2740 80. If a message was submitted locally by a user whose login name contained one
2741 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2742 It caused a spool format error. I have fixed the spool reading code. A
2743 related problem was that the "from" clause in the Received: line became
2744 illegal because of the space(s). It is now covered by ${quote_local_part.
2746 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2749 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2751 83. Patch from Sieve maintainer to fix unterminated string problem in
2752 "vacation" handling.
2754 84. Some minor changes to the Linux configuration files to help with other
2755 OS variants using glibc.
2757 85. One more patch for Sieve to update vacation handling to latest spec.
2760 ----------------------------------------------------
2761 See the note above about the 4.44 and 4.50 releases.
2762 ----------------------------------------------------
2768 1. Change 4.43/35 introduced a bug that caused file counts to be
2769 incorrectly computed when quota_filecount was set in an appendfile
2772 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2773 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2776 3. Give more explanation in the error message when the command for a transport
2777 filter fails to execute.
2779 4. There are several places where Exim runs a non-Exim command in a
2780 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2781 was being done only for the command run by the queryprogram router. It is
2782 now done for all such subprocesses. The other cases are: ${run, transport
2783 filters, and the commands run by the lmtp and pipe transports.
2785 5. Some older OS have a limit of 256 on the maximum number of file
2786 descriptors. Exim was using setrlimit() to set 1000 as a large value
2787 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2788 systems. I've change it so that if it can't get 1000, it tries for 256.
2790 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2791 was an oversight, and furthermore, ever since the addition of extra
2792 controls (e.g. 4.43/32), the checks on when to allow different forms of
2793 "control" were broken. There should now be diagnostics for all cases when a
2794 control that does not make sense is encountered.
2796 7. $recipients is now available in the predata ACL (oversight).
2798 8. Tidy the search cache before the fork to do a delivery from a message
2799 received from the command line. Otherwise the child will trigger a lookup
2800 failure and thereby defer the delivery if it tries to use (for example) a
2801 cached ldap connection that the parent has called unbind on.
2803 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2804 of $address_data from the recipient verification was clobbered by the
2805 sender verification.
2807 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2808 was its contents. (It was OK if the option was not defined at all.)
2810 11. A "Completed" log line is now written for messages that are removed from
2811 the spool by the -Mrm option.
2813 12. $host_address is now set to the target address during the checking of
2814 ignore_target_hosts.
2816 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2817 being passed; this would have caused $sender_host_name to have been used if
2818 matching the list had actually called for a host name (not very likely,
2819 since this list is usually IP addresses). A host name is now passed as
2822 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2823 code that specifies a non-restarting handler (typically sigaction() in
2824 modern systems) in an attempt to fix a rare and obscure crash bug.
2826 15. Narrowed the window for a race in the daemon that could cause it to ignore
2827 SIGCHLD signals. This is not a major problem, because they are used only to
2828 wake it up if nothing else does.
2830 16. A malformed maildirsize file could cause Exim to calculate negative values
2831 for the mailbox size or file count. Odd effects could occur as a result.
2832 The maildirsize information is now recalculated if the size or filecount
2835 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2836 support for a long time. Removed HAVE_SYS_VFS_H.
2838 18. Updated exipick to current release from John Jetmore.
2840 19. Allow an empty sender to be matched against a lookup in an address list.
2841 Previously the only cases considered were a regular expression, or an
2844 20. Exim went into a mad DNS lookup loop when doing a callout where the
2845 host was specified on the transport, if the DNS lookup yielded more than
2848 21. The RFC2047 encoding function was originally intended for short strings
2849 such as real names; it was not keeping to the 75-character limit for
2850 encoded words that the RFC imposes. It now respects the limit, and
2851 generates multiple encoded words if necessary. To be on the safe side, I
2852 have increased the buffer size for the ${rfc2047: expansion operator from
2855 22. Failure to deliver a bounce message always caused it to be frozen, even if
2856 there was an errors_to setting on the router. The errors_to setting is now
2859 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2860 canonical form (fully expanded) before being placed in
2861 $sender_host_address.
2863 24. Updated eximstats to version 1.33
2865 25. Include certificate and key file names in error message when GnuTLS fails
2866 to set them up, because the GnuTLS error message doesn't include the name
2867 of the failing file when there is a problem reading it.
2869 26. Expand error message when OpenSSL has problems setting up cert/key files.
2872 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2873 (this can affect the format of dates).
2875 28. exim_tidydb, when checking for the continued existence of a message for
2876 which it has found a message-specific retry record, was not finding
2877 messages that were in split spool directories. Consequently, it was
2878 deleting retry records that should have stayed in existence.
2880 29. eximstats updated to version 1.35
2881 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2882 1.35 - bugfix such that pie charts by volume are generated correctly
2884 30. The SPA authentication driver was not abandoning authentication and moving
2885 on to the next authenticator when an expansion was forced to fail,
2886 contradicting the general specification for all authenticators. Instead it
2887 was generating a temporary error. It now behaves as specified.
2889 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2890 (the order specifies the preference for clients). The order is now AES256,
2891 AES128, 3DES, ARCFOUR128.
2893 31. Small patch to Sieve code - explicitly set From: when generating an
2896 32. Exim crashed if a remote delivery caused a very long error message to be
2897 recorded - for instance if somebody sent an entire SpamAssassin report back
2898 as a large number of 550 error lines. This bug was coincidentally fixed by
2899 increasing the size of one of Exim's internal buffers (big_buffer) that
2900 happened as part of the Exiscan merge. However, to be on the safe side, I
2901 have made the code more robust (and fixed the comments that describe what
2904 33. Some experimental protocols are using DNS PTR records for new purposes. The
2905 keys for these records are domain names, not reversed IP addresses. The
2906 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2907 leaves it alone. Component reversal etc. now happens only for IP addresses.
2910 34. The host_aton() function is supposed to be passed a string that is known
2911 to be a valid IP address. However, in the case of IPv6 addresses, it was
2912 not checking this. This is a hostage to fortune. Exim now panics and dies
2913 if the condition is not met. A case was found where this could be provoked
2914 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2915 components; fortuitously, this particular loophole had already been fixed
2916 by change 4.50/55 or 4.44/33 above.
2918 If there are any other similar loopholes, the new check in host_aton()
2919 itself should stop them being exploited. The report I received stated that
2920 data on the command line could provoke the exploit when Exim was running as
2921 exim, but did not say which command line option was involved. All I could
2922 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2923 running as the user.
2926 35. There was a buffer overflow vulnerability in the SPA authentication code
2927 (which came originally from the Samba project). I have added a test to the
2928 spa_base64_to_bits() function which I hope fixes it.
2931 36. The daemon start-up calls getloadavg() while still root for those OS that
2932 need the first call to be done as root, but it missed one case: when
2933 deliver_queue_load_max is set with deliver_drop_privilege. This is
2934 necessary for the benefit of the queue runner, because there is no re-exec
2935 when deliver_drop_privilege is set.
2937 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2938 was in use, was not putting the data itself into the right store pool;
2939 consequently, it could be overwritten for a subsequent message in the same
2940 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2943 38. Sometimes the final signoff response after QUIT could fail to get
2944 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2945 < 0 before doing a fflush(). This bug looks as though it goes back to the
2946 introduction of TLS in release 3.20, but "sometimes" must have been rare
2947 because the tests only now provoked it.
2953 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2954 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2955 Now it returns an integer. A number of calls were still expecting a T/F
2956 return. Fortuitously, in all cases, the tests worked in OK situations,
2957 which is the norm. However, things would have gone wrong on any write
2958 failures on the smtp file descriptor. This function is used when sending
2959 messages over SMTP and also when doing verify callouts.
2961 2. When Exim is called to do synchronous delivery of a locally submitted
2962 message (the -odf or -odi options), it no longer closes stderr before doing
2965 3. Implemented the mua_wrapper option.
2967 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2969 5. Implemented the functions header_remove(), header_testname(),
2970 header_add_at_position(), and receive_remove_recipient(), and exported them
2973 6. If an ACL "warn" statement specified the addition of headers, Exim already
2974 inserted X-ACL-Warn: at the start if there was no header name. However, it
2975 was not making this test for the second and subsequent header lines if
2976 there were newlines in the string. This meant that an invalid header could
2977 be inserted if Exim was badly configured.
2979 7. Allow an ACL "warn" statement to add header lines at the start or after all
2980 the Received: headers, as well as at the end.
2982 8. Added the rcpt_4xx retry error code.
2984 9. Added postmaster_mailfrom=xxx to callout verification option.
2986 10. Added mailfrom=xxxx to the callout verification option, for verify=
2989 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2990 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2992 12. Inserted some casts to stop certain compilers warning when using pointer
2993 differences as field lengths or precisions in printf-type calls (mostly
2994 affecting debugging statements).
2996 13. Added optional readline() support for -be (dynamically loaded).
2998 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2999 same clock tick as a message's arrival, so that its received time was the
3000 same as the "first fail" time on the retry record, and that message
3001 remained on the queue past the ultimate address timeout, every queue runner
3002 would try a delivery (because it was past the ultimate address timeout) but
3003 after another failure, the ultimate address timeout, which should have then
3004 bounced the address, did not kick in. This was a "< instead of <=" error;
3005 in most cases the first failure would have been in the next clock tick
3006 after the received time, and all would be well.
3008 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3009 being recognized when the domain list was tested by the match_domain
3010 condition in an expansion string.
3012 16. Added the ${str2b64: operator.
3014 17. Exim was always calling setrlimit() to set a large limit for the number of
3015 processes, without checking whether the existing limit was already
3016 adequate. (It did check for the limit on file descriptors.) Furthermore,
3017 errors from getrlimit() and setrlimit() were being ignored. Now they are
3018 logged to the main and panic logs, but Exim does carry on, to try to do its
3019 job under whatever limits there are.
3021 18. Imported PCRE 5.0.
3023 19. Trivial typo in log message " temporarily refused connection" (the leading
3026 20. If the log selector return_path_on_delivery was set and an address was
3027 redirected to /dev/null, the delivery process crashed because it assumed
3028 that a return path would always be set for a "successful" delivery. In this
3029 case, the whole delivery is bypassed as an optimization, and therefore no
3032 21. Internal re-arrangement: the function for sending a challenge and reading
3033 a response while authentication was assuming a zero-terminated challenge
3034 string. It's now changed to take a pointer and a length, to allow for
3035 binary data in such strings.
3037 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3039 23. Exim was not respecting finduser_retries when seeking the login of the
3040 uid under which it was called; it was always trying 10 times. (The default
3041 setting of finduser_retries is zero.) Also, it was sleeping after the final
3042 failure, which is pointless.
3044 24. Implemented tls_on_connect_ports.
3046 25. Implemented acl_smtp_predata.
3048 26. If the domain in control=submission is set empty, Exim assumes that the
3049 authenticated id is a complete email address when it generates From: or
3050 Sender: header lines.
3052 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3053 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3054 chown and chgrp in /bin and hostname in /usr/bin.
3056 28. Exim was keeping the "process log" file open after each use, just as it
3057 does for the main log. This opens the possibility of it remaining open for
3058 long periods when the USR1 signal hits a daemon. Occasional processlog
3059 errors were reported, that could have been caused by this. Anyway, it seems
3060 much more sensible not to leave this file open at all, so that is what now
3063 29. The long-running daemon process does not normally write to the log once it
3064 has entered its main loop, and it closes the log before doing so. This is
3065 so that log files can straightforwardly be renamed and moved. However,
3066 there are a couple of unusual error situations where the daemon does write
3067 log entries, and I had neglected to close the log afterwards.
3069 30. The text of an SMTP error response that was received during a remote
3070 delivery was being truncated at 512 bytes. This is too short for some of
3071 the long messages that one sometimes sees. I've increased the limit to
3074 31. It is now possible to make retry rules that apply only when a message has a
3075 specific sender, in particular, an empty sender.
3077 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3078 it possible to be selective about when SMTP synchronization is enforced.
3080 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3082 32. Implemented hosts_connection_nolog.
3084 33. Added an ACL for QUIT.
3086 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3089 35. Added mailbox_size and mailbox_filecount to appendfile.
3091 36. Added control = no_multiline_responses to ACLs.
3093 37. There was a bug in the logic of the code that waits for the clock to tick
3094 in the case where the clock went backwards by a substantial amount such
3095 that the microsecond fraction of "now" was more than the microsecond
3096 fraction of "then" (but the whole seconds number was less).
3098 38. Added support for the libradius Radius client library this is found on
3099 FreeBSD (previously only the radiusclient library was supported).
3105 1. When certain lookups returned multiple values in the form name=value, the
3106 quoting of the values was not always being done properly. Specifically:
3107 (a) If the value started with a double quote, but contained no whitespace,
3109 (b) If the value contained whitespace other than a space character (i.e.
3110 tabs or newlines or carriage returns) it was not quoted.
3111 This fix has been applied to the mysql and pgsql lookups by writing a
3112 separate quoting function and calling it from the lookup code. The fix
3113 should probably also be applied to nisplus, ibase and oracle lookups, but
3114 since I cannot test any of those, I have not disturbed their existing code.
3116 2. A hit in the callout cache for a specific address caused a log line with no
3117 reason for rejecting RCPT. Now it says "Previous (cached) callout
3118 verification failure".
3120 3. There was an off-by-one bug in the queryprogram router. An over-long
3121 return line was truncated at 256 instead of 255 characters, thereby
3122 overflowing its buffer with the terminating zero. As well as fixing this, I
3123 have increased the buffer size to 1024 (and made a note to document this).
3125 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3126 when Exim is waiting for an SMTP response from a remote server, Exim
3127 restarts its select() call on the socket, thereby resetting its timeout.
3128 This is not a problem when such interrupts are rare. Somebody set up a cron
3129 job to run exiwhat every 2 minutes, which is less than the normal select()
3130 timeout (5 or 10 minutes). This meant that the select() timeout never
3131 kicked in because it was always reset. I have fixed this by comparing the
3132 time when an interrupt arrives with the time at the start of the first call
3133 to select(). If more time than the timeout has elapsed, the interrupt is
3134 treated as a timeout.
3136 5. Some internal re-factoring in preparation for the addition of Sieve
3137 extensions (by MH). In particular, the "personal" test is moved to a
3138 separate function, and given an option for scanning Cc: and Bcc: (which is
3139 not set for Exim filters).
3141 6. When Exim created an email address using the login of the caller as the
3142 local part (e.g. when creating a From: or Sender: header line), it was not
3143 quoting the local part when it contained special characters such as @.
3145 7. Installed new OpenBSD configuration files.
3147 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3148 try to make them clearer.
3150 9. Callout options, other than the timeout value, were being ignored when
3151 verifying sender addresses in header lines. For example, when using
3153 verify = header_sender/callout=no_cache
3155 the cache was (incorrectly) being used.
3157 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3158 only the Cygwin environment.
3160 11. When return_path_on_delivery was set as a log selector, if different remote
3161 addresses in the same message used different return paths and parallel
3162 remote delivery occurred, the wrong values would sometimes be logged.
3163 (Whenever a remote delivery process finished, the return path value from
3164 the most recently started remote delivery process was logged.)
3166 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3167 lines when AUTH and/or TLS are in use. This is the "received protocol"
3168 field. Exim used to use "asmtp" for authenticated SMTP, without any
3169 indication (in the protocol name) for TLS use. Now it follows the RFC and
3170 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3171 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3172 names appear in log lines as well as in Received: header lines.
3174 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3175 extensions, and comparison tests, and to fix some bugs.
3177 14. Changes to the "personal" filter test:
3179 (1) The test was buggy in that it was just doing the equivalent of
3180 "contains" tests on header lines. For example, if a user's address was
3181 anne@some.where, the "personal" test would incorrectly be true for
3183 To: susanne@some.where
3185 This test is now done by extracting each address from the header in turn,
3186 and checking the entire address. Other tests that are part of "personal"
3187 are now done using regular expressions (for example, to check local parts
3188 of addresses in From: header lines).
3190 (2) The list of non-personal local parts in From: addresses has been
3191 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3192 taken from the Sieve specification recommendations.
3194 (3) If the message contains any header line starting with "List-" it is
3195 treated as non-personal.
3197 (4) The test for "circular" in the Subject: header line has been removed
3198 because it now seems ill-conceived.
3200 15. Minor typos in src/EDITME comments corrected.
3202 16. Installed latest exipick from John Jetmore.
3204 17. If headers_add on a router specified a text string that was too long for
3205 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3206 of string_sprintf() is now avoided.
3208 18. $message_body_size was not set (it was always zero) when running the DATA
3209 ACL and the local_scan() function.
3211 19. For the "mail" command in an Exim filter, no default was being set for
3212 the once_repeat time, causing a random time value to be used if "once" was
3213 specified. (If the value happened to be <= 0, no repeat happened.) The
3214 default is now 0s, meaning "never repeat". The "vacation" command was OK
3215 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3216 (I found it when inspecting the code).
3218 20. There is now an overall timeout for performing a callout verification. It
3219 defaults to 4 times the callout timeout, which applies to individual SMTP
3220 commands during the callout. The overall timeout applies when there is more
3221 than one host that can be tried. The timeout is checked before trying the
3222 next host. This prevents very long delays if there are a large number of
3223 hosts and all are timing out (e.g. when the network connections are timing
3224 out). The value of the overall timeout can be changed by specifying an
3225 additional sub-option for "callout", called "maxwait". For example:
3227 verify = sender/callout=5s,maxwait=20s
3229 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3230 to the end before writing, but this should make it even safer).
3232 22. Exim was forgetting that it had advertised PIPELINING for the second and
3233 subsequent messages on an SMTP connection. It was also not resetting its
3234 memory on STARTTLS and an internal HELO.
3236 23. When Exim logs an SMTP synchronization error within a session, it now
3237 records whether PIPELINING has been advertised or not.
3239 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3240 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3241 rather than long int.
3243 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3245 26. Added the never_mail option to autoreply.
3251 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3252 crash if the getsockname() call failed; this can happen if a connection is
3253 closed very soon after it is established. The problem was simply in the
3254 order in which certain operations were done, causing Exim to try to write
3255 to the SMTP stream before it had set up the file descriptor. The bug has
3256 been fixed by making things happen in the correct order.
3262 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3263 before the connection was closed, thus losing the rejection response.
3265 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3266 some early Solaris releases, but causes trouble in current releases where
3267 socklen_t is defined.
3269 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3272 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3275 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3276 that releasing the top of it at the end releases what was used for sub-
3277 expansions (unless the block got too big). However, discard this block if
3278 the first thing is a variable or header, so that we can use its block when
3279 it is dynamic (useful for very large $message_headers, for example).
3281 6. Lookups now cache *every* query, not just the most recent. A new, separate
3282 store pool is used for this. It can be recovered when all lookup caches are
3283 flushed. Lookups now release memory at the end of their result strings.
3284 This has involved some general refactoring of the lookup sources.
3286 7. Some code has been added to the store_xxx() functions to reduce the amount
3287 of flapping under certain conditions.
3289 8. log_incoming_interface used to affect only the <= reception log lines. Now
3290 it causes the local interface and port to be added to several more SMTP log
3291 lines, for example "SMTP connection from", and rejection lines.
3293 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3295 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3297 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3298 could be overwritten at the end of the current message (or the start of a
3299 new message if it was set in a HELO ACL). The value is now preserved for
3300 the duration of the SMTP connection.
3302 12. If a transport had a headers_rewrite setting, and a matching header line
3303 contained an unqualified address, that address was qualified, even if it
3304 did not match any rewriting rules. The underlying bug was that the values
3305 of the flags that permit the existence of unqualified sender and recipient
3306 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3307 non-local messages, and by -bnq for local messages) were not being
3308 preserved with the message after it was received.
3310 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3311 "next input=" as part of the text comprising the host identity instead of
3312 the correct text. The code was using the same buffer for two different
3313 strings. However, depending on which order the printing function evaluated
3314 its arguments, the bug did not always show up. Under Linux, for example, my
3315 test suite worked just fine.
3317 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3318 doesn't work with some older versions of Perl. It has been changed to "my",
3319 which in any case is probably the better facility to use.
3321 15. A really picky compiler found some instances of statements for creating
3322 error messages that either had too many or two few arguments for the format
3325 16. The size of the buffer for calls to the DNS resolver has been increased
3326 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3327 for addresses that have a lot of PTR records. This alleviates a problem; it
3328 does not fully solve it.
3330 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3331 buffer now truncates the list and logs the incident, which is the same
3332 action as happens when Exim is looking up a host name and its aliases.
3333 Previously in this situation something unpredictable would happen;
3334 sometimes it was "internal error: store_reset failed".
3336 18. If a server dropped the connection unexpectedly when an Exim client was
3337 using GnuTLS and trying to read a response, the client delivery process
3338 crashed while trying to generate an error log message.
3340 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3341 string, for example:
3343 warn message = H1: something\nH2: something
3345 the text was added as a single header line from Exim's point of view
3346 though it ended up OK in the delivered message. However, searching for the
3347 second and subsequent header lines using $h_h2: did not work. This has been
3348 fixed. Similarly, if a system filter added multiple headers in this way,
3349 the routers could not see them.
3351 20. Expanded the error message when iplsearch is called with an invalid key to
3352 suggest using net-iplsearch in a host list.
3354 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3355 ACLs are no longer actually imposed (and a message to that effect is
3358 22. If a "gecos" field in a passwd entry contained escaped characters, in
3359 particular, if it contained a \" sequence, Exim got it wrong when building
3360 a From: or a Sender: header from that name. A second bug also caused
3361 incorrect handling when an unquoted " was present following a character
3362 that needed quoting.
3364 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3365 was not being matched caselessly.
3367 24. Arranged for all hyphens in the exim.8 source to be escaped with
3370 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3371 properly. Recipient callout cache records were still being keyed to include
3372 the sender, even when use_sender was set false. This led to far more
3373 callouts that were necessary. The sender is no longer included in the key
3374 when use_sender is false.
3376 26. Added "control = submission" modifier to ACLs.
3378 27. Added the ${base62d: operator to decode base 62 numbers.
3380 28. dnsdb lookups can now access SRV records.
3382 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3383 the configuration file.
3385 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3386 (-v) mode. This makes the output for a verbose queue run more intelligible.
3388 31. Added a use_postmaster feature to recipient callouts.
3390 32. Added the $body_zerocount variable, containing the number of binary zero
3391 bytes in the message body.
3393 33. The time of last modification of the "new" subdirectory is now used as the
3394 "mailbox time last read" when there is a quota error for a maildir
3397 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3399 35. Added +ignore_unknown as a special item in host lists.
3401 36. Code for decoding IPv6 addresses in host lists is now included, even if
3402 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3403 address was recognized as an IP address, but was then not correctly decoded
3404 into binary, causing unexpected and incorrect effects when compared with
3411 1. Very minor rewording of debugging text in manualroute to say "list of
3412 hosts" instead of "hostlist".
3414 2. If verify=header_syntax was set, and a header line with an unqualified
3415 address (no domain) and a large number of spaces between the end of the
3416 name and the colon was received, the reception process suffered a buffer
3417 overflow, and (when I tested it) crashed. This was caused by some obsolete
3418 code that should have been removed. The fix is to remove it!
3420 3. When running in the test harness, delay a bit after writing a bounce
3421 message to get a bit more predictability in the log output.
3423 4. Added a call to search_tidyup() just before forking a reception process. In
3424 theory, someone could use a lookup in the expansion of smtp_accept_max_
3425 per_host which, without the tidyup, could leave open a database connection.
3427 5. Added the variables $recipient_data and $sender_data which get set from a
3428 lookup success in an ACL "recipients" or "senders" condition, or a router
3429 "senders" option, similar to $domain_data and $local_part_data.
3431 6. Moved the writing of debug_print from before to after the "senders" test
3434 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3435 problems for message scanning, either using a data ACL, or using
3436 local_scan() because the Received: header was not generated till after they
3437 were called (in order to set the time as the time of reception completion).
3438 I have revised the way this works. The header is now generated after the
3439 body is received, but before the ACL or local_scan() are called. After they
3440 are run, the timestamp in the header is updated.
3446 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3447 before starting a queue runner without re-exec. This happened only when
3448 deliver_drop_privilege was set or when the Exim user was set to root. The
3449 effect of the bug was that timeouts during subsequent deliveries caused
3450 crashes instead of being properly handled. The handler is now left at its
3451 default (and expected) setting.
3453 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3454 message, again when deliver_drop_privilege is set or Exim is run as root.
3455 The bug described in (1) was not present in this case, but the tidying up
3456 of the other signals was missing. I have made the two cases consistent.
3458 3. The ignore_target_hosts setting on a manualroute router was being ignored
3459 for hosts that were looked up using the /MX notation.
3461 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3464 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3465 operated on the sender address. After changing the $sender_address to <>
3466 for the sender address verify, Exim was re-instated it as the original
3467 (before rewriting) address, but remembering that it had rewritten it, so it
3468 wasn't rewriting it again. This bug also had the effect of breaking the
3469 sender address verification caching when the sender address was rewritten.
3471 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3472 This has been changed so that if the ip literal address matches
3473 ignore_target_hosts, the router declines.
3475 7. Added expansion conditions match_domain, match_address, and match_local_
3476 part (NOT match_host).
3478 8. The placeholder for the Received: header didn't have a length field set.
3480 9. Added code to Exim itself and to exim_lock to test for a specific race
3481 condition that could lead to file corruption when using MBX delivery. The
3482 issue is with the lockfile that is created in /tmp. If this file is removed
3483 after a process has opened it but before that process has acquired a lock,
3484 there is the potential for a second process to recreate the file and also
3485 acquire a lock. This could lead to two Exim processes writing to the file
3486 at the same time. The added code performs the same test as UW imapd; it
3487 checks after acquiring the lock that its file descriptor still refers to
3488 the same named file.
3490 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3491 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3494 11. Added the smtp_active_hostname option. If used, this will typically be made
3495 to depend on the incoming interface address. Because $interface_address is
3496 not set up until the daemon has forked a reception process, error responses
3497 that can happen earlier (such as "too many connections") no longer contain
3500 12. If an expansion in a condition on a "warn" statement fails because a lookup
3501 defers, the "warn" statement is abandoned, and the next ACL statement is
3502 processed. Previously this caused the whole ACL to be aborted.
3504 13. Added the iplsearch lookup type.
3506 14. Added ident_timeout as a log selector.
3508 15. Added tls_certificate_verified as a log selector.
3510 16. Added a global option tls_require_ciphers (compare the smtp transport
3511 option of the same name). This controls incoming TLS connections.
3513 17. I finally figured out how to make tls_require_ciphers do a similar thing
3514 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3515 before starting the TLS session.
3517 18. Tabs are now shown as \t in -bP output.
3519 19. If the log selector return_path_on_delivery was set, Exim crashed when
3520 bouncing a message because it had too many Received: header lines.
3522 20. If two routers both had headers_remove settings, and the first one included
3523 a superfluous trailing colon, the final name in the first list and the
3524 first name in the second list were incorrectly joined into one item (with a
3525 colon in the middle).
3531 1. Added -C and -D options to the exinext utility, mainly to make it easier
3532 to include in the automated testing, but these could be helpful when
3533 multiple configurations are in use.
3535 2. The exinext utility was not formatting the output nicely when there was
3536 an alternate port involved in the retry record key, nor when there was a
3537 message id as well (for retries that were specific to a specific message
3538 and a specific host). It was also confused by IPv6 addresses, because of
3539 the additional colons they contain. I have fixed the IPv4 problem, and
3540 patched it up to do a reasonable job for IPv6.
3542 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3543 delivery, the log line now contains "pipelined" if PIPELINING was used.
3545 4. An SMTP transport process used to panic and die if the bind() call to set
3546 an explicit outgoing interface failed. This has been changed; it is now
3547 treated in the same way as a connect() failure.
3549 5. A reference to $sender_host_name in the part of a conditional expansion
3550 that was being skipped was still causing a DNS lookup. This no longer
3553 6. The def: expansion condition was not recognizing references to header lines
3554 that used bh_ and bheader_.
3556 7. Added the _cache feature to named lists.
3558 8. The code for checking quota_filecount in the appendfile transport was
3559 allowing one more file than it should have been.
3561 9. For compatibility with Sendmail, the command line option
3569 and sets the incoming protocol and host name (for trusted callers). The
3570 host name and its colon can be omitted when only the protocol is to be set.
3571 Note the Exim already has two private options, -pd and -ps, that refer to
3572 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3573 "s", but I don't think that's a major issue.
3575 10. A number of refactoring changes to the code, none of which should affect
3578 (a) The number of logging options was getting close to filling up the
3579 32-bit word that was used as a bit map. I have split them into two classes:
3580 those that are passed in the argument to log_write(), and those that are
3581 only ever tested independently outside of that function. These are now in
3582 separate 32-bit words, so there is plenty of room for expansion again.
3583 There is no change in the user interface or the logging behaviour.
3585 (b) When building, for example, log lines, the code previously used a
3586 macro that called string_cat() twice, in order to add two strings. This is
3587 not really sufficiently general. Furthermore, there was one instance where
3588 it was actually wrong because one of the argument was used twice, and in
3589 one call a function was used. (As it happened, calling the function twice
3590 did not affect the overall behaviour.) The macro has been replaced by a
3591 function that can join an arbitrary number of extra strings onto a growing
3594 (c) The code for expansion conditions now uses a table and a binary chop
3595 instead of a serial search (which was left over from when there were very
3596 few conditions). Also, it now recognizes conditions like "pam" even when
3597 the relevant support is not compiled in: a suitably worded error message is
3598 given if an attempt is made to use such a condition.
3600 11. Added ${time_interval:xxxxx}.
3602 12. A bug was causing one of the ddress fields not to be passed back correctly
3603 from remote delivery subprocesses. The field in question was not being
3604 subsequently used, so this caused to problems in practice.
3606 13. Added new log selectors queue_time and deliver_time.
3608 14. Might have fixed a bug in maildirsizefile handling that threw up
3609 "unexpected character" debug warnings, and recalculated the data
3610 unnecessarily. In any case, I expanded the warning message to give more
3613 15. Added the message "Restricted characters in address" to the statements in
3614 the default ACL that block characters like @ and % in local parts.
3616 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3617 Three changes have been made:
3619 (a) There was a serious bug; a negative response to MAIL caused the whole
3620 recipient domain to be cached as invalid, thereby blocking all messages
3621 to all local parts at the same domain, from all senders. This bug has
3622 been fixed. The domain is no longer cached after a negative response to
3623 MAIL if the sender used is not empty.
3625 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3628 (c) A new callout option, "use_sender" has been added for people who want
3629 the modified behaviour.
3635 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3638 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3639 indeed breaks things for older releases.
3641 3. Added additional logging to the case where there is a problem reading data
3642 from a filter that is running in a subprocess using a pipe, in order to
3643 try to track down a specific problem.
3645 4. Testing facility fudge: when running in the test harness and attempting
3646 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3647 getting "No route to host". Convert this to a timeout.
3649 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3652 6. Some OS don't have socklen_t but use size_t instead. This affects the
3653 fifth argument of getsockopt() amongst other things. This is now
3654 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3655 can be set for individual OS. I have set it for SunOS5, OSF1, and
3656 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3657 some earlier ones do not.
3659 7. Change 4.30/15 was not doing the test caselessly.
3661 8. The standard form for an IPv6 address literal was being rejected by address
3662 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3663 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3664 this, as well as the form without the "IPv6" on the front (but only when
3665 address literals are enabled, of course).
3667 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3669 10. Exim crashed if a message with an empty sender address specified by -f
3670 encountered a router with an errors_to setting. This could be provoked only
3671 by a command such as
3675 where an empty string was supplied; "<>" did not hit this bug.
3677 11. Installed PCRE release 4.5.
3679 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3680 remained set. It is now erased.
3682 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3683 times from message ids (which are base 36 rather than the normal 62).
3685 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3686 were being counted as actual protocol errors, and logged if the log
3687 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3688 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3689 and DATA following a set of rejected RCPTs do not count as protocol errors.
3690 In other words, Exim assumes they were pipelined, though this may not
3691 actually be the case. Of course, in all cases the client gets an
3692 appropriate error code.
3694 15. If a lookup fails in an ACL condition, a message about the failure may
3695 be available; it is used if testing the ACL cannot continue, because most
3696 such messages specify what the cause of the deferral is. However, some
3697 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3698 that caused an old message to be retained and used if a later statement
3699 caused a defer, replacing the real cause of the deferral.
3701 16. If an IP address had so many PTR records that the DNS lookup buffer
3702 was not large enough to hold them, Exim could crash while trying to process
3703 the truncated data. It now detects and logs this case.
3705 17. Further to 4.21/58, another change has been made: if (and only if) the
3706 first line of a message (the first header line) ends with CRLF, a bare LF
3707 in a subsequent header line has a space inserted after it, so as not to
3708 terminate the header.
3710 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3711 unnecessarily, used atoi() instead of strtol(), and didn't check the
3712 termination when getting file sizes from file names by regex.
3714 19. Completely re-implemented the support for maildirsize files, in the light
3715 of a number of problems with the previous contributed implementation
3716 (4.30/29). In particular:
3718 . If the quota is zero, the maildirsize file is maintained, but no quota is
3721 . If the maildir directory does not exist, it is created before any attempt
3722 to write a maildirsize file.
3724 . The quota value in the file is just a cache; if the quota is changed in
3725 the transport, the new value overrides.
3727 . A regular expression is available for excluding directories from the
3730 20. The autoreply transport checks the characters in options that define the
3731 message's headers; it allows continued headers, but it was checking with
3732 isspace() after an embedded newline instead of explicitly looking for a
3735 21. If all the "regular" hosts to which an address was routed had passed their
3736 expiry times, and had not reached their retry times, the address was
3737 bounced, even if fallback hosts were defined. Now Exim should go on to try
3740 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3741 equivalent code in the SMTP transport. Some hosts send humungous responses
3742 to HELO/EHLO, more than 1024 it seems.
3744 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3745 gives compiler warnings in some environments. I've now done it "properly",
3748 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3749 (because of problems with the built-in one) was declared to return uschar *
3750 instead of char *, causing compiler failure.
3752 25. Fixed a file descriptor leak when processing alias/forward files.
3754 26. Fixed a minor format string issue in dbfn.c.
3756 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3758 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3759 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3762 29. When a sender address is verified, it is cached, to save repeating the test
3763 when there is more than one recipient in a message. However, when the
3764 verification involves a callout, it is possible for different callout
3765 options to be set for different recipients. It is too complicated to keep
3766 track of this in the cache, so now Exim always runs a verification when a
3767 callout is required, relying on the callout cache for the optimization.
3768 The overhead is duplication of the address routing, but this should not be
3771 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3772 to be verified with callout=postmaster, and the main callout worked but the
3773 postmaster check failed, the verification correctly failed. However, if a
3774 subsequent RCPT command asked for sender verification *without* the
3775 postmaster check, incorrect caching caused this verification also to fail,
3778 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3779 it was not caching the DNS options (qualify_single, search_parents) that
3780 were used when the lookup failed. A subsequent lookup with different
3781 options therefore always gave the same answer, though there were cases
3782 where it should not have. (Example: a "domains = !$mx_any" option on a
3783 dnslookup router: the "domains" option is always processed without any
3784 widening, but the router might have qualify_single set.) Now Exim uses the
3785 cached value only when the same options are set.
3787 32. Added John Jetmore's "exipick" utility to the distribution.
3789 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3790 than a timeout (e.g. a certificate is required, and is not provided), an
3791 Exim server now closes the connection immediately. Previously it waited for
3792 the client to close - but if the client is SSL, it seems that they each
3793 wait for each other, leading to a delay before one of them times out.
3795 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3796 maintained 0.8.x compatibility because I don't think many are using it, and
3797 it is clearly obsolete.
3799 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3802 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3803 client certificate was expired. A simple patch fixes this, though I don't
3804 understand the full logic of why the verify callback is called multiple
3807 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3808 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3809 which causes problems with some clients (such as the Certicom SSL Plus
3810 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3811 disables the coutermeasure allowing Eudora to connect."
3813 38. Exim was not checking that a write() to a log file succeeded. This could
3814 lead to Bad Things if a log got too big, in particular if it hit a file
3815 size limit. Exim now panics and dies if it cannot write to a log file, just
3816 as it does if it cannot open a log file.
3818 39. Modified OS/Makefile-Linux so that it now contains
3820 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3822 The two -D definitions ensure that Exim is compiled with large file
3823 support, which makes it possible to handle log files that are bigger than
3826 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3827 instance) a domain was checked against a named list that involved a lookup,
3828 causing $domain_data to be set, then another domain was checked against the
3829 same list, then the first domain was re-checked, the value of $domain_data
3830 after the final check could be wrong. In particular, if the second check
3831 failed, it could be set empty. This bug probably also applied to
3834 41. The strip_trailing_dot option was not being applied to the address given
3835 with the -f command-line option.
3837 42. The code for reading a message's header from the spool was incrementing
3838 $received_count, but never initializing it. This meant that the value was
3839 incorrect (doubled) while delivering a message in the same process in which
3840 it was received. In the most common configuration of Exim, this never
3841 happens - a fresh exec is done - but it can happen when
3842 deliver_drop_privilege is set.
3844 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3845 it now includes up to 150 characters of the unexpected data in the log
3848 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3849 and building data strings. The size of both of these buffers was 10 000
3850 bytes - far larger than anybody would *ever* want, thought I. Needless to
3851 say, somebody hit the limit. I have increased the maximum line length to
3852 20 000 and the maximum data length of concatenated lines to 100 000. I have
3853 also fixed two bugs, because there was no checking on these buffers. Tsk,
3854 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3855 buffer is too small.
3857 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3858 lsearch lookups. Now it does.
3860 46. When parsing a route_list item in a manualroute router, a fixed-length
3861 buffer was used for the list of hosts. I made this 1024 bytes long,
3862 thinking that nobody would ever have a list of hosts that long. Wrong.
3863 Somebody had a whole pile of complicated expansion conditions, and the
3864 string was silently truncated, leading to an expansion error. It turns out
3865 that it is easier to change to an unlimited length (owing to other changes
3866 that have happened since this code was originally written) than to build
3867 structure for giving a limitation error. The length of the item that
3868 expands into the list of hosts is now unlimited.
3870 47. The lsearch lookup could not handle data where the length of text line was
3871 more than 4095 characters. Such lines were truncated, leading to shortened
3872 data being returned. It should now handle lines of any length.
3874 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3875 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3878 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3879 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3880 no longer happens in comment lines. A list of replacements is now placed
3881 at the head of all of the source files, except those whose only change is
3882 to replace PERL_COMMAND in the very first #! line.
3884 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3885 messages on the queue, with a bottom-up merge sort, using code contributed
3886 by Michael Haardt. This should make operations like -bp somewhat faster on
3887 large queues. It won't affect queue runners, except when queue_run_in_order
3890 51. Installed eximstats 1.31 in the distribution.
3892 52. Added support for SRV lookups to the dnslookup router.
3894 53. If an ACL referred to $message_body or $message_body_end, the value was not
3895 reset for any messages that followed in the same SMTP session.
3897 54. The store-handling optimization for building very long strings was not
3898 differentiating between the different store pools. I don't think this
3899 actually made any difference in practice, but I've tidied it.
3901 55. While running the routers to verify a sender address, $sender_address
3902 was still set to the sender address. This is wrong, because when routing to
3903 send a bounce to the sender, it would be empty. Therefore, I have changed
3904 it so that, while verifying a sender address, $sender_address is set to <>.
3905 (There is no change to what happens when verifying a recipient address.)
3907 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3908 target A or AAAA records (if not already returned) without resetting the
3909 qualify_single or search_parents options of the DNS resolver. These are
3910 inappropriate in this case because the targets of MX and SRV records must
3911 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3912 target that, when qualified, matched something in the local domain. These
3913 two options are now turned off when doing these lookups.
3915 57. It seems that at least some releases of Reiserfs (which does not have the
3916 concept of a fixed number of inodes) returns zero and not -1 for the
3917 number of available inodes. This interacted badly with check_spool_inodes,
3918 which assumed that -1 was the "no such thing" setting. What I have done is
3919 to check that the total number of inodes is greater than zero before doing
3920 the test of how many are available.
3922 58. When a "warn" ACL statement has a log_message modifier, the message is
3923 remembered, and not repeated. This is to avoid a lot of repetition when a
3924 message has many recipients that cause the same warning to be written.
3925 Howewer, Exim was preserving the list of already written lines for an
3926 entire SMTP session, which doesn't seem right. The memory is now reset if a
3927 new message is started.
3929 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3930 debugging output unless log_rewrite was also set.
3932 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3933 of (int)(handle) when we know that handle contains (void *)(-1).
3935 61. The Exim daemon panic-logs an error return when it closes the incoming
3936 connection. However "connection reset by peer" seems to be common, and
3937 isn't really an error worthy of noting specially, so that particular error
3940 62. When Exim is trying to find all the local interfaces, it used to panic and
3941 die if the ioctl to get the interface flags failed. However, it seems that
3942 on at least one OS (Solaris 9) it is possible to have an interface that is
3943 included in the list of interfaces, but for which you get a failure error
3944 for this call. This happens when the interface is not "plumbed" into a
3945 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3946 failure of the "get flags" call assumes that the interface is down.
3948 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3949 makes life easier for people who are doing arithmetic on fields extracted
3950 from dates, where you often get leading zeros that should not be
3951 interpreted as octal.
3953 64. Added qualify_domain to the redirect router, to override the global
3956 65. If a pathologically long header line contained very many addresses (the
3957 report of this problem mentioned 10 000) and each of them was rewritten,
3958 Exim could use up a very large amount of memory. (It kept on making new
3959 copies of the header line as it rewrote, and never released the old ones.)
3960 At the expense of a bit more processing, the header rewriting function has
3961 been changed so that it no longer eats memory in this way.
3963 66. The generation of the Received: header has been moved from the time that a
3964 message starts to be received, to the time that it finishes. The timestamp
3965 in the Received: header should now be very close to that of the <= log
3966 line. There are two side-effects of this change:
3968 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3969 logged header lines no longer include the local Received: line, because
3970 it has not yet been created. The same applies to a copy of the message
3971 that is returned to a non-SMTP sender when a message is rejected.
3973 (b) When a filter file is tested using -bf, no additional Received: header
3974 is added to the test message. After some thought, I decided that this
3977 This change does not affect the value of $received_for. It is still set
3978 after address rewriting, but before local_scan() is called.
3980 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3982 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3983 gave an unhelpful panic error message, and a defer error. I have managed to
3984 change this behaviour so that it now rejects any supplied certificate,
3985 which seems right, as the list of acceptable certificates is empty.
3987 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3988 gave an unhelpful defer error. I have not managed to make this reject any
3989 supplied certificates, but the error message it gives is "no certificate
3990 supplied", which is not helpful.
3992 70. exigrep's output now also includes lines that are not associated with any
3993 message, but which match the given pattern. Implemented by a patch from
3994 Martin Sluka, which also tidied up the Perl a bit.
3996 71. Recipient callout verification, like sender verification, was using <> in
3997 the MAIL FROM command. This isn't really the right thing, since the actual
3998 sender may affect whether the remote host accepts the recipient or not. I
3999 have changed it to use the actual sender in the callout; this means that
4000 the cache record is now keyed on a recipient/sender pair, not just the
4001 recipient address. There doesn't seem to be a real danger of callout loops,
4002 since a callout by the remote host to check the sender would use <>.
4003 [SEE ABOVE: changed after hitting problems.]
4005 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4006 temporary errors. However, in the case of such a code being given after
4007 the end of a data transmission (i.e. after ".") Exim was failing to write
4008 a retry record for the message. (Yes, there was some broken host that was
4009 actually sending 8xx at this point.)
4011 73. An unknown lookup type in a host list could cause Exim to panic-die when
4012 the list was checked. (An example that provoked this was putting <; in the
4013 middle of a list instead of at the start.) If this happened during a DATA
4014 ACL check, a -D file could be left lying around. This kind of configuration
4015 error no longer causes Exim to die; instead it causes a defer errror. The
4016 incident is still logged to the main and panic logs.
4018 74. Buglet left over from Exim 3 conversion. The message "too many messages
4019 in one connection" was written to the rejectlog but not the mainlog, except
4020 when address rewriting (yes!) was being logged.
4022 75. Added write_rejectlog option.
4024 76. When a system filter was run not as root (that is, when system_filter_user
4025 was set), the values of the $n variables were not being returned to the
4026 main process; thus, they were not subsequently available in the $sn
4029 77. Added +return_path_on_delivery log selector.
4031 78. A connection timeout was being treated differently from recipients deferred
4032 when testing hosts_max_try with a message that was older than the host's
4033 retry timeout. (The host should not be counted, thus allowing all hosts to
4034 be tried at least once before bouncing.) This may have been the cause of an
4035 occasionally reported bug whereby a message would remain on the queue
4036 longer than the retry timeout, but would be bounced if a delivery was
4037 forced. I say "may" because I never totally pinned down the problem;
4038 setting up timeout/retry tests is difficult. See also the next item.
4040 79. The ultimate address timeout was not being applied to errors that involved
4041 a combination of host plus message (for example, a timeout on a MAIL
4042 command). When an address resolved to a number of possible hosts, and they
4043 were not all tried for each delivery (e.g. because of hosts_max_try), a
4044 message could remain on the queue longer than the retry timeout.
4046 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4049 81. Fixed an obscure SMTP outgoing bug which required at least the following
4050 conditions: (a) there was another message waiting for the same server;
4051 (b) the server returned 5xx to all RCPT commands in the first message so
4052 that the message was not completed; (c) the server dropped the connection
4053 or gave a negative response to the RSET that Exim sends to abort the
4054 transaction. The observed case was a dropped connection after DATA that had
4055 been sent in pipelining mode. That is, the server had advertised PIPELINING
4056 but was not implementing it correctly. The effect of the bug was incorrect
4057 behaviour, such as trying another host, and this could lead to a crash.
4063 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4064 and daemon.c were passed as pointers to ints; they should have been
4065 pointers to socklen_t variables (which are typically unsigned ints).
4067 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4070 3. Fixed a really odd bug that affected only the testing scheme; patching a
4071 certain fixed string in the binary changed the value of another string that
4072 happened to be identical to the end of the original first string.
4074 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4075 name", it returns that address as the IP address. On some operating
4076 systems (e.g. Solaris), it also passes back the IP address string as the
4077 "host name". However, on others (e.g. Linux), it passes back an empty
4078 string. Exim wasn't checking for this, and was changing the host name to an
4079 empty string, assuming it had been canonicized.
4081 5. Although rare, it is permitted to have more than one PTR record for a given
4082 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4083 all the names associated with an address, because they do in Solaris.
4084 However, it seems that they do not in Linux for data that comes from the
4085 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4086 I found this out when I moved to a new Linux workstation and tried to run
4087 the Exim test suite.
4089 To get round this problem I have changed the code so that it now does its
4090 own call to the DNS to look up PTR records when searching for a host name.
4091 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4092 addresses that are only in /etc/hosts are still found.
4094 This behaviour is, however, controlled by an option called host_lookup_
4095 order, which defaults to "bydns:byaddr". If people want to use the other
4096 order, or indeed, just use one or the other means of lookup, they can
4097 specify it in this variable.
4099 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4100 some operating systems, this comes back from gethostbyaddr() as an empty
4101 string, and this is what Exim used to test for. However, it seems that in
4102 other systems, "." is yielded. Exim now tests for this case too.
4104 7. The values of check_spool_space and check_log_space are now held internally
4105 as a number of kilobytes instead of an absolute number of bytes. If a
4106 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4107 kilobyte. This means that much larger values can be stored.
4109 8. Exim monitor: an attempt to get the action menu when not actually pointing
4110 at a message produces an empty menu entitled "No message selected". This
4111 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4112 no entries in it ("Shell widget menu has zero width and/or height"). So I
4113 have added a single, blank menu entry in this case.
4115 9. Added ${quote_local_part.
4117 10. MIME decoding is now applied to the contents of Subject: header lines when
4120 11. Now that a reference to $sender_host_address automatically causes a reverse
4121 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4122 host lookup before query-style lookups in lists that might use this
4123 variable. This has therefore been abolished, and the "net-" prefix is no
4124 longer necessary for query-style lookups.
4126 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4127 have been a typo for LFLAGS, so it has been changed.
4129 13. The install script calls Exim with "-C /dev/null" in order to find the
4130 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4131 to be output. Howeve, since Exim outputs its version number before the
4132 error, it didn't break the script. It just looked ugly. I fixed this by
4133 always allowing "-C /dev/null" if the caller is root.
4135 14. Ignore overlarge ACL variable number when reading spool file - insurance
4136 against a later release with more variables having written the file.
4138 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4139 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4140 this, as well as the form without the "IPv6" on the front.
4142 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4143 OS/Makefile-Darwin file.
4145 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4146 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4148 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4149 variable. However, some OS use ints for inodes. I've added cast to long int
4150 to get rid of the compiler warning.
4152 19. I had forgotten to lock out "/../" in configuration file names when
4153 ALT_CONFIG_PREFIX was set.
4155 20. Routers used for verification do not need to specify transports. However,
4156 if such a router generated a host list, and callout was configured, Exim
4157 crashed, because it could not find a port number from the (non-existent)
4158 transport. It now assumes port 25 in this circumstance.
4160 21. Added the -t option to exigrep.
4162 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4163 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4164 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4165 them is set, LOOKUP_LSEARCH is forced.
4167 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4169 24. Added sender and host information to the "rejected by local_scan()" log
4170 line; previously there was no indication of these.
4172 25. Added .include_if_exists.
4174 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4175 Linux. It turns out that not all file systems support this. Apparently some
4176 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4177 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4178 not supported on the file descriptor, is now ignored when Exim is trying to
4179 sync a directory. This applies only to Linux.
4181 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4183 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4184 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4187 29. Added support for maildirsize files from supplied patch (modified a bit).
4189 30. The use of :fail: followed by an empty string could lead Exim to respond to
4190 sender verification failures with (e.g.):
4192 550 Verification failed for <xxx>
4193 550 Sender verify failed
4195 where the first response line was missing the '-' that indicates it is not
4196 the final line of the response.
4198 31. The loop for finding the name of the user that called Exim had a hardwired
4199 limit of 10; it now uses the value of finduser_retries, which is used for
4200 all other user lookups.
4202 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4205 33. Exim was neglecting to zero errno before one call of strtol() when
4206 expanding a string and expecting an integer value. On some systems this
4207 resulted in spurious "integer overflow" errors. Also, it was casting the
4208 result into an int without checking.
4210 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4211 did not work. The code looks as if it has *never* worked, though it appears
4212 to have been documented since at least releast 1.62. I have made it work.
4214 35. The "timeout_DNS" error in retry rules, also documented since at least
4215 1.62, also never worked. As it isn't clear exactly what this means, and
4216 clearly it isn't a major issue, I have abolished the feature by treating it
4217 as "timeout", and writing a warning to the main and panic logs.
4219 36. The display of retry rules for -brt wasn't always showing the error code
4222 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4223 timeout_connect_A, timeout_connect_MX.
4225 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4226 to the empty sender.
4228 39. The daemon was not analysing the content of -oX till after it had closed
4229 stderr and disconnected from the controlling terminal. This meant that any
4230 syntax errors were only noted on the panic log, and the return code from
4231 the command was 0. By re-arranging the code a little, I've made the
4232 decoding happen first, so such errors now appear on stderr, and the return
4233 code is 1. However, the actual setting up of the sockets still happens in
4234 the disconnected process, so errors there are still only recorded on the
4237 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4238 connections (as happens on some IP stacks) was logged at start up time as
4239 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4240 it from "IPv6 and IPv4", which means that two separate sockets are being
4243 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4244 says whether AF_INET or AF_INET6 was passed as an argument.
4246 42. Exiwhat output was messed up when time zones were included in log
4249 43. Exiwhat now gives more information about the daemon's listening ports,
4250 and whether -tls-on-connect was used.
4252 44. The "port" option of the smtp transport is now expanded.
4254 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4255 silently ignored. Now an error message is written to the main and panic
4258 46. There's a new ACL modifier called "logwrite" which writes to a log file
4259 as soon as it is encountered.
4261 47. Added $local_user_uid and $local_user_gid at routing time.
4263 48. Exim crashed when trying to verify a sender address that was being
4266 49. Exim was recognizing only a space character after ".include". It now also
4267 recognizes a tab character.
4269 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4270 extracting the relevant information from the specification. The man page no
4271 longer contains scrambled data for the -d option, and I've added a section
4272 at the front about calling Exim under different names.
4274 51. Added "extra_headers" argument to the "mail" command in filter files.
4276 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4279 53. Installed eximstats 1.29.
4281 54. Added transport_filter_timeout as a generic transport option.
4283 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4284 Cc: header lines. This was required by RFC 822, but it not required by RFC
4287 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4288 incoming messages that did not have them. Now it does so only if the
4289 message originates locally, that is, if there is no associated remote host
4290 address. When Resent- header lines are present, this applies to the Resent-
4291 lines rather than the non-Resent- lines.
4293 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4294 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4296 58. Messages for configuration errors now include the name of the main
4297 configuration file - useful now that there may be more than one file in a
4298 list (.included file names were always shown).
4300 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4301 for those rare installations that do not start the daemon as root or run it
4302 setuid root. I've cut out the call to initgroups() if the daemon is not
4305 60. The Exim user and group can now be bound into the binary as text strings
4306 that are looked up at the start of Exim's processing.
4308 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4310 62. Added $mailstore_basename variable.
4312 63. Installed patch to sieve.c from Michael Haardt.
4314 64. When Exim failed to open the panic log after failing to open the main log,
4315 the original message it was trying to log was written to stderr and debug
4316 output, but if they were not available (the usual case in production), it
4317 was lost. Now it is written to syslog before the two lines that record the
4318 failures to open the logs.
4320 65. Users' Exim filters run in subprocesses under the user's uid. It is
4321 possible for a "deliver" command or an alias in a "personal" command to
4322 provoke an address rewrite. If logging of address rewriting is configured,
4323 this fails because the process is not running as root or exim. There may be
4324 a better way of dealing with this, but for the moment (because 4.30 needs
4325 to be released), I have disabled address rewrite logging when running a
4326 filter in a non-root, non-exim process.
4332 1. The buildconfig auxiliary program wasn't quoting the value set for
4333 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4334 not defined. This bug was masked in 4.22 by the effect that was fixed in
4337 2. Some messages that were rejected after a message id was allocated were
4338 shown as "incomplete" by exigrep. It no longer does this for messages that
4339 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4341 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4342 have allow_domain_literals set, the ID did not get logged in the <= line.
4343 Domain literals are now always recognized in Message-ID: header lines.
4345 4. The first argument for a ${extract expansion item is the key name or field
4346 number. Leading and trailing spaces in this item were not being ignored,
4347 causing some misleading effects.
4349 5. When deliver_drop_privilege was set, single queue runner processes started
4350 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4351 same command in the process it spins off) were not dropping privilege.
4353 6. When the daemon running as "exim" started a queue runner, it always
4354 re-executed Exim in the spun-off process. This is a waste of effort when
4355 deliver_drop_privilege is set. The new process now just calls the
4356 queue-runner function directly.
4362 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4365 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4366 ignored. Though the use of -oP was forcing the writing of a pid file, it
4367 was always written to the default place.
4369 3. If the message "no IP address found for host xxxx" is generated during
4370 incoming verification, it is now followed by identification of the incoming
4371 connection (so you can more easily find what provoked it).
4373 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4375 5. Added some features to "harden" Exim a bit more against certain attacks:
4377 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4378 be put in Local/Makefile. This is like the never_users runtime option,
4379 but it cannot be overridden. The default setting is "root".
4381 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4382 prefix string with which any file named in a -C command line option
4385 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4386 is retained for -C and -D only if the caller of Exim is root. Without
4387 it, the exim user may also use -C and -D and retain privilege.
4389 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4390 command line option is disabled.
4392 6. Macro names set by the -D option must start with an upper case letter, just
4393 like macro names defined in the configuration file.
4395 7. Added "dereference=" facility to LDAP.
4397 8. Two instances of the typo "uknown" in the source files are fixed.
4399 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4400 the Configure-Makefile script screwed up while processing it.
4402 10. Incorporated PCRE 4.4.
4404 11. The SMTP synchronization check was not operating right at the start of an
4405 SMTP session. For example, it could not catch a HELO sent before the client
4406 waited for the greeting. There is now a check for outstanding input at the
4407 point when the greeting is written. Because of the duplex, asynchronous
4408 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4409 way, but not yet received, when the check is performed.
4411 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4412 on TCP/IP sockets, because this apparently causes some broken clients to
4415 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4416 unchanged) from the Cygwin maintainer.
4418 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4419 was supported instead of testing for mbx. Effectively a typo.
4421 15. The spa authenticator server code was not checking that the input it
4422 received was valid base64.
4424 16. The debug output line for the "set" modifier in ACLs was not showing the
4425 name of the variable that was being set.
4427 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4429 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4430 lookup on its own. Something else was needed to trigger the lookup. For
4431 example, a match in host_lookup or the need for a host name in a host list.
4432 Now, if $sender_host_name is referenced and the host name has not yet been
4433 looked up, a lookup is performed. If the lookup fails, the variable remains
4434 empty, and $host_lookup_failed is set to "1".
4436 19. Added "eqi" as a case-independent comparison operator.
4438 20. The saslauthd authentication condition could segfault if neither service
4439 nor realm was specified.
4441 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4442 error message that was logged was misleading, and incoming SMTP
4443 connections were dropped. The message is now more accurate, and temporary
4444 errors are given to SMTP connections.
4446 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4447 (except RSET) with a 421 error, until QUIT is received. However, it was
4448 failing to send a response to QUIT.
4450 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4451 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4452 ACL is now run after the helo_try_verify_hosts code.
4454 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4455 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4456 case-independent, so other case variants are also recognized). Apparently
4457 some systems use these upper case variants.
4459 25. If more than two messages were waiting for the same host, and a transport
4460 filter was specified for the transport, Exim sent two messages over the
4461 same TCP/IP connection, and then failed with "socket operation on non-
4462 socket" when it tried to send the third.
4464 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4466 27. The extern definition of crypt16() in expand.c was not being excluded when
4467 the OS had its own crypt16() function.
4469 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4470 as a preferred synonym for return_size_limit, both as an option and as an
4473 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4475 30. Changed the default configuration ACL to relax the local part checking rule
4476 for addresses that are not in any local domains. For these addresses,
4477 slashes and pipe symbols are allowed within local parts, but the sequence
4478 /../ is explicitly forbidden.
4480 31. SPA server authentication was not clearing the challenge buffer before
4483 32. log_message in a "warn" ACL statement was writing to the reject log as
4484 well as to the main log, which contradicts the documentation and doesn't
4485 seem right (because no rejection is happening). So I have stopped it.
4487 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4488 However, I am unable to do any testing of this.
4490 34. Fixed an infelicity in the appendfile transport. When checking directories
4491 for a mailbox, to see if any needed to be created, it was accidentally
4492 using path names with one or more superfluous leading slashes; tracing
4493 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4495 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4496 contents are added to the log line that is written for every discarded
4497 recipient. (Previously a log_message setting was ignored.)
4499 36. The ${quote: operator now quotes the string if it is empty.
4501 37. The install script runs exim in order to find its version number. If for
4502 some reason other than non-existence or emptiness, which it checks, it
4503 could not run './exim', it was installing it with an empty version number,
4504 i.e. as "exim-". This error state is now caught, and the installation is
4507 38. An argument was missing from the function that creates an error message
4508 when Exim fails to connect to the socket for saslauthd authentication.
4509 This could cause Exim to crash, or give a corrupted message.
4511 39. Added isip, isip4, and isip6 to ${if conditions.
4513 40. The ACL variables $acl_xx are now saved with the message, and can be
4514 accessed later in routers, transports, and filters.
4516 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4517 strings in the file are not string-expanded.
4519 42. If a MAIL command specified a SIZE value that was too large to fit into an
4520 int variable, the check against message_size_limit failed. Such values are
4521 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4522 day this will have to be increased, but I don't think I want to be around
4523 when emails are that large.
4530 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4531 iconv() is not standard in FreeBSD.
4533 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4534 IPv6 enabled. The observed symptom was a segmentation fault on return from
4535 the function os_common_find_running_interfaces() in src/os.c.
4537 3. In the check_special_case() function in daemon.c I had used "errno" as an
4538 argument name, which causes warnings on some systems. This was basically a
4539 typo, since it was named "eno" in the comments!
4541 4. The code that waits for the clock to tick (at a resolution of some fraction
4542 of a second) so as to ensure message-id uniqueness was always waiting for
4543 at least one whole tick, when it could have waited for less. [This is
4544 almost certainly not relevant at current processor speeds, where it is
4545 unlikely to ever wait at all. But we try to future-proof.]
4547 5. The function that sleeps for a time interval that includes fractions of a
4548 second contained a race. It did not block SIGALRM between setting the
4549 timer, and suspending (a couple of lines later). If the interval was short
4550 and the sigsuspend() was delayed until after it had expired, the suspension
4551 never ended. On busy systems this could lead to processes getting stuck for
4554 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4555 process, before it forks any delivery processes. The open lookup caching
4556 mechanism meant that the open file or database connection was passed into
4557 the delivery process. The problem was that delivery processes always tidy
4558 up cached lookup data. This could cause a problem for the next delivery
4559 process started by the queue runner, because the external queue runner
4560 process does not know about the closure. So the next delivery process
4561 still has data in the lookup cache. In the case of a file lookup, there was
4562 no problem because closing a file descriptor in a subprocess doesn't affect
4563 the parent. However, if the lookup was caching a connection to a database,
4564 the connection was closed, and the second delivery process was likely to
4565 see errors such as "PGSQL: query failed: server closed the connection
4566 unexpectedly". The problem has been fixed by closing all cached lookups
4567 in a queue runner before running a delivery process.
4569 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4570 seem to have the "const" qualifier which it has on other OS. I've
4573 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4574 *of the same type* (client or server) with the same public name that an
4575 error should be diagnosed.
4577 9. When Exim looked up a host name for an IP address, but failed to find the
4578 original IP address when looking up the host name (a safety check), it
4579 output the message "<ip address> does not match any IP for NULL", which was
4580 confusing, to say the least. The bug was that the host name should have
4581 appeared instead of "NULL".
4583 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4584 user that is built into the binary, and the -C or -D options is used, root
4585 privilege is dropped before the configuration file is read. In addition,
4586 logging is switched to stderr instead of the normal log files. If the
4587 configuration then re-defines the Exim user, the unprivileged environment
4588 is probably not what is expected, so Exim logs a panic warning message (but
4591 However, if deliver_drop_privilege is set, the unprivileged state may well
4592 be exactly what is intended, so the warning has been cut out in that case,
4593 and Exim is allowed to try to write to its normal log files.
4599 1. smtp_return_error_details was not giving details for temporary sender
4600 or receiver verification errors.
4602 2. Diagnose a configuration error if two authenticators have the same public
4605 3. Exim used not to create the message log file for a message until the first
4606 delivery attempt. This could be confusing when incoming messages were held
4607 for policy or load reasons. The message log file is now created at the time
4608 the message is received, and an initial "Received" line is written to it.
4610 4. The automatically generated man page for command line options had a minor
4611 bug that caused no ill effects; however, a more serious problem was that
4612 the procedure for building the man page automatically didn't always
4613 operate. Consequently, release 4.20 contains an out-of-date version. This
4614 shouldn't happen again.
4616 5. When building Exim with embedded Perl support, the script that builds the
4617 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4618 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4620 6. The freeze_tell option was not being used for messages that were frozen on
4621 arrival, either by an ACL or by local_scan().
4623 7. Added the smtp_incomplete_transaction log selector.
4625 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4626 was accepting AUTH without a new EHLO.
4628 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4629 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4632 10. Logging of TCP/IP connections (when configured) now happens in the main
4633 daemon process instead of the child process, so that the TCP/IP connection
4634 count is more accurate (but it can never be perfect).
4636 11. The use of "drop" in a nested ACL was not being handled correctly in the
4637 outer ACL. Now, if condition failure induced by the nested "drop" causes
4638 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4639 or "require"), the connection is dropped.
4641 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4642 that yield "discard" can now be used with an "accept" or a "discard" verb,
4643 but an error is generated for any others (because I can't see a useful way
4644 to define what should happen).
4646 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4647 are now processed in the same way as lines in the Exim configuration file.
4648 In particular, continuation lines are supported.
4650 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4652 15. Added -ti meaning -t -i.
4654 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4655 domains, and warn by logging if others are found.
4657 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4658 returned from GIFCONF when Exim is trying to find the list of interfaces on
4659 a host. The code in os.c has been modified to copy each ifreq to an aligned
4660 structure in all cases.
4662 Also, in some cases, the returned ifreq's were being copied to a 'struct
4663 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4664 means the last couple of bytes of an IPv6 address could be chopped if the
4665 ifreq contained only a normal sockaddr (14 bytes storage).
4667 18. Named domain lists were not supported in the hosts_treat_as_local option.
4668 An entry such as +xxxx was not recognized, and was treated as a literal
4671 19. Ensure that header lines added by a DATA ACL are included in the reject log
4672 if the ACL subsequently rejects the message.
4674 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4675 MD5 (which is deprecated).
4677 21. When testing a filter file using -bf, Exim was writing a message when it
4678 took the sender from a "From " line in the message, but it was not doing so
4679 when it took $return_path from a Return-Path: header line. It now does.
4681 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4682 with a valid header line field name (a series of printing characters
4683 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4685 23. Changed "disc" in the source to "disk" to conform to the documentation and
4686 the book and for uniformity.
4688 24. Ignore Sendmail's -Ooption=value command line item.
4690 25. When execve() failed while trying to run a command in a pipe transport,
4691 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4692 could be confused with a return value of 69 from the command itself. This
4693 has been changed to 127, the value the shell returns if it is asked to run
4694 a non-existent command. The wording for the related log line suggests a
4695 non-existent command as the problem.
4697 26. If received_header_text expands to an empty string, do not add a Received:
4698 header line to the message. (Well, it adds a token one on the spool, but
4699 marks it "old" so that it doesn't get used or transmitted.)
4701 27. Installed eximstats 1.28 (addition of -nt option).
4703 28. There was no check for failure on the call to getsockname() in the daemon
4704 code. This can fail if there is a shortage of resources on the system, with
4705 ENOMEM, for example. A temporary error is now given on failure.
4707 29. Contrary to the C standard, it seems that in some environments, the
4708 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4709 program. Exim now does this explicitly; it affects the formatting of
4710 timestamps using strftime().
4712 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4713 complaints. I've now initialized all the variables, to avoid this.
4715 32. Header lines added by a system filter were not being "seen" during
4716 transport-time rewrites.
4718 33. The info_callback() function passed to OpenSSL is set up with type void
4719 (*)(SSL *, int, int), as described somewhere. However, when calling the
4720 function (actually a macro) that sets it up, the type void(*)() is
4721 expected. I've put in a cast to prevent warnings from picky compilers.
4723 34. If a DNS black list lookup found a CNAME record, but there were no A
4724 records associated with the domain it pointed at, Exim crashed.
4726 35. If a DNS black list lookup returned more than one A record, Exim ignored
4727 all but the first. It now scans all returned addresses if a particular IP
4728 value is being sought. In this situation, the contents of the
4729 $dnslist_value variable are a list of all the addresses, separated by a
4732 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4733 to accept a host name and all its aliases if the forward lookup for any of
4734 them yielded the IP address of the incoming connection. Now it accepts only
4735 those names whose forward lookup yields the correct IP address. Any other
4736 names are discarded. This closes a loophole whereby a rogue DNS
4737 administrator could create reverse DNS records to break through a
4738 wildcarded host restriction in an ACL.
4740 37. If a user filter or a system filter that ran in a subprocess used any of
4741 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4742 the wrong values were passed to the pipe command ($thisaddress had the
4743 value of $0, $0 had the value of $1, etc). This bug was introduced by
4744 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4746 38. Improved the line breaking for long SMTP error messages from ACLs.
4747 Previously, if there was no break point between 40 and 75 characters, Exim
4748 left the rest of the message alone. Two changes have been made: (a) I've
4749 reduced the minimum length to 35 characters; (b) if it can't find a break
4750 point between 35 and 75 characters, it looks ahead and uses the first one
4751 that it finds. This may give the occasional overlong line, but at least the
4752 remaining text gets split now.
4754 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4755 file descriptors might be low, and that setting 1000 would always raise it.
4756 It turns out that in some environments, the limit is already over 1000 and
4757 that lowering it causes trouble. So now Exim takes care not to decrease it.
4759 40. When delivering a message, the value of $return_path is set to $sender_
4760 address at the start of routing (routers may change the value). By an
4761 oversight, this default was not being set up when an address was tested by
4762 -bt or -bv, which affected the outcome if any router or filter referred to
4765 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4766 log only when "message" or "log_message" are set. However, if one of the
4767 conditions was an address verification, or a call to a nested ACL, the
4768 messages generated by the underlying test were being passed through. This
4769 no longer happens. The underlying message is available in $acl_verify_
4770 message for both "message" and "log_message" expansions, so it can be
4771 passed through if needed.
4773 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4774 new expansion $bh_ to give the encoded byte string without charset
4775 translation. Translation happens only if iconv() is available; HAVE_ICONV
4776 indicates this at build time. HEADERS_CHARSET gives the charset to
4777 translate to; headers_charset can change it in the configuration, and
4778 "headers charset" can change it in an individual filter file.
4780 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4781 that creates RFC 2047 encoded "words" labels them as that charset instead
4782 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4783 expansion operator; (ii) when Exim creates a From: line for a local
4784 message; (iii) when a header line is rewritten to include a "phrase" part.
4786 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4787 buggy, causing it to skip the first lines of messages whose message ID
4788 ended in 'D'. This would not have bitten before Exim release 4.14, because
4789 message IDs were unlikely to end in 'D' before then. The effect was to have
4790 incorrect size information for certain domains.
4792 45. #include "config.h" was missing at the start of the crypt16.c module. This
4793 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4796 46. If there was a timeout during a "random" callout check, Exim treated it as
4797 a failure of the random address, and carried on sending RSET and the real
4798 address. If the delay was just some slowness somewhere, the response to the
4799 original RCPT would be taken as a response to RSET and so on, causing
4800 mayhem of various kinds.
4802 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4803 when I implemented it. It didn't allow for the fact that some option values
4804 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4805 the right test for positive values.
4807 48. Domain names in DNS records are case-independent. Exim always looks them up
4808 in lower case. Some resolvers return domain names in exactly the case they
4809 appear in the zone file, that is, they may contain uppercase letters. Not
4810 all resolvers do this - some return always lower case. Exim was treating a
4811 change of case by a resolver as a change of domain, similar to a widening
4812 of a domain abbreviation. This triggered its re-routing code and so it was
4813 trying to route what was effectively the same domain again. This normally
4814 caused routing to fail (because the router wouldn't handle the domain
4815 twice). Now Exim checks for this case specially, and just changes the
4816 casing of the domain that it ultimately uses when it transmits the message
4819 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4822 50. If a filter generated a file delivery with a non-absolute name (possible if
4823 no home directory exists for the router), the forbid_file option was not
4826 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4827 the existing equality matching.
4829 52. Exim was using ints instead of ino_t variables in some places where it was
4830 dealing with inode numbers.
4832 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4833 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4834 called TMPDIR, and if it finds it is different, it changes its value.
4836 54. The smtp_printf() function is now made available to local_scan() so
4837 additional output lines can be written before returning. There is also an
4838 smtp_fflush() function to enable the detection of a dropped connection.
4839 The variables smtp_input and smtp_batched_input are exported to
4842 55. Changed the default runtime configuration: the message "Unknown user"
4843 has been removed from the ACL, and instead placed on the localuser router,
4844 using the cannot_route_message feature. This means that any verification
4845 failures that generate their own messages won't get overridden. Similarly,
4846 the "Unrouteable address" message that was in the ACL for unverifiable
4847 relay addresses has also been removed.
4849 56. Added hosts_avoid_esmtp to the smtp transport.
4851 57. The exicyclog script was not checking for the esoteric option
4852 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4853 will work only if exicyclog is run under the appropriate euid.
4855 58. Following a discussion on the list, the rules by which Exim recognises line
4856 endings on incoming messages have been changed. The -dropcr and drop_cr
4857 options are now no-ops, retained only for backwards compatibility. The
4858 following line terminators are recognized: LF CRLF CR. However, special
4859 processing applies to CR:
4861 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4862 nor a local message in the state where . is a terminator.
4864 (ii) If a bare CR is encountered in a header line, an extra space is added
4865 after the line terminator so as not to end the header. The reasoning
4866 behind this is that bare CRs in header lines are most likely either
4867 to be mistakes, or people trying to play silly games.
4869 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4870 was being incorrectly given as 18 bytes larger than it should have been.
4871 This is a VOB (very old bug).
4873 60. This may never have affected anything current, but just in case it has:
4874 When the local host is found other than at the start of a list of hosts,
4875 the local host, those with the same MX, and any that follow, are discarded.
4876 When the list in question was part of a longer list of hosts, the following
4877 hosts (not currently being processed) were also being discarded. This no
4878 longer happens. I'm not sure if this situation could ever has previously
4881 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4884 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4885 header. This is something that is recommended in a new Internet Draft, and
4886 is something that is documented as being done by Sendmail. There are two
4887 possible values. For messages generated by the autoreply transport, Exim
4890 Auto-Submitted: auto-replied
4892 whereas for all other generated messages (e.g. bounces) it adds
4894 Auto-Submitted: auto-generated
4896 63. The "personal" condition in filters now includes a test for the
4897 Auto-Submitted: header. If it contains the string "auto-" the message it
4898 not considered personal.
4900 64. Added rcpt_include_affixes as a generic transport option.
4902 65. Added queue_only_override (default true).
4904 66. Added the syslog_duplication option.
4906 67. If what should have been the first header line of a message consisted of
4907 a space followed by a colon, Exim was mis-interpreting it as a header line.
4908 It isn't of course - it is syntactically invalid and should therefore be
4909 treated as the start of the message body. The misbehaviour could have
4910 caused a number of strange effects, including loss of data in subsequent
4911 header lines, and spool format errors.
4913 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4914 client host had authenticated. This control can now be exercised by an ACL
4915 for more flexibility.
4917 69. By default, callouts do not happen when testing with -bh. There is now a
4918 variant, -bhc, which does actually run the callout code, including
4919 consulting and updating the callout cache.
4921 70. Added support for saslauthd authentication, courtesy of Alexander
4924 71. If statvfs() failed on the spool or log directories while checking their
4925 size for availability, Exim confusingly gave the error "space shortage".
4926 Furthermore, in debugging mode it crashed with a floating point exception.
4927 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4928 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4929 serious problem, Exim now writes to the main and panic logs when this
4930 happens, with details of the failure. It then refuses to accept the
4931 incoming message, giving the message "spool directory problem" or "log
4932 directory problem" with a 421 code for SMTP messages.
4934 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4935 0, 1, and 2 exist, because some OS complain for execs without them (see
4936 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4937 descriptors. However, the code omitted to check that the open succeeded,
4938 causing mysterious errors if for some reason the permissions on /dev/null
4939 got screwed. Now Exim writes a message to the main and panic logs, and
4940 bombs out if it can't open /dev/null.
4942 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4943 interact so that it is all more flexible. It is supposed to remain
4944 backwards compatible. Also added extra_local_interfaces.
4946 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4947 to bomb out with an assertion failure - to the client this appears as a
4948 connection drop. This problem occurs in the part of the code that was taken
4949 from the Samba project. Fortunately, the assertion is in a very simple
4950 function, so I have fixed this by reproducing the function inline in the
4951 one place where it is called, and arranging for authentication to fail
4952 instead of killing the process with assert().
4954 75. The SPA client code was not working when the server requested OEM rather
4955 than Unicode encoding.
4957 76. Added code to make require_files with a specific uid setting more usable in
4958 the case where statting the file as root fails - usually a non-root-mounted
4959 NFS file system. When this happens and the failure is EACCES, Exim now
4960 forks a subprocess and does the per-uid checking as the relevant uid.
4962 77. Added process_log_path.
4964 78. If log_file_path was not explicitly set, a setting of check_log_space or
4965 check_log_inodes was ignored.
4967 79. If a space check for the spool or log partitions fails, the incident is now
4968 logged. Of course, in the latter case the data may get lost...
4970 80. Added the %p formatting code to string_format() so that it can be used to
4971 print addresses in debug_print(). Adjusted all the address printing in the
4972 debugging in store.c to use %p rather than %d.
4974 81. There was a concern that a line of code in smtp_in.c could overflow a
4975 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4976 initially expressed, the concern was not well-founded, because trailing
4977 spaces are removed early. However, if the trailing spaces were followed by
4978 a NULL, they did not get removed, so the overflow was possible. Two fixes
4981 (a) I re-wrote the offending code in a cleaner fashion.
4982 (b) If an incoming SMTP command contains a NULL character, it is rejected
4985 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4986 runs initgroups(), so that if the Exim user is in any additional groups,
4987 they will be used during message reception.
4993 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob