1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.575 2009/10/19 11:47:43 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
130 TK/01 Add preliminary DKIM support. Currently requires a forked version of
131 ALT-N's libdkim that I have put here:
132 http://duncanthrax.net/exim-experimental/
134 Note to Michael Haardt: I had to rename some vars in sieve.c. They
135 were called 'true' and it seems that C99 defines that as a reserved
136 keyword to be used with 'bool' variable types. That means you could
137 not include C99-style headers which use bools without triggering
138 build errors in sieve.c.
140 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
141 as mailq or other aliases. Changed the --help handling significantly
142 to do whats expected. exim_usage() emits usage/help information.
144 SC/01 Added the -bylocaldomain option to eximstats.
146 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
148 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
150 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
156 PH/01 Another patch from the Sieve maintainer.
158 PH/02 When an IPv6 address is converted to a string for single-key lookup
159 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
160 dots are used instead of colons so that keys in lsearch files need not
161 contain colons. This was done some time before quoting was made available
162 in lsearch files. However, iplsearch files do require colons in IPv6 keys
163 (notated using the quote facility) so as to distinguish them from IPv4
164 keys. This meant that lookups for IP addresses in host lists did not work
165 for iplsearch lookups.
167 This has been fixed by arranging for IPv6 addresses to be expressed with
168 colons if the lookup type is iplsearch. This is not incompatible, because
169 previously such lookups could never work.
171 The situation is now rather anomolous, since one *can* have colons in
172 ordinary lsearch keys. However, making the change in all cases is
173 incompatible and would probably break a number of configurations.
175 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
178 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
179 conversion specification without a maximum field width, thereby enabling
180 a rogue spamd server to cause a buffer overflow. While nobody in their
181 right mind would setup Exim to query an untrusted spamd server, an
182 attacker that gains access to a server running spamd could potentially
183 exploit this vulnerability to run arbitrary code as the Exim user.
185 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
186 $primary_hostname instead of what libspf2 thinks the hosts name is.
188 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
189 a directory entry by the name of the lookup key. Previously, if a
190 symlink pointed to a non-existing file or a file in a directory that
191 Exim lacked permissions to read, a lookup for a key matching that
192 symlink would fail. Now it is enough that a matching directory entry
193 exists, symlink or not. (Bugzilla 503.)
195 PH/03 The body_linecount and body_zerocount variables are now exported in the
198 PH/04 Added the $dnslist_matched variable.
200 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
201 This means they are set thereafter only if the connection becomes
204 PH/06 Added the client_condition to authenticators so that some can be skipped
205 by clients under certain conditions.
207 PH/07 The error message for a badly-placed control=no_multiline_responses left
208 "_responses" off the end of the name.
210 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
212 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
213 (without spaces) instead of just copying the configuration text.
215 PH/10 Added the /noupdate option to the ratelimit ACL condition.
217 PH/11 Added $max_received_linelength.
219 PH/12 Added +ignore_defer and +include_defer to host lists.
221 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
222 way in which PCRE > 7.0 is built.
224 PH/14 Implemented queue_only_load_latch.
226 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
227 MAIL command. The effect was to mangle the value on 64-bit systems.
229 PH/16 Another patch from the Sieve maintainer.
231 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
233 PH/18 If a system quota error occurred while trying to create the file for
234 a maildir delivery, the message "Mailbox is full" was not appended to the
235 bounce if the delivery eventually timed out. Change 4.67/27 below applied
236 only to a quota excession during the actual writing of the file.
238 PH/19 It seems that peer DN values may contain newlines (and other non-printing
239 characters?) which causes problems in log lines. The DN values are now
240 passed through string_printing() before being added to log lines.
242 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
243 and InterBase are left for another time.)
245 PH/21 Added message_body_newlines option.
247 PH/22 Guard against possible overflow in moan_check_errorcopy().
249 PH/23 POSIX allows open() to be a macro; guard against that.
251 PH/24 If the recipient of an error message contained an @ in the local part
252 (suitably quoted, of course), incorrect values were put in $domain and
253 $local_part during the evaluation of errors_copy.
259 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
260 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
263 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
264 issue a MAIL command.
266 PH/02 In an ACL statement such as
268 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
270 if a client was not listed at all, or was listed with a value other than
271 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
272 the condition was not true (as it should be), so access was not denied.
273 The bug was that the ! inversion was incorrectly passed on to the second
274 item. This has been fixed.
276 PH/03 Added additional dnslists conditions == and =& which are different from
277 = and & when the dns lookup returns more than one IP address.
279 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
280 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
282 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
283 FSYNC, which compiles an option called disable_fsync that allows for
284 bypassing fsync(). The documentation is heavily laced with warnings.
286 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
288 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
289 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
290 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
291 including adding "make clean"; (3) Added -fPIC when compiling the test
292 dynamically loaded module, to get rid of a warning.
294 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
295 message fails, move_frozen_messages = true and ignore_bounce_errors_after
296 = 0s. The bug is otherwise harmless.
298 PH/07 There was a bug in the dovecot authenticator such that the value of
299 $auth1 could be overwritten, and so not correctly preserved, after a
300 successful authentication. This usually meant that the value preserved by
301 the server_setid option was incorrect.
303 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
305 PH/09 Installed PCRE release 7.0.
307 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
308 run for batched SMTP input. It is now run at the start of every message
309 in the batch. While fixing this I discovered that the process information
310 (output by running exiwhat) was not always getting set for -bs and -bS
311 input. This is fixed, and it now also says "batched" for BSMTP.
313 PH/11 Added control=no_pipelining.
315 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
316 patch, slightly modified), and move the expansion of helo_data till after
317 the connection is made in the smtp transport (so it can use these
320 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
322 PH/14 Added log_selector = +pid.
324 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
326 PH/16 Add ${if forany and ${if forall.
328 PH/17 Added dsn_from option to vary the From: line in DSNs.
330 PH/18 Flush SMTP output before performing a callout, unless control =
331 no_callout_flush is set.
333 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
334 was true (the default) a successful delivery failed to delete the retry
335 item, thus causing premature timeout of the address. The bug is now
338 PH/20 Added hosts_avoid_pipelining to the smtp transport.
340 PH/21 Long custom messages for fakedefer and fakereject are now split up
341 into multiline reponses in the same way that messages for "deny" and
342 other ACL rejections are.
344 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
345 with slight modification.
347 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
348 draft, changing the syntax and factoring some duplicate code".
350 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
351 for deliveries of the second and subsequent messages over the same SMTP
354 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
355 ${reduce, with only minor "tidies".
357 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
359 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
360 expansion side effects.
362 PH/27 When a message times out after an over-quota error from an Exim-imposed
363 quota, the bounce message says "mailbox is full". This message was not
364 being given when it was a system quota that was exceeded. It now should
367 MH/03 Made $recipients available in local_scan(). local_scan() already has
368 better access to the recipient list through recipients_list[], but
369 $recipients can be useful in postmaster-provided expansion strings.
371 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
372 in the case of a MAIL command with additional options following the
373 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
374 were accidentally chopped off.
376 PH/29 SMTP synchronization checks are implemented when a command is read -
377 there is a check that no more input is waiting when there shouldn't be
378 any. However, for some commands, a delay in an ACL can mean that it is
379 some time before the response is written. In this time, more input might
380 arrive, invalidly. So now there are extra checks after an ACL has run for
381 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
382 pipelining has not been advertised.
384 PH/30 MH's patch to allow iscntrl() characters to be list separators.
386 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
387 returned in the SMTP response when smtp_return_error_details was false.
390 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
391 instead of the C I/O that was originally supplied, because problems were
394 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
395 Exim which did not show up earlier: it was assuming that a call to
396 SSL_CTX_set_info_callback() might give an error value. In fact, there is
397 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
398 was a macro that became an assignment, so it seemed to work. This has
399 changed to a proper function call with a void return, hence the compile
400 error. Exim's code has been fixed.
402 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
405 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
407 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
409 PH/37 If a message is not accepted after it has had an id assigned (e.g.
410 because it turns out to be too big or there is a timeout) there is no
411 "Completed" line in the log. When some messages of this type were
412 selected by exigrep, they were listed as "not completed". Others were
413 picked up by some special patterns. I have improved the selection
414 criteria to be more general.
416 PH/38 The host_find_failed option in the manualroute router can now be set
417 to "ignore", to completely ignore a host whose IP address cannot be
418 found. If all hosts are ignored, the behaviour is controlled by the new
419 host_all_ignored option.
421 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
422 homing or because of multiple MX records with /mx) generated more than
423 one IP address, and the following item turned out to be the local host,
424 all the secondary addresses of the first item were incorrectly removed
425 from the list, along with the local host and any following hosts (which
426 is what is supposed to happen).
428 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
429 whoever called Exim into the -H file. In the case of the daemon it was
430 behaving confusingly. When first started, it used values for whoever
431 started the daemon, but after a SIGHUP it used the Exim user (because it
432 calls itself on a restart). I have changed the code so that it now always
435 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
436 message are rejected with the same error (e.g. no authentication or bad
437 sender address), and a DATA command is nevertheless sent (as can happen
438 with PIPELINING or a stupid MUA), the error message that was given to the
439 RCPT commands is included in the rejection of the DATA command. This is
440 intended to be helpful for MUAs that show only the final error to their
443 PH/42 Another patch from the Sieve maintainer.
445 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
446 Eximstats - Fixed some broken HTML links and added missing column headers
448 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
449 columns for Rejects, Temp Rejects, Ham, and Spam rows.
451 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
453 PH/43 Yet another patch from the Sieve maintainer.
455 PH/44 I found a way to check for a TCP/IP connection going away before sending
456 the response to the final '.' that terminates a message, but only in the
457 case where the client has not sent further data following the '.'
458 (unfortunately, this is allowed). However, in many cases there won't be
459 any further data because there won't be any more messages to send. A call
460 to select() can be used: if it shows that the input is "ready", there is
461 either input waiting, or the socket has been closed. An attempt to read
462 the next input character can distinguish the two cases. Previously, Exim
463 would have sent an OK response which the client would never have see.
464 This could lead to message repetition. This fix should cure that, at
465 least in a lot of common cases.
467 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
468 advertised in response to EHLO.
474 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
475 fixed by 4.65/MH/01 (is this a record?) are fixed:
477 (i) An empty string was always treated as zero by the numeric comparison
478 operators. This behaviour has been restored.
480 (ii) It is documented that the numeric comparison operators always treat
481 their arguments as decimal numbers. This was broken in that numbers
482 starting with 0 were being interpreted as octal.
484 While fixing these problems I realized that there was another issue that
485 hadn't been noticed. Values of message_size_limit (both the global option
486 and the transport option) were treated as octal if they started with 0.
487 The documentation was vague. These values are now always treated as
488 decimal, and I will make that clear in the documentation.
494 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
495 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
498 MH/01 Don't check that the operands of numeric comparison operators are
499 integers when their expansion is in "skipping" mode (fixes bug
500 introduced by 4.64-PH/07).
502 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
503 child addresses, Exim now panics and dies. Previously, because the count
504 is held in a short int, deliveries were likely to be lost. As such a
505 large number of recipients for a single message is ridiculous
506 (performance will be very, very poor), I have chosen to impose a limit
507 rather than extend the field.
513 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
514 leftover -K file (the existence of which was triggered by #402).
515 While we were at it, introduced process PID as part of the -K
516 filename. This should rule out race conditions when creating
519 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
520 processing considerably. Previous code took too long for large mails,
521 triggering a timeout which in turn triggers #401.
523 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
524 in the DK code in transports.c. sendfile() is not really portable,
525 hence the _LINUX specificness.
527 TF/01 In the add_headers option to the mail command in an Exim filter,
528 there was a bug that Exim would claim a syntax error in any
529 header after the first one which had an odd number of characters
532 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
533 callout verification, Exim cached a "reject" for the entire domain. This
534 is correct for most verifications, but it is not correct for a recipient
535 verification with use_sender or use_postmaster set, because in that case
536 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
537 case of MAIL FROM:<> rejection from other early rejections (e.g.
538 rejection of HELO). When verifying a recipient using a non-null MAIL
539 address, the cache is ignored if it shows MAIL FROM:<> rejection.
540 Whatever the result of the callout, the value of the domain cache is
541 left unchanged (for any other kind of callout, getting as far as trying
542 RCPT means that the domain itself is ok).
544 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
547 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
548 manifest itself as EPIPE rather than ECONNECT. When tidying away a
549 session, the daemon ignores ECONNECT errors and logs others; it now
550 ignores EPIPE as well.
552 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
553 (quoted-printable decoding).
555 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
556 later the small subsequent patch to fix an introduced bug.
558 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
560 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
562 PH/08 An error is now given if message_size_limit is specified negative.
564 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
565 to be given (somewhat) arbitrary names.
567 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
570 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
571 miscellaneous code fixes
573 PH/10 Added the log_reject_target ACL modifier to specify where to log
576 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
577 hostname. This is wrong, because it relates to the incoming message (and
578 probably the interface on which it is arriving) and not to the outgoing
579 callout (which could be using a different interface). This has been
580 changed to use the value of the helo_data option from the smtp transport
581 instead - this is what is used when a message is actually being sent. If
582 there is no remote transport (possible with a router that sets up host
583 addresses), $smtp_active_hostname is used.
585 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
586 tweaks were necessary in order to get it to work (see also 21 below):
587 (a) The code assumed that strncpy() returns a negative number on buffer
588 overflow, which isn't the case. Replaced with Exim's string_format()
590 (b) There were several signed/unsigned issues. I just did the minimum
591 hacking in of casts. There is scope for a larger refactoring.
592 (c) The code used strcasecmp() which is not a standard C function.
593 Replaced with Exim's strcmpic() function.
594 (d) The code set only $1; it now sets $auth1 as well.
595 (e) A simple test gave the error "authentication client didn't specify
596 service in request". It would seem that Dovecot has changed its
597 interface. Fortunately there's a specification; I followed it and
598 changed what the client sends and it appears to be working now.
600 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
603 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
604 address A is aliased to B and C, where B exists and C does not. Without
605 -v the output is "A verified" because verification stops after a
606 successful redirection if more than one address is generated. However,
607 with -v the child addresses are also verified. Exim was outputting "A
608 failed to verify" and then showing the successful verification for C,
609 with its parentage. It now outputs "B failed to verify", showing B's
610 parentage before showing the successful verification of C.
612 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
613 look up a TXT record in a specific list after matching in a combined
616 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
617 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
618 they consult the DNS. I had assumed they would set it the way they
619 wanted; and indeed my experiments on Linux seem to show that in some
620 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
621 To be on the safe side, however, I have now made the interface to
622 host_find_byname() similar to host_find_bydns(), with an argument
623 containing the DNS resolver options. The host_find_byname() function now
624 sets these options at its start, just as host_find_bydns() does. The smtp
625 transport options dns_qualify_single and dns_search_parents are passed to
626 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
627 of host_find_byname() use the default settings of RES_DEFNAMES
628 (qualify_single) but not RES_DNSRCH (search_parents).
630 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
631 spool_read_header() do less string testing, by means of a preliminary
632 switch on the second character of optional "-foo" lines. (This is
633 overdue, caused by the large number of possibilities that now exist.
634 Originally there were few.) While I was there, I also converted the
635 str(n)cmp tests so they don't re-test the leading "-" and the first
636 character, in the hope this might squeeze out yet more improvement.
638 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
639 flag allowing group syntax was set by the header_syntax check but not
640 turned off, possible causing trouble later; (2) The flag was not being
641 set at all for the header_verify test, causing "group"-style headers to
642 be rejected. I have now set it in this case, and also caused header_
643 verify to ignore an empty address taken from a group. While doing this, I
644 came across some other cases where the code for allowing group syntax
645 while scanning a header line wasn't quite right (mostly, not resetting
646 the flag correctly in the right place). These bugs could have caused
647 trouble for malformed header lines. I hope it is now all correct.
649 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
650 with the "reply" argument non-NULL. The code, however (which originally
651 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
652 but it didn't always do it. This confused somebody who was copying the
653 code for some other use. I have removed all the tests.
655 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
656 feature that was used to support insecure browsers during the U.S. crypto
657 embargo. It requires special client support, and Exim is probably the
658 only MTA that supported it -- and would never use it because real RSA is
659 always available. This code has been removed, because it had the bad
660 effect of slowing Exim down by computing (never used) parameters for the
661 RSA_EXPORT functionality.
663 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
664 authenticator to fail if there's a tab character in the incoming data
665 (there should never be unless someone is messing about, as it's supposed
666 to be base64-encoded). Also added, on Timo's advice, the "secured" option
667 if the connection is using TLS or if the remote IP is the same as the
668 local IP, and the "valid-client-cert option" if a client certificate has
671 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
672 authenticators. This can be used for authorization after authentication
673 succeeds. (In the case of plaintext, it servers for both authentication
676 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
677 if any retry times were supplied.
679 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
680 connection, where there is no client IP address to check. In this
681 situation, the verify now always succeeds.
683 PH/25 Applied John Jetmore's -Mset patch.
685 PH/26 Added -bem to be like -Mset, but loading a message from a file.
687 PH/27 In a string expansion for a processed (not raw) header when multiple
688 headers of the same name were present, leading whitespace was being
689 removed from all of them, but trailing whitespace was being removed only
690 from the last one. Now trailing whitespace is removed from each header
691 before concatenation. Completely empty headers in a concatenation (as
694 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
695 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
697 PH/29 [Removed. This was a change that I later backed out, and forgot to
698 correct the ChangeLog entry (that I had efficiently created) before
699 committing the later change.]
701 PH/30 Exim was sometimes attempting to deliver messages that had suffered
702 address errors (4xx response to RCPT) over the same connection as other
703 messages routed to the same hosts. Such deliveries are always "forced",
704 so retry times are not inspected. This resulted in far too many retries
705 for the affected addresses. The effect occurred only when there were more
706 hosts than the hosts_max_try setting in the smtp transport when it had
707 the 4xx errors. Those hosts that it had tried were not added to the list
708 of hosts for which the message was waiting, so if all were tried, there
709 was no problem. Two fixes have been applied:
711 (i) If there are any address or message errors in an SMTP delivery, none
712 of the hosts (tried or untried) are now added to the list of hosts
713 for which the message is waiting, so the message should not be a
714 candidate for sending over the same connection that was used for a
715 successful delivery of some other message. This seems entirely
716 reasonable: after all the message is NOT "waiting for some host".
717 This is so "obvious" that I'm not sure why it wasn't done
718 previously. Hope I haven't missed anything, but it can't do any
719 harm, as the worst effect is to miss an optimization.
721 (ii) If, despite (i), such a delivery is accidentally attempted, the
722 routing retry time is respected, so at least it doesn't keep
723 hammering the server.
725 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
726 in ${readsocket because some servers need this prod.
728 PH/32 Added some extra debug output when updating a wait-xxx database.
730 PH/33 The hint "could be header name not terminated by colon", which has been
731 given for certain expansion errors for a long time, was not being given
732 for the ${if def:h_colon_omitted{... case.
734 PH/34 The spec says: "With one important exception, whenever a domain list is
735 being scanned, $domain contains the subject domain." There was at least
736 one case where this was not true.
738 PH/35 The error "getsockname() failed: connection reset by peer" was being
739 written to the panic log as well as the main log, but it isn't really
740 panic-worthy as it just means the connection died rather early on. I have
741 removed the panic log writing for the ECONNRESET error when getsockname()
744 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
745 runs only) independently of the message's sender address. This meant
746 that, if the 4xx error was in fact related to the sender, a different
747 message to the same recipient with a different sender could confuse
748 things. In particualar, this can happen when sending to a greylisting
749 server, but other circumstances could also provoke similar problems.
750 I have changed the default so that the retry time for these errors is now
751 based a combination of the sender and recipient addresses. This change
752 can be overridden by setting address_retry_include_sender=false in the
755 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
756 remote server are returned as part of bounce messages. This was not
757 happening for LMTP over a pipe (the lmtp transport), but now it is the
758 same for both kinds of LMTP.
760 PH/38 Despite being documented as not happening, Exim was rewriting addresses
761 in header lines that were in fact CNAMEs. This is no longer the case.
763 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
764 and queue runs started by the daemon processed all messages. This has
765 been fixed so that -R and -S can now usefully be given with -q<time>.
767 PH/40 Import PCRE release 6.7 (fixes some bugs).
769 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
771 PH/42 Give an error if -q is specified more than once.
773 PH/43 Renamed the variables $interface_address and $interface_port as
774 $received_ip_address and $received_port, to make it clear that these
775 values apply to message reception, and not to the outgoing interface when
776 a message is delivered. (The old names remain recognized, of course.)
778 PH/44 There was no timeout on the connect() call when using a Unix domain
779 socket in the ${readsocket expansion. There now is.
781 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
782 be meaningful with "accept".
784 SC/01 Eximstats V1.43
785 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
787 SC/02 Eximstats V1.44
788 Use a glob alias rather than an array ref in the generated
789 parser. This improves both readability and performance.
791 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
792 Collect SpamAssassin and rejection statistics.
793 Don't display local sender or destination tables unless
794 there is data to show.
795 Added average volumes into the top table text output.
797 SC/04 Eximstats V1.46
798 Collect data on the number of addresses (recipients)
799 as well as the number of messages.
801 SC/05 Eximstats V1.47
802 Added 'Message too big' to the list of mail rejection
803 reasons (thanks to Marco Gaiarin).
805 SC/06 Eximstats V1.48
806 Mainlog lines which have GMT offsets and are too short to
807 have a flag are now skipped.
809 SC/07 Eximstats V1.49 (Alain Williams)
810 Added the -emptyok flag.
812 SC/08 Eximstats V1.50
813 Fixes for obtaining the IP address from reject messages.
815 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
816 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
817 whitesspace changes from 4.64-PH/27
819 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
822 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
823 are found, allow negative numbers in numeric criteria)
825 JJ/06 exipick.20061117.2, added new $message_body_missing variable
827 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
828 to match changes made in 4.64-PH/43
830 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
832 PH/47 Put in an explicit test for a DNS lookup of an address record where the
833 "domain" is actually an IP address, and force a failure. This locks out
834 those revolvers/nameservers that support "A-for-A" lookups, in
835 contravention of the specifications.
837 PH/48 When a host name was looked up from an IP address, and the subsequent
838 forward lookup of the name timed out, the host name was left in
839 $sender_host_name, contrary to the specification.
841 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
842 restricted to single-key lookups, Exim was not diagnosing an error if
843 * or *@ was used with a query-style lookup.
845 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
847 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
848 long ago, but noone interested enough thought of it. Let's just say that
849 the "1.1" means that there are some new functions that weren't there at
850 some point in the past.
852 PH/51 Error processing for expansion failure of helo_data from an smtp
853 transport during callout processing was broken.
855 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
856 tested/used via the -bh/-bhc/-bs options.
858 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
859 bug, fixed in subsequent PCRE releases).
861 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
862 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
864 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
870 SC/01 Use a glob alias rather than an array ref in eximstats generated
871 parser. This improves both readability and performance.
873 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
874 Don't display local sender or destination tables in eximstats unless
875 there is data to show.
876 Added average volumes into the eximstats top table text output.
878 SC/03 Collect data on the number of addresses (recipients) as well
879 as the number of messages in eximstats.
881 TF/01 Correct an error in the documentation for the redirect router. Exim
882 does (usually) call initgroups() when daemonizing.
884 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
885 with consistent privilege compared to when running as a daemon.
887 TF/03 Note in the spec that $authenticated_id is not set for local
888 submissions from trusted users.
890 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
891 Thanks to Dean Brooks <dean@iglou.com> for the patch.
893 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
894 by adding some example configuration directives to the default
895 configuration file. A little bit of work is required to uncomment the
896 directives and define how usernames and passwords are checked, but
897 there is now a framework to start from.
899 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
900 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
901 without this. I don't know how relevant this is to other LDAP libraries.
903 PH/02 Add the verb name to the "unknown ACL verb" error.
905 PH/03 Magnus Holmgren's patch for filter_prepend_home.
907 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
909 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
910 directory not expanded when it should be if an expanded home directory
911 was set for the address (which is overridden by the transport).
913 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
916 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
917 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
918 because it is too late at that time, and has no effect.
920 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
921 security issue with \' (bugzilla #107). I could not use the
922 PQescapeStringConn() function, because it needs a PGconn value as one of
925 PH/08 When testing addresses using -bt, indicate those final addresses that
926 are duplicates that would not cause an additional delivery. At least one
927 person was confused, thinking that -bt output corresponded to deliveries.
928 (Suppressing duplicates isn't a good idea as you lose the information
929 about possibly different redirections that led to the duplicates.)
931 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
932 systems where poll() doesn't work, in particular OS X.
934 PH/10 Added more information to debugging output for retry time not reached.
936 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
937 operations in malware.c.
939 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
942 PH/13 If write_rejectlog was set false when logging was sent to syslog with
943 syslog_duplication set false, log lines that would normally be written
944 both the the main log and to the reject log were not written to syslog at
947 PH/14 In the default configuration, change the use of "message" in ACL warn
948 statements to "add_header".
950 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
951 not followed by a command (e.g. "seen endif").
953 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
954 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
957 PH/17 Added extra conditions to the default value of delay_warning_condition
961 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
962 { match{$h_precedence:}{(?i)bulk|list|junk} } \
963 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
966 The Auto-Submitted: and various List- headers are standardised, whereas I
967 don't think Precedence: ever was.
969 PH/18 Refactored debugging code in route_finduser() to show more information,
970 in particular, the error code if getpwnam() issues one.
972 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
973 This is apparently needed in addition to the PH/07 change above to avoid
974 any possible encoding problems.
976 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
977 but not after initializing Perl.
979 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
980 output them only if debugging. By default they are written stderr,
981 apparently, which is not desirable.
983 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
986 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
989 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
991 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
992 authenticated or an ident call has been made. Suppress the default
993 values for $authenticated_id and $authenticated_sender (but permit -oMai
994 and -oMas) when testing with -bh.
996 PH/24 Re-jigged the order of the tests in the default configuration so that the
997 tests for valid domains and recipients precede the DNS black list and CSA
998 tests, on the grounds that those ones are more expensive.
1000 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1001 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1002 This bug exists in every version of Exim that I still have, right back to
1005 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1006 However, an attempt to turn on case-sensitivity in a regex key by
1007 including (?-i) didn't work because the subject string was already
1008 lowercased, and the effects were non-intuitive. It turns out that a
1009 one-line patch can be used to allow (?-i) to work as expected.
1015 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1016 other effects) broke the use of negated acl sub-conditions.
1018 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1021 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1022 "Deny" causes Exim to reject the incoming connection with a 554 error.
1023 Unfortunately, if there is a major crisis, such as a disk failure,
1024 tcp-wrappers gives "deny", whereas what one would like would be some
1025 kind of temporary error. A kludge has been added to help with this.
1026 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1027 554 error is used if errno is still zero or contains ENOENT (which occurs
1028 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1031 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1033 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1034 errors. Otherwise a message that provokes a temporary error (when other
1035 messages do not) can cause a whole host to time out.
1037 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1038 addresses were routed directly to files or pipes from a redirect router.
1039 File deliveries just didn't batch; pipe deliveries might have suffered
1042 PH/06 A failure to get a lock for a hints database would erroneously always say
1043 "Failed to get write lock", even when it was really a read lock.
1045 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1046 of 0600. This has been changed to use the value of the lockfile_mode
1047 option (which defaults to 0600).
1049 PH/08 Applied small patch from the Sieve maintainer.
1051 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1052 folder from quota calculations, a direct delivery into this folder messed
1053 up the contents of the maildirsize file. This was because the regex was
1054 used only to exclude .Trash (or whatever) when the size of the mailbox
1055 was calculated. There was no check that a delivery was happening into an
1056 excluded directory. This bug has been fixed by ignoring all quota
1057 processing for deliveries into excluded directories.
1059 PH/10 Added the maildirfolder_create_regex option to appendfile.
1065 PH/01 The code for finding all the local interface addresses on a FreeBSD
1066 system running IPv6 was broken. This may well have applied to all BSD
1067 systems, as well as to others that have similar system calls. The broken
1068 code found IPv4 interfaces correctly, but gave incorrect values for the
1069 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1070 that it would not match correctly against @[] and not recognize the IPv6
1073 PH/02 The ipliteral router was not recognizing addresses of the form user@
1074 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1076 PH/03 Added disable_ipv6.
1078 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1079 decoded form, because it is most often used to construct To: headers
1080 lines in autoreplies, and the decoded form may well be syntactically
1081 invalid. However, $reply_address has leading white space removed, and all
1082 newlines turned into spaces so that the autoreply transport does not
1085 PH/05 If group was specified without a user on a router, and no group or user
1086 was specified on a transport, the group from the router was ignored.
1088 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1089 for visible compile-time settings that can be used to change these
1090 numbers, for those that want even more. Backwards compatibility with old
1091 spool files has been maintained. However, going back to a previous Exim
1092 release will lost any variables that are in spool files.
1094 PH/07 Two small changes when running in the test harness: increase delay when
1095 passing a TCP/IP connection to a new process, in case the original
1096 process has to generate a bounce, and remove special handling of
1097 127.0.0.2 (sic), which is no longer necessary.
1099 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1100 be the same on different OS.
1102 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1105 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1106 whether --show-vars was specified or not
1108 JJ/02 exipick: Added support for new ACL variable spool format introduced
1111 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1112 syntactically invalid From: or Reply-to: line, and a filter used this to
1113 generate an autoreply, and therefore failed to obtain an address for the
1114 autoreply, Exim could try to deliver to a non-existent relative file
1115 name, causing unrelated and misleading errors. What now happens is that
1116 it logs this as a hard delivery error, but does not attempt to create a
1119 PH/11 The exinext utility has a -C option for testing purposes, but although
1120 the given file was scanned by exinext itself; it wasn't being passed on
1121 when Exim was called.
1123 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1124 an end-of-file indication when reading a command response.
1126 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1127 compiled. In many other places in Exim, IPv6 addresses are always
1128 recognized, so I have changed this. It also means that IPv4 domain
1129 literals of the form [IPV4:n.n.n.n] are now always recognized.
1131 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1132 used if the router is not running as root, for example, when verifying at
1133 ACL time, or when using -bh. The debugging output from this situation was
1134 non-existent - all you got was a failure to exec. I have made two
1137 (a) Failures to set uid/gid, the current directory, or a process leader
1138 in a subprocess such as that created by queryprogram now generate
1139 suitable debugging ouput when -d is set.
1141 (b) The queryprogram router detects when it is not running as root,
1142 outputs suitable debugging information if -d is set, and then runs
1143 the subprocess without attempting to change uid/gid.
1145 PH/15 Minor change to Makefile for building test_host (undocumented testing
1148 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1149 additional section of a DNS packet that returns MX or SRV records.
1150 Instead, it always explicitly searches for A/AAAA records. This avoids
1151 major problems that occur when a DNS server includes only records of one
1152 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1153 fixed another bug: if SRV records were looked up and the corresponding
1154 address records were *not* found in the additional section, the port
1155 values from the SRV records were lost.
1157 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1158 using the correct key (the original address) when searching the retry
1159 rules in order to find which one to use for generating the retry hint.
1161 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1162 adding the default one. Similarly, if it contains a Reply-To: header, the
1163 errors_reply_to option, if set, is not used.
1165 PH/19 When calculating a retry time, Exim used to measure the "time since
1166 failure" by looking at the "first failed" field in the retry record. Now
1167 it does not use this if it is later than than the arrival time of the
1168 message. Instead it uses the arrival time. This makes for better
1169 behaviour in cases where some deliveries succeed, thus re-setting the
1170 "first failed" field. An example is a quota failure for a huge message
1171 when small messages continue to be delivered. Without this change, the
1172 "time since failure" will always be short, possible causing more frequent
1173 delivery attempts for the huge message than are intended.
1174 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1176 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1177 $1, $2, $3) because the numerical variables can be reset during some
1178 expansion items (e.g. "match"), thereby losing the authentication data.
1180 PH/21 Make -bV show the size of off_t variables so that the test suite can
1181 decide whether to run tests for quotas > 2G.
1183 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1184 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1185 filecount value is greater than 2G or if a quota value is greater than 2G
1186 on a system where the size of off_t is not greater than 4, a panic error
1189 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1190 never match. The debug and -bh output now contains an explicit error
1191 message indicating a malformed IPv4 address or mask.
1193 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1194 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1195 PH/23 above applies.
1197 PH/25 Do not write to syslog when running in the test harness. The only
1198 occasion when this arises is a failure to open the main or panic logs
1199 (for which there is an explicit test).
1201 PH/26 Added the /no_tell option to "control=freeze".
1203 PH/27 If a host name lookup failed very early in a connection, for example, if
1204 the IP address matched host_lookup and the reverse lookup yielded a name
1205 that did not have a forward lookup, an error message of the form "no IP
1206 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1207 could be logged. Now it outputs the IP address instead of "NULL".
1209 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1210 allows the sender and the authenticated sender to be set when
1211 submitting a message from within Exim. Since child_open_exim() is
1212 documented for local_scan(), the new function should be too.
1214 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1215 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1216 results in an empty string is now treated as unset.
1218 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1220 PH/31 Added sender_verify_fail logging option.
1222 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1223 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1224 tidied the source and removed it altogether.
1226 PH/33 When a queue run was abandoned because the load average was too high, a
1227 log line was always written; now it is written only if the queue_run log
1228 selector is set. In addition, the log line for abandonment now contains
1229 information about the queue run such as the pid. This is always present
1230 in "start" and "stop" lines but was omitted from the "abandon" line.
1232 PH/34 Omit spaces between a header name and the colon in the error message that
1233 is given when verify = headers_syntax fails (if there are lots of them,
1234 the message gets confusing).
1236 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1237 names, as there are now some PTR records that contain slashes. This check
1238 is only to protect against broken name servers that fall over on strange
1239 characters, so the fact that it applies to all lookups doesn't matter.
1241 PH/36 Now that the new test suite is complete, we can remove some of the
1242 special code in Exim that was needed for the old test suite. For example,
1243 sorting DNS records because real resolvers return them in an arbitrary
1244 order. The new test suite's fake resolver always returns records in the
1247 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1248 bounces) except when queue_only is set, to avoid logging races between
1249 the different processes.
1251 PH/38 Panic-die if .include specifies a non-absolute path.
1253 PH/39 A tweak to the "H" retry rule from its user.
1255 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1256 a label. They prevented compilation on older perls.
1258 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1259 a warning to be raised on newish perls.
1261 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1262 on queue. Changes to match documented behaviour of showing count of
1263 messages matching specified criteria.
1265 PH/40 Changed the default ident timeout from 30s to 5s.
1267 PH/41 Added support for the use of login_cap features, on those BSD systems
1268 that have them, for controlling the resources used by pipe deliveries.
1270 PH/42 The content-scanning code uses fopen() to create files in which to put
1271 message data. Previously it was not paying any attention to the mode of
1272 the files. Exim runs with umask(0) because the rest of the code creates
1273 files with open(), and sets the required mode explicitly. Thus, these
1274 files were ending up world-writeable. This was not a big issue, because,
1275 being within the spool directory, they were not world-accessible. I have
1276 created a function called modefopen, which takes an additional mode
1277 argument. It sets umask(777), creates the file, chmods it to the required
1278 mode, then resets the umask. All the relevant calls to fopen() in the
1279 content scanning code have been changed to use this function.
1281 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1282 to 24 hours. This avoids potential overflow problems when processing G
1283 and H retry rules. I suspect nobody ever tinkers with this value.
1285 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1287 PH/45 When the plaintext authenticator is running as a client, the server's
1288 challenges are checked to ensure they are valid base64 strings. By
1289 default, the authentication attempt is cancelled if an invalid string is
1290 received. Setting client_ignore_invalid_base64 true ignores these errors.
1291 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1292 they are received. Thus, the responses can be made to depend on the
1293 challenges. If an invalid string is ignored, an empty string is placed in
1296 PH/46 Messages that are created by the autoreply transport now contains a
1297 References: header, in accordance with RFCs 2822 and 3834.
1299 PH/47 Added authenticated_sender_force to the smtp transport.
1301 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1303 PH/49 Installed latest patch from the Sieve maintainer.
1305 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1306 was also set, the appendfile transport was unnecessarily scanning a
1307 directory of message files (e.g. for maildir delivery) to find the count
1308 of files (along with the size), even though it did not need this
1309 information. It now does the scan only if it needs to find either the
1310 size of the count of files.
1312 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1314 PH/52 Two bugs concerned with error handling when the smtp transport is
1317 (i) Exim was not creating retry information for temporary errors given
1318 for individual recipients after the DATA command when the smtp transport
1319 was used in LMTP mode. This meant that they could be retried too
1320 frequently, and not timed out correctly.
1322 (ii) Exim was setting the flag that allows error details to be returned
1323 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1324 recipients that were returned after the DATA command.
1326 PH/53 This is related to PH/52, but is more general: for any failing address,
1327 when detailed error information was permitted to be returned to the
1328 sender, but the error was temporary, then after the final timeout, only
1329 "retry timeout exceeded" was returned. Now it returns the full error as
1330 well as "retry timeout exceeded".
1332 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1333 do this, and (what is worse) MTAs that accept it.
1335 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1336 will now be deprecated.
1338 PH/56 New os.c-cygwin from the Cygwin maintainer.
1340 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1341 formats (previously only available in exim formats via -bpr, -bpru,
1342 and -bpra. Now also available in native and exiqgrep formats)
1344 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1345 with very large, slow to parse queues
1347 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1349 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1351 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1352 responses to authentication challenges, though it was showing the
1353 challenges; (ii) I've removed the CR characters from the debug output for
1356 PH/58 Allow for the insertion of a newline as well as a space when a string
1357 is turned into more than one encoded-word during RFC 2047 encoding. The
1358 Sieve code now uses this.
1360 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1361 data_4xx, lost_connection, tls_required.
1363 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1364 message was being sent as an SMTP response.
1366 PH/61 Add -l and -k options to exicyclog.
1368 PH/62 When verifying, if an address was redirected to one new address, so that
1369 verification continued, and the new address failed or deferred after
1370 having set something in $address_data, the value of $address_data was not
1371 passed back to the ACL. This was different to the case when no
1372 redirection occurred. The value is now passed back in both cases.
1374 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1375 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1376 use login_cap.h, so on its own it isn't the distinguishing feature. The
1377 new name refers directly to the setclassresources() function.
1379 PH/65 Added configuration files for NetBSD3.
1381 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1383 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1384 is preferred over IPv4.
1386 PH/68 The bounce_return_message and bounce_return_body options were not being
1387 honoured for bounces generated during the reception of non-SMTP messages.
1388 In particular, this applied to messages rejected by the ACL. This bug has
1389 been fixed. However, if bounce_return_message is true and bounce_return_
1390 body is false, the headers that are returned for a non-SMTP message
1391 include only those that have been read before the error was detected.
1392 (In the case of an ACL rejection, they have all been read.)
1394 PH/69 The HTML version of the specification is now built in a directory called
1395 spec_html instead of spec.html, because the latter looks like a path with
1396 a MIME-type, and this confuses some software.
1398 PH/70 Catch two compiler warnings in sieve.c.
1400 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1401 function verify_get_ident() calls ip_connect() to connect a socket, but
1402 if the "connect()" function timed out, ip_connect() used to close the
1403 socket. However, verify_get_ident() also closes the socket later, and in
1404 between Exim writes to the log, which may get opened at this point. When
1405 the socket was closed in ip_connect(), the log could get the same file
1406 descriptor number as the socket. This naturally causes chaos. The fix is
1407 not to close the socket in ip_connect(); the socket should be closed by
1408 the function that creates it. There was only one place in the code where
1409 this was missing, in the iplookup router, which I don't think anybody now
1410 uses, but I've fixed it anyway.
1412 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1413 well as to direct DNS lookups. Otherwise the handling of names in host
1414 lists is inconsistent and therefore confusing.
1420 PH/01 Two changes to the default runtime configuration:
1422 (1) Move the checks for relay_from_hosts and authenticated clients from
1423 after to before the (commented out) DNS black list checks.
1425 (2) Add control=submission to the relay_from_hosts and authenticated
1426 clients checks, on the grounds that messages accepted by these
1427 statements are most likely to be submissions.
1429 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1431 (1) Generate an error if the third argument for the ${prvs expansion is
1434 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1437 (3) Reset the variables that are obtained from the first argument of
1438 ${prvscheck and used in the second argument before leaving the code,
1439 because their memory is reclaimed, so using them afterwards may do
1442 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1443 one (it's much easier than Tom thought :-).
1445 (5) Because of (4), we can now allow for the use of $prvscheck_result
1446 inside the third argument.
1448 PH/03 For some reason, the default setting of PATH when running a command from
1449 a pipe transport was just "/usr/bin". I have changed it to
1452 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1453 anything to be listed in the output from -bV.
1455 PH/05 When a filter generated an autoreply, the entire To: header line was
1456 quoted in the delivery log line, like this:
1458 => >A.N.Other <ano@some.domain> <original@ddress> ...
1460 This has been changed so that it extracts the operative address. There
1461 may be more than one such address. If so, they are comma-separated, like
1464 => >ano@some.domain,ona@other.domain <original@ddress> ...
1466 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1467 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1468 not being looked up in the rDNS to get a host name, Exim was showing the
1469 IP address twice in Received: lines, even though the IP addresses were
1470 identical. For example:
1472 Received: from [1.2.3.4] (helo=[1.2.3.4])
1474 However, if the real host name was known, it was omitting the HELO data
1475 if it matched the actual IP address. This has been tidied up so that it
1476 doesn't show the same IP address twice.
1478 PH/07 When both +timestamp and +memory debugging was on, the value given by
1479 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1480 called by the debug printing, thereby overwriting the timestamp buffer.
1481 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1483 PH/08 When the original message was included in an autoreply transport, it
1484 always said "this is a copy of the message, including all the headers",
1485 even if body_only or headers_only was set. It now gives an appropriate
1488 PH/09 Applied a patch from the Sieve maintainer which:
1490 o fixes some comments
1491 o adds the (disabled) notify extension core
1492 o adds some debug output for the result of if/elsif tests
1493 o points to the current vacation draft in the documentation
1494 and documents the missing references header update
1498 o fixes a bug in processing the envelope test (when testing
1499 multiple envelope elements, the last element determinted the
1502 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1503 Electronic Mail") by including:
1505 Auto-submitted: auto-generated
1507 in the messages that it generates (bounce messages and others, such as
1508 warnings). In the case of bounce messages for non-SMTP mesages, there was
1509 also a typo: it was using "Auto_submitted" (underscore instead of
1510 hyphen). Since every message generated by Exim is necessarily in response
1511 to another message, thes have all been changed to:
1513 Auto-Submitted: auto-replied
1515 in accordance with these statements in the RFC:
1517 The auto-replied keyword:
1519 - SHOULD be used on messages sent in direct response to another
1520 message by an automatic process,
1522 - MUST NOT be used on manually-generated messages,
1524 - MAY be used on Delivery Status Notifications (DSNs) and Message
1525 Disposition Notifications (MDNs),
1527 - MUST NOT be used on messages generated by automatic or periodic
1528 processes, except for messages which are automatic responses to
1531 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1532 to the default Received: header definition.
1534 PH/12 Added log selector acl_warn_skipped (default on).
1536 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1537 variables because (a) they are in the wrong storage pool and (b) even if
1538 they were copied, it wouldn't work properly because of the caching.
1540 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1541 checking when decoding. Apparently there are clients that generate
1542 overlong encoded strings. Why am I not surprised?
1544 PH/15 If the first argument of "${if match_address" was not empty, but did not
1545 contain an "@" character, Exim crashed. Now it writes a panic log message
1546 and treats the condition as false.
1548 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1550 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1551 extension "envelope-auth". The code is finished and in agreement with
1552 other implementations, but there is no documentation so far and in fact,
1553 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1554 not changing the active code.
1556 Print executed "if" and "elsif" statements when debugging is used. This
1557 helps a great deal to understand what a filter does.
1559 Document more things not specified clearly in RFC3028. I had all this
1560 sorted out, when out of a sudden new issues came to my mind. Oops."
1562 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1565 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1566 canonical form (as documented). However, after a host name lookup from
1567 the IP address, check_host() was doing a simple string comparison with
1568 addresses acquired from the DNS when checking that the found name did
1569 have the original IP as one of its addresses. Since any found IPv6
1570 addresses are likely to be in abbreviated form, the comparison could
1571 fail. Luckily, there already exists a function for doing the comparison
1572 by converting both addresses to binary, so now that is used instead of
1573 the text comparison.
1575 PH/20 There was another similar case to PH/19, when a complete host name was
1576 given in a host list; looking up its IP address could give an abbreviated
1577 form, whereas the current host's name might or might not be abbreviated.
1578 The same fix has been applied.
1584 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1585 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1588 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1589 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1591 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1593 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1594 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1595 Also, for "Auto-Submitted", treat anything other than "no" as
1596 non-personal, in accordance with RFC 3834. (Previously it treated
1597 anything starting "auto-" as non-personal.)
1599 TF/01 The control=submission/name=... option had a problem with syntax
1600 errors if the name included a slash character. The /name= option
1601 now slurps the rest of the string, so it can include any characters
1602 but it must come last in the list of options (after /sender_retain
1605 PH/05 Some modifications to the interface to the fake nameserver for the new
1613 TK/01 Added the "success_on_redirect" address verification option. See
1614 NewStuff for rationale and an example.
1616 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1618 PH/02 Patch to exigrep to allow it to work on syslog lines.
1620 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1621 fread() to skip over the body file's header line, because in Cygwin the
1622 header line is locked and is inaccessible.
1624 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1625 co-exist for some time) to make it clear that it is the Exim ID that is
1626 referenced, not the Message-ID: header line.
1628 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1629 string_format() function, because snprintf() does not exist on all
1632 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1633 ${stat: expansion item.
1635 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1636 protocol synchronization error", to keep the pedants happy.
1638 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1639 well as for IRIX systems, when gcc is being used. See the host.c source
1642 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1644 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1647 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1648 transport and to the smtp transport in LMTP mode.
1650 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1652 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1653 run a filter in a subprocess. This could lead to confusion in subsequent
1654 lookups in the parent process. There should also be a search_tidyup() at
1655 the end of the subprocess.
1657 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1658 only if the host matched helo_try_verify_hosts, which caused the
1659 verification to occur when the EHLO/HELO command was issued. The ACL just
1660 tested the remembered result. Now, if a previous verification attempt has
1661 not happened, "verify = helo" does it there and then.
1663 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1665 TK/03 Fix log output including CR from clamd.
1667 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1668 exist provoked a memory error which could cause a segfault.
1670 PH/15 Installed PCRE 6.2
1672 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1674 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1675 of the problem. Specifically, suggested +O2 rather than +O1 for the
1678 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1680 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1681 record, the port was not correctly logged when the outgoing_port log
1682 selector was set (it logged the transort's default port).
1684 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1685 fallback_hosts, and "hosts" in the smtp transport.
1687 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1688 host errors such as "Connection refused".
1690 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1691 authentication with radiusclient 0.4.9:
1693 - Error returned from rc_read_config was caught wrongly
1694 - Username/password not passed on to radius server due to wrong length.
1696 The presumption is that some radiusclient API changes for 4.51/PH/17
1697 were not taken care of correctly. The code is still untested by me (my
1698 Linux distribution still has 0.3.2 of radiusclient), but it was
1699 contributed by a Radius user.
1701 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1702 expanding the "port" option of the smtp transport.
1704 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1705 while reading a MIME header. Thanks to Tom Hughes for a patch.
1707 PH/24 Include config.h inside local_scan.h so that configuration settings are
1710 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1711 that in an ACL for RCPT (for example), you can examine exactly what was
1714 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1715 commands, but it was not correctly comparing the address with the actual
1716 client host address. Thus, it would show the EHLO address in Received:
1717 header lines when this was not necessary.
1719 PH/27 Added the % operator to ${eval:}.
1721 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1722 it should be ignoring failures (because with -C, for example, it has lost
1723 privilege). It wasn't ignoring creation failures other than "already
1726 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1729 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1730 because an input line was too long, either on its own, or by virtue of
1731 too many continuations, the temporary file was not being removed, and the
1732 return code was incorrect.
1734 PH/31 Missing "BOOL" in function definition in filtertest.c.
1736 PH/32 Applied Sieve patches from the maintainer.
1738 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1740 PH/33 Added "verify = not_blind".
1742 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1743 Local/Makefile (with some defaults set). These are used in built scripts
1744 such as exicyclog, but they have never been used in the exim_install
1745 script (though there are many overriding facilities there). I have
1746 arranged that the exim_install script now takes note of these two
1749 PH/35 Installed configuration files for Dragonfly.
1751 PH/36 When a locally submitted message by a trusted user did not contain a
1752 From: header, and the sender address was obtained from -f or from an SMTP
1753 MAIL command, and the trusted user did not use -F to supply a sender
1754 name, $originator_name was incorrectly used when constructing a From:
1755 header. Furthermore, $originator_name was used for submission mode
1756 messages from external hosts without From: headers in a similar way,
1757 which is clearly wrong.
1759 PH/37 Added control=suppress_local_fixups.
1761 PH/38 When log_selector = +received_sender was set, and the addition of the
1762 sender made the log line's construction buffer exactly full, or one byte
1763 less than full, an overflow happened when the terminating "\n" was
1766 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1767 when the result of a list match is failure because a DNS lookup failed.
1769 PH/40 RM_COMMAND is now used in the building process.
1771 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1772 the "build-* directories that it finds.
1774 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1775 address in a domain literal was a prefix of an interface address.
1777 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1778 when verifying a sender address, unless rewrite_headers is false.
1780 PH/44 Wrote a long comment about why errors_to addresses are verified as
1781 recipients, not senders.
1783 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1784 the ratelimit ACL was added.
1786 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1788 PH/46 Added extra information about PostgreSQL errors to the error string.
1790 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1791 avoiding the need to install special zones in a real server. This is
1792 backwards compatible; if it can't find the fake resolver, it drops back.
1793 Thus, both old and new test suites can be run.
1795 TF/02 Added util/ratelimit.pl
1797 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1798 clock is set back in time.
1800 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1801 Candler <B.Candler@pobox.com>.
1803 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1804 OK for addresses that are the result of redirections.
1806 PH/48 A number of further additions for the benefit of the new test suite,
1807 including a fake gethostbyname() that interfaces to the fake DNS resolver
1810 TF/06 The fix for widen_domains has also been applied to qualify_single and
1811 search_parents which are the other dnslookup options that can cause
1814 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1817 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1819 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1820 service" if it encountered an MX record with an empty target hostname.
1821 The message is now "an MX or SRV record indicated no SMTP service".
1823 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1824 if the DNS of the sending site is misconfigured. This is quite a
1825 common situation. This change restores the behaviour of treating a
1826 helo verification defer as a failure.
1828 PH/51 If self=fail was set on a router, the bounce message did not include the
1829 actual error message.
1835 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1837 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1838 command itself ended in error, the underlying message about the transport
1839 filter timeout was being overwritten with the pipe command error. Now the
1840 underlying error message should be appended to the second error message.
1842 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1844 PH/02 Reduce the amount of output that "make" produces by default. Full output
1845 can still be requested.
1847 PH/03 The warning log line about a condition test deferring for a "warn" verb
1848 was being output only once per connection, rather than after each
1849 occurrence (because it was using the same function as for successful
1850 "warn" verbs). This seems wrong, so I have changed it.
1852 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1853 it should not have, which might have caused a crash in the right
1854 circumstances, but probably never did.
1856 PH/04 Installed a modified version of Tony Finch's patch to make submission
1857 mode fix the return path as well as the Sender: header line, and to
1858 add a /name= option so that you can make the user's friendly name appear
1861 TF/03 Added the control = fakedefer ACL modifier.
1863 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1864 Mark Lowes for thorough testing.
1866 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1868 TK/03 Merged latest SRS patch from Miles Wilton.
1870 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1871 with the definition in sysexits.h (which is #included earlier).
1872 Fortunately, Exim does not actually use EX_OK. The code used to try to
1873 preserve the sysexits.h value, by assumimg that macro definitions were
1874 scanned for macro replacements. I have been disabused of this notion,
1875 so now the code just undefines EX_OK before #including unistd.h.
1877 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1878 in the smtp transport. When a block could not be written in a single
1879 write() function, the timeout was being re-applied to each part-write.
1880 This seems wrong - if the receiver was accepting one byte at a time it
1881 would take for ever. The timeout is now adjusted when this happens. It
1882 doesn't have to be particularly precise.
1884 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1885 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1887 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1888 without a domain if the check to <postmaster@domain> fails.
1890 SC/01 Eximstats: added -xls and the ability to specify output files
1891 (patch written by Frank Heydlauf).
1893 SC/02 Eximstats: use FileHandles for outputing results.
1895 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1897 SC/04 Eximstats: fixed display of large numbers with -nvr option
1899 SC/05 Eximstats: fixed merging of reports with empty tables.
1901 SC/06 Eximstats: added the -include_original_destination flag
1903 SC/07 Eximstats: removed tabs and trailing whitespace.
1905 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1907 TK/06 MBOX spool code: Add real "From " MBOX separator line
1908 so the .eml file is really in mbox format (even though
1909 most programs do not really care). Patch from Alex Miller.
1911 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1912 The latter is generated from $received_to and is only set if the
1913 message has one envelope recipient. SA can use these headers,
1914 obviously out-of-the-box. Patch from Alex Miller.
1916 PH/08 The ${def test on a variable was returning false if the variable's
1917 value was "0", contrary to what the specification has always said!
1918 The result should be true unless the variable is empty.
1920 PH/09 The syntax error of a character other than { following "${if
1921 def:variable_name" (after optional whitespace) was not being diagnosed.
1922 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1923 accidental colon was present, for example, could give incorrect results.
1925 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1926 result is used (not including appendfile, where other changes are about
1929 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1930 This involved changing a lot of size variables from int to off_t. It
1931 should work with maildirs and everything.
1933 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1934 spamd dying while we are connected to it.
1936 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1939 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1940 function for BDB 4.3.
1942 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1944 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1945 the sources for PCRE have been re-arranged and more files are now
1948 PH/15 The code I had for printing potentially long long variables in PH/11
1949 above was not the best (it lost precision). The length of off_t variables
1950 is now inspected at build time, and an appropriate printing format (%ld
1951 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1952 to be "long long int" or "long int". This is needed for the internal
1953 formatting function string_vformat().
1955 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1956 the configuration file to be ":syslog", then the script "guesses" where
1957 the logs files are, rather than using the compiled in default. In our
1958 case the guess is not the same as the compiled default, so the script
1959 suddenly stopped working when I started to use syslog. The patch checks
1960 to see if log_file_path is "". If so, it attempts to read it from exim
1961 with no configuration file to get the compiled in version, before it
1962 falls back to the previous guessing code."
1964 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1965 implementing BATV in an Exim configuration. See NewStuff for the gory
1968 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1969 Makefile that are specific to HP-UX.
1971 PH/18 If the "use_postmaster" option was set for a recipient callout together
1972 with the "random" option, the postmaster address was used as the MAIL
1973 FROM address for the random test, but not for the subsequent recipient
1974 test. It is now used for both.
1976 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1977 patch removes a few documentation additions to RFC 3028, because the
1978 latest draft now contains them. It adds the new en;ascii-case comparator
1979 and a new error check for 8bit text in MIME parts. Comparator and
1980 require names are now matched exactly. I enabled the subaddress
1981 extension, but it is not well tested yet (read: it works for me)."
1983 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1984 rework some of the code of TK/09 above to avoid the hardwired use of
1985 "%lld" and "long long". Replaced the call to snprintf() with a call to
1988 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1989 records point to non-existent hosts", "retry timeout exceeded", and
1990 "retry time not reached for any host after a long failure period".
1992 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1993 experimental DomainKeys support:
1995 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1996 (2) On an error such as an illegally used "control", the wrong name for
1997 the control was given.
1999 These problems did NOT occur unless DomainKeys support was compiled.
2001 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2003 PH/24 Added ${if match_ip condition.
2005 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2006 databases so that it will be absolutely obvious if a crash occurs in the
2007 DB library. This is a regular occurrence (often caused by mis-matched
2010 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2011 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2012 picked up on a user's system that detects such things. There doesn't seem
2013 to be a gcc warning option for this - only an attribute that has to be
2014 put on the function's prototype. It seems that in Fedora Core 4 they have
2015 set this on a number of new functions. No doubt there will be more in due
2018 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2019 specify a transport. However, if an address that was verified by such a
2020 router was the subject of a callout, Exim crashed because it tried to
2021 read the rcpt_include_affixes from the non-existent transport. Now it
2022 just assumes that the setting of that option is false. This bug was
2023 introduced by 4.51/PH/31.
2025 PH/28 Changed -d+all to exclude +memory, because that information is very
2026 rarely of interest, but it makes the output a lot bigger. People tend to
2027 do -d+all out of habit.
2029 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2030 code in os-type was giving problems when libc.so lives in lib64, like on
2033 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2034 aren't the modern standard, and it seems that some systems' include files
2035 don't always have them. Exim was already checking for some of the newer
2036 ones like T_AAAA, and defining it itself. I've added checks for all the
2037 record types that Exim uses.
2039 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2040 not automatically generating a new one, as it is supposed to. This
2041 prevented TLS from working. If the file did exist, but contained invalid
2042 data, a new version was generated, as expected. It was only the case of a
2043 non-existent file that was broken.
2045 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2046 with a change in libdomainkeys > 0.64.
2048 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2049 from DNS. If the selector record carries the flag, it now has
2050 precedence over the domain-wide flag.
2052 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2054 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2055 the use of a port name that isn't defined in /etc/services) occurred, the
2056 message was deferred as in a normal delivery, and thus remained on the
2057 spool, instead of being failed because of the mua_wrapper setting. This
2058 is now fixed, and I tidied up some of the mua_wrapper messages at the
2061 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2062 the messages in a hash of arrays rather than using individual hashes.
2063 This is a bit cleaner and results in dramatic memory savings, albeit
2064 at a slight CPU cost.
2066 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2067 as requested by Marc Sherman.
2069 SC/10 Eximstats: added histograms for user specified patterns as requested
2072 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2074 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2075 fopen() in the content-scanning modules that did not already have it.
2081 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2082 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2084 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2086 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2088 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2089 to test Sieve filters that use "vacation".
2091 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2092 that changes the way the GnuTLS parameters are stored in the cache file.
2093 The new format can be generated externally. For backward compatibility,
2094 if the data in the cache doesn't make sense, Exim assumes it has read an
2095 old-format file, and it generates new data and writes a new file. This
2096 means that you can't go back to an older release without removing the
2099 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2100 work if there are any delivery delays because "one_time" forces the
2101 parent to be marked "delivered", so its unseen clone is never tried
2102 again. For this reason, Exim now forbids the simultaneous setting of
2105 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2106 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2107 entry if you want to know the details. The fix, however, neglected to
2108 consider the case when local delivery batching is involved. The test for
2109 "previously delivered" was not happening when checking to see if an
2110 address could be batched with a previous (undelivered) one; under
2111 certain circumstances this could lead to multiple deliveries to the same
2114 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2115 in its include files, and this causes problems building Exim.
2117 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2118 header_syntax) but Exim was just ignoring anything given after a slash.
2119 In particular, this caused confusion with an attempt to use "verify =
2120 reverse_host_lookup/defer_ok". An error is now given when options are
2121 supplied for verify items that do not have them. (Maybe reverse_host_
2122 lookup should have a defer_ok option, but that's a different point.)
2124 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2125 defined by RFC 821) to 2048, because there were problems with some AUTH
2126 commands, and RFC 1869 says the size should be increased for extended
2127 SMTP commands that take arguments.
2129 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2132 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2133 "unknown" error; now it says that the functionality isn't in the binary.
2135 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2136 an address' error message when a string expansion fails (syntax or
2137 whatever). Otherwise the password may appear in the log. Following change
2138 PH/42 below, there is no longer a chance of it appearing in a bounce
2141 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2143 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2144 of its addresses was ever tried. (Bugzilla bug #2.)
2146 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2147 the result incorrectly in the debug output. (It correctly added a newline
2148 to what was transported.)
2150 TF/01 Added $received_time.
2152 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2153 commented out examples of how to interface to a virus scanner and to
2154 SpamAssassin. Also added commented examples of av_scanner and
2155 spamd_address settings.
2157 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2158 and controls are allowed in which ACLs. There were a couple of minor
2159 errors. Some of the entries in the conditions table (which is a table of
2160 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2161 negation of where the condition IS allowed.
2163 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2165 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2166 header file does not have a version number, so I've had to invent a new
2167 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2168 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2169 radiusclient), but it was contributed by a Radius user.
2171 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2172 files or directories, for OpenSSL.
2174 PH/19 When an Exim process that is running as root has to create an Exim log
2175 file, it does so in a subprocess that runs as exim:exim so as to get the
2176 ownership right at creation (otherwise, other Exim processes might see
2177 the file with the wrong ownership). There was no test for failure of this
2178 fork() call, which would lead to the process getting stuck as it waited
2179 for a non-existent subprocess. Forks do occasionally fail when resources
2180 run out. I reviewed all the other calls to fork(); they all seem to check
2183 PH/20 When checking for unexpected SMTP input at connect time (before writing
2184 the banner), Exim was not dealing correctly with a non-positive return
2185 from the read() function. If the client had disconnected by this time,
2186 the result was a log entry for a synchronization error with an empty
2187 string after "input=" when read() returned zero. If read() returned -1
2188 (an event I could not check), uninitialized data bytes were printed.
2189 There were reports of junk text (parts of files, etc) appearing after
2192 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2194 PH/22 Added support for macro redefinition, and (re)definition in between
2195 driver and ACL definitions.
2197 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2198 forgetting to use the resulting value; it was using the unexpanded value.
2200 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2201 hadn't been configured. The fix is from Juergen Kreileder, who
2202 understands it better than I do:
2204 "Here's what I see happening with three configured cyrus_sasl
2205 authenticators configured (plain, login, cram-md5):
2207 On startup auth_cyrus_sasl_init() gets called for each of these.
2208 This means three calls to sasl_listmech() without a specified mech_list.
2209 => SASL tests which mechs of all available mechs actually work
2210 => three warnings about OTP not working
2211 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2213 With the patch, sasl_listmech() also gets called three times. But now
2214 SASL's mech_list option is set to the server_mech specified in the the
2215 authenticator. Or in other words, the answer from sasl_listmech()
2216 gets limited to just the mech you're testing for (which is different
2218 => the return list contains just 'plain' or 'login', 'cram-md5' or
2219 nothing depending on the value of ob->server_mech.
2221 I've just tested the patch: Authentication still works fine,
2222 unavailable mechs specified in the exim configuration are still
2223 caught, and the auth.log warnings about OTP are gone."
2225 PH/25 When debugging is enabled, the contents of the command line are added
2226 to the debugging output, even when log_selector=+arguments is not
2229 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2230 answer is "GNU", and only if the return is "GNU/something" is the answer
2233 PH/27 $acl_verify_message is now set immediately after the failure of a
2234 verification in an ACL, and so is available in subsequent modifiers. In
2235 particular, the message can be preserved by coding like this:
2237 warn !verify = sender
2238 set acl_m0 = $acl_verify_message
2240 Previously, $acl_verify_message was set only while expanding "message"
2241 and "log_message" when a very denied access.
2243 PH/28 Modified OS/os.c-Linux with
2245 -#ifndef OS_LOAD_AVERAGE
2246 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2248 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2249 nomenclature these days.)
2251 PH/29 Installed patch from the Sieve maintainer that adds the options
2252 sieve_useraddress and sieve_subaddress to the redirect router.
2254 PH/30 In these circumstances:
2255 . Two addresses routed to the same list of hosts;
2256 . First host does not offer TLS;
2257 . First host accepts first address;
2258 . First host gives temporary error to second address;
2259 . Second host offers TLS and a TLS session is established;
2260 . Second host accepts second address.
2261 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2262 and peerdn, if requested) that were in fact used only for the second
2265 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2266 attention to any local part prefix or suffix that was matched by the
2267 router that accepted the address. It now behaves in the same way as it
2268 does for delivery: the affixes are removed from the local part unless
2269 rcpt_include_affixes is set on the transport.
2271 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2272 timeout during the DATA phase of an incoming message.
2274 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2275 applied a patch sanctioned by the Sieve maintainer.
2277 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2278 the uid or gid is negative. A case of a negative gid caused this to be
2279 noticed. The fix allows for either to be negative.
2281 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2282 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2285 PH/36 The change PH/12 above was broken. Fixed it.
2287 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2288 the grounds that routing the same address twice would always produce the
2289 same answer. This might have been true once, but it is certainly no
2290 longer true now. Routing a child address may depend on the previous
2291 routing that produced that child. Some complicated redirection strategies
2292 went wrong when messages had multiple recipients, and made Exim's
2293 behaviour dependent on the order in which the addresses were given.
2295 I have moved the duplicate checking until after the routing is complete.
2296 Exim scans the addresses that are assigned to local and remote
2297 transports, and removes any duplicates. This means that more work will be
2298 done, as duplicates will always all be routed, but duplicates are
2299 presumably rare, so I don't expect this is of any significance.
2301 For deliveries to pipes, files, and autoreplies, the duplicate checking
2302 still happens during the routing process, since they are not going to be
2305 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2306 It corrects a timeout issue with spamd. This is Ian's comment: "The
2307 background is that sometimes spamd either never reads data from a
2308 connection it has accepted, or it never writes response data. The exiscan
2309 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2310 blindly assumes that writes won't block so it may never time out."
2312 PH/39 Allow G after quota size as well as K and M.
2314 PH/40 The value set for $authenticated_id in an authenticator may not contain
2315 binary zeroes or newlines because the value is written to log lines and
2316 to spool files. There was no check on this. Now the value is run through
2317 the string_printing() function so that such characters are converted to
2318 printable escape sequences.
2320 PH/41 $message_linecount is a new variable that contains the total number of
2321 lines in the message. Compare $body_linecount, which is the count for the
2324 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2325 bounce and delay warning messages, except in certain special cases, which
2328 (a) An SMTP error message from a remote host;
2329 (b) A message specified in a :fail: redirection;
2330 (c) A message specified in a "fail" command in a system filter;
2331 (d) A message specified in a FAIL return from the queryprogram router;
2332 (e) A message specified by the cannot_route_message router option.
2334 In these cases only, Exim does include the error details in bounce and
2335 warning messages. There are also a few cases where bland messages such
2336 as "unrouteable address" or "local delivery error" are given.
2338 PH/43 $value is now also set for the "else" part of a ${run expansion.
2340 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2341 being worked on, but at least Exim now implements the latest version to
2344 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2345 process to complete was treated as a delivery failure, a timeout while
2346 writing the message to the pipe was logged, but erroneously treated as a
2347 successful delivery. Such timeouts include transport filter timeouts. For
2348 consistency with the overall process timeout, these timeouts are now
2349 treated as errors, giving rise to delivery failures by default. However,
2350 there is now a new Boolean option for the pipe transport called
2351 timeout_defer, which, if set TRUE, converts the failures into defers for
2352 both kinds of timeout. A transport filter timeout is now identified in
2355 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2356 systems where "make" and "gmake" are different, calling "gmake" at top
2357 level broke things. I've arranged for the value of $(MAKE) to be passed
2358 from the Makefile to this script so that it can call the same version of
2362 A note about Exim versions 4.44 and 4.50
2363 ----------------------------------------
2365 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2366 changes of various kinds. As a consequence, a big documentation update was
2367 needed. This delayed the release for rather longer than seemed good, especially
2368 in the light of a couple of (minor) security issues. Therefore, the changes
2369 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2370 release. So 4.44 and 4.50 are in effect two different branches that both start
2373 I have left the 4.50 change log unchanged; it contains all the changes since
2374 4.43. The change log for 4.44 is below; many of its items are identical to
2375 those for 4.50. This seems to be the most sensible way to preserve the
2376 historical information.
2382 1. Minor wording change to the doc/README.SIEVE file.
2384 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2385 computation of the current number of files was incorrect.
2387 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2388 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2391 4. Give more explanation in the error message when the command for a transport
2392 filter fails to execute.
2394 5. There are several places where Exim runs a non-Exim command in a
2395 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2396 was being done only for the command run by the queryprogram router. It is
2397 now done for all such subprocesses. The other cases are: ${run, transport
2398 filters, and the commands run by the lmtp and pipe transports.
2400 6. Added CONFIGURE_GROUP build-time option.
2402 7. Some older OS have a limit of 256 on the maximum number of file
2403 descriptors. Exim was using setrlimit() to set 1000 as a large value
2404 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2405 systems. I've change it so that if it can't get 1000, it tries for 256.
2407 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2408 was an oversight, and furthermore, ever since the addition of extra
2409 controls (e.g. 4.43/32), the checks on when to allow different forms of
2410 "control" were broken. There should now be diagnostics for all cases when a
2411 control that does not make sense is encountered.
2413 9. Added the /retain_sender option to "control=submission".
2415 10. $recipients is now available in the predata ACL (oversight).
2417 11. Tidy the search cache before the fork to do a delivery from a message
2418 received from the command line. Otherwise the child will trigger a lookup
2419 failure and thereby defer the delivery if it tries to use (for example) a
2420 cached ldap connection that the parent has called unbind on.
2422 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2423 of $address_data from the recipient verification was clobbered by the
2424 sender verification.
2426 13. The value of address_data from a sender verification is now available in
2427 $sender_address_data in subsequent conditions in the ACL statement.
2429 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2431 15. Added a new option "connect=<time>" to callout options, to set a different
2434 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2435 was its contents. (It was OK if the option was not defined at all.)
2437 17. A "Completed" log line is now written for messages that are removed from
2438 the spool by the -Mrm option.
2440 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2441 information about exactly what failed.
2443 19. Added -dd to debug only the daemon process.
2445 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2446 handles timeouts, both on the server side and network timeouts. Renamed the
2447 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2449 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2450 It is now set to "smtps".
2452 22. $host_address is now set to the target address during the checking of
2453 ignore_target_hosts.
2455 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2456 being passed; this would have caused $sender_host_name to have been used if
2457 matching the list had actually called for a host name (not very likely,
2458 since this list is usually IP addresses). A host name is now passed as
2461 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2462 code that specifies a non-restarting handler (typically sigaction() in
2463 modern systems) in an attempt to fix a rare and obscure crash bug.
2465 25. Narrowed the window for a race in the daemon that could cause it to ignore
2466 SIGCHLD signals. This is not a major problem, because they are used only to
2467 wake it up if nothing else does.
2469 26. A malformed maildirsize file could cause Exim to calculate negative values
2470 for the mailbox size or file count. Odd effects could occur as a result.
2471 The maildirsize information is now recalculated if the size or filecount
2474 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2475 support for a long time. Removed HAVE_SYS_VFS_H.
2477 28. Installed the latest version of exipick from John Jetmore.
2479 29. In an address list, if the pattern was not a regular expression, an empty
2480 subject address (from a bounce message) matched only if the pattern was an
2481 empty string. Non-empty patterns were not even tested. This was the wrong
2482 because it is perfectly reasonable to use an empty address as part of a
2483 database query. An empty address is now tested by patterns that are
2484 lookups. However, all the other forms of pattern expect the subject to
2485 contain a local part and a domain, and therefore, for them, an empty
2486 address still always fails if the pattern is not itself empty.
2488 30. Exim went into a mad DNS loop when attempting to do a callout where the
2489 host was specified on an smtp transport, and looking it up yielded more
2490 than one IP address.
2492 31. Re-factored the code for checking spool and log partition space into a
2493 function that finds that data and another that does the check. The former
2494 is then used to implement four new variables: $spool_space, $log_space,
2495 $spool_inodes, and $log_inodes.
2497 32. The RFC2047 encoding function was originally intended for short strings
2498 such as real names; it was not keeping to the 75-character limit for
2499 encoded words that the RFC imposes. It now respects the limit, and
2500 generates multiple encoded words if necessary. To be on the safe side, I
2501 have increased the buffer size for the ${rfc2047: expansion operator from
2504 33. It is now permitted to omit both strings after an "if" condition; if the
2505 condition is true, the result is "true". As before, when the second string
2506 is omitted, a false condition yields an empty string. This makes it less
2507 cumbersome to write custom ACL and router conditions.
2509 34. Failure to deliver a bounce message always caused it to be frozen, even if
2510 there was an errors_to setting on the router. The errors_to setting is now
2513 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2514 canonical form (fully expanded) before being placed in
2515 $sender_host_address.
2517 36. The table in the code that translates DNS record types into text (T_A to
2518 "A" for instance) was missing entries for NS and CNAME. It is just possible
2519 that this could have caused confusion if both these types were looked up
2520 for the same domain, because the text type is used as part of Exim's
2521 per-process caching. But the chance of anyone hitting this buglet seems
2524 37. The dnsdb lookup has been extended in a number of ways.
2526 (1) There is a new type, "zns", which walks up the domain tree until it
2527 finds some nameserver records. It should be used with care.
2529 (2) There is a new type, "mxh", which is like "mx" except that it returns
2530 just the host names, not the priorities.
2532 (3) It is now possible to give a list of domains (or IP addresses) to be
2533 looked up. The behaviour when one of the lookups defers can be
2534 controlled by a keyword.
2536 (4) It is now possible to specify the separator character for use when
2537 multiple records are returned.
2539 38. The dnslists ACL condition has been extended: it is now possible to supply
2540 a list of IP addresses and/or domains to be looked up in a particular DNS
2543 39. Added log_selector=+queue_time_overall.
2545 40. When running the queue in the test harness, wait just a tad after forking a
2546 delivery process, to get repeatability of debugging output.
2548 41. Include certificate and key file names in error message when GnuTLS fails
2549 to set them up, because the GnuTLS error message doesn't include the name
2550 of the failing file when there is a problem reading it.
2552 42. Allow both -bf and -bF in the same test run.
2554 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2556 44. The "Exiscan patch" is now merged into the mainline Exim source.
2558 45. Sometimes the final signoff response after QUIT could fail to get
2559 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2560 < 0 before doing a fflush(). This bug looks as though it goes back to the
2561 introduction of TLS in release 3.20, but "sometimes" must have been rare
2562 because the tests only now provoked it.
2564 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2565 (this can affect the format of dates).
2567 47. exim_tidydb, when checking for the continued existence of a message for
2568 which it has found a message-specific retry record, was not finding
2569 messages that were in split spool directories. Consequently, it was
2570 deleting retry records that should have stayed in existence.
2572 48. Steve fixed some bugs in eximstats.
2574 49. The SPA authentication driver was not abandoning authentication and moving
2575 on to the next authenticator when an expansion was forced to fail,
2576 contradicting the general specification for all authenticators. Instead it
2577 was generating a temporary error. It now behaves as specified.
2579 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2580 (the order specifies the preference for clients). The order is now AES256,
2581 AES128, 3DES, ARCFOUR128.
2583 51. Small patch to Sieve code - explicitly set From: when generating an
2586 52. Exim crashed if a remote delivery caused a very long error message to be
2587 recorded - for instance if somebody sent an entire SpamAssassin report back
2588 as a large number of 550 error lines. This bug was coincidentally fixed by
2589 increasing the size of one of Exim's internal buffers (big_buffer) that
2590 happened as part of the Exiscan merge. However, to be on the safe side, I
2591 have made the code more robust (and fixed the comments that describe what
2594 53. Now that there can be additional text after "Completed" in log lines (if
2595 the queue_time_overall log selector is set), a one-byte patch to exigrep
2596 was needed to allow it to recognize "Completed" as not the last thing in
2599 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2600 patch that reportedly fixes this has been added. I am not expert enough to
2601 create a test for it. This is what the patch creator wrote:
2603 "I found a little strange behaviour of ldap code when working with
2604 Windows 2003 AD Domain, where users was placed in more than one
2605 Organization Units. When I tried to give exim partial DN, the exit code
2606 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2607 But simultaneously result of request was absolutely normal ldap result,
2608 so I produce this patch..."
2610 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2611 REFERENCE, so I have modified the code to exclude the patch when that macro
2614 55. Some experimental protocols are using DNS PTR records for new purposes. The
2615 keys for these records are domain names, not reversed IP addresses. The
2616 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2617 leaves it alone. Component reversal etc. now happens only for IP addresses.
2620 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2622 57. Double the size of the debug message buffer (to 2048) so that more of very
2623 long debug lines gets shown.
2625 58. The exicyclog utility now does better if the number of log files to keep
2626 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2628 59. Two changes related to the smtp_active_hostname option:
2630 (1) $smtp_active_hostname is now available as a variable.
2631 (2) The default for smtp_banner uses $smtp_active_hostname instead
2632 of $primary_hostname.
2634 60. The host_aton() function is supposed to be passed a string that is known
2635 to be a valid IP address. However, in the case of IPv6 addresses, it was
2636 not checking this. This is a hostage to fortune. Exim now panics and dies
2637 if the condition is not met. A case was found where this could be provoked
2638 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2639 components; fortuitously, this particular loophole had already been fixed
2640 by change 4.50/55 above.
2642 If there are any other similar loopholes, the new check in host_aton()
2643 itself should stop them being exploited. The report I received stated that
2644 data on the command line could provoke the exploit when Exim was running as
2645 exim, but did not say which command line option was involved. All I could
2646 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2647 running as the user.
2650 61. There was a buffer overflow vulnerability in the SPA authentication code
2651 (which came originally from the Samba project). I have added a test to the
2652 spa_base64_to_bits() function which I hope fixes it.
2655 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2656 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2658 63. The daemon start-up calls getloadavg() while still root for those OS that
2659 need the first call to be done as root, but it missed one case: when
2660 deliver_queue_load_max is set with deliver_drop_privilege. This is
2661 necessary for the benefit of the queue runner, because there is no re-exec
2662 when deliver_drop_privilege is set.
2664 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2665 This has been fixed.
2667 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2668 was in use, was not putting the data itself into the right store pool;
2669 consequently, it could be overwritten for a subsequent message in the same
2670 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2673 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2675 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2676 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2677 the function were treating the return as a boolean value, which happened to
2678 work because 0=false and not-0=true, but is not correct code.
2680 68. The host_aton() function was not handling scoped IPv6 addresses (those
2681 with, for example, "%eth0" on the end) correctly.
2683 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2684 negated items (that is, ~something) in unsigned ints. Some compilers
2685 apparently mutter when there is no cast.
2687 70. If an address verification called from an ACL failed, and did not produce a
2688 user-specific message (i.e. there was only a "system" message), nothing was
2689 put in $acl_verify_message. In this situation, it now puts the system
2692 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2693 session; change 4.31/43 added the unwanted input to the log line - except
2694 that it did not do this in the start of session case. It now does.
2696 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2697 This is wrong and can cause the other end to generate a synchronization
2698 error if it is another Exim or anything else that does the synchronization
2699 check. A QUIT command is no longer sent after a timeout.
2701 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2702 during host lookups.
2704 74. The defer_ok option of callout verification was not working if it was used
2705 when verifying addresses in header lines, that is, for this case:
2707 verify = header_sender/callout=defer_ok
2709 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2710 those file descriptors could be used for SMTP connections. If anything
2711 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2712 could be sent to the SMTP client, causing chaos. The daemon now opens
2713 stdin, stdout, and stderr to /dev/null when it puts itself into the
2716 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2717 log by default. The user can override this with suitable Perl magic.
2719 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2720 the log message when discard triggers, was not working for the DATA ACL or
2721 for the non-SMTP ACL.
2723 78. Error message wording change in sieve.c.
2725 79. If smtp_accept_max_per_host was set, the number of connections could be
2726 restricted to fewer than expected, because the daemon was trying to set up
2727 a new connection before checking whether the processes handling previous
2728 connections had finished. The check for completed processes is now done
2729 earlier. On busy systems, this bug wouldn't be noticed because something
2730 else would have woken the daemon, and it would have reaped the completed
2733 80. If a message was submitted locally by a user whose login name contained one
2734 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2735 It caused a spool format error. I have fixed the spool reading code. A
2736 related problem was that the "from" clause in the Received: line became
2737 illegal because of the space(s). It is now covered by ${quote_local_part.
2739 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2742 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2744 83. Patch from Sieve maintainer to fix unterminated string problem in
2745 "vacation" handling.
2747 84. Some minor changes to the Linux configuration files to help with other
2748 OS variants using glibc.
2750 85. One more patch for Sieve to update vacation handling to latest spec.
2753 ----------------------------------------------------
2754 See the note above about the 4.44 and 4.50 releases.
2755 ----------------------------------------------------
2761 1. Change 4.43/35 introduced a bug that caused file counts to be
2762 incorrectly computed when quota_filecount was set in an appendfile
2765 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2766 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2769 3. Give more explanation in the error message when the command for a transport
2770 filter fails to execute.
2772 4. There are several places where Exim runs a non-Exim command in a
2773 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2774 was being done only for the command run by the queryprogram router. It is
2775 now done for all such subprocesses. The other cases are: ${run, transport
2776 filters, and the commands run by the lmtp and pipe transports.
2778 5. Some older OS have a limit of 256 on the maximum number of file
2779 descriptors. Exim was using setrlimit() to set 1000 as a large value
2780 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2781 systems. I've change it so that if it can't get 1000, it tries for 256.
2783 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2784 was an oversight, and furthermore, ever since the addition of extra
2785 controls (e.g. 4.43/32), the checks on when to allow different forms of
2786 "control" were broken. There should now be diagnostics for all cases when a
2787 control that does not make sense is encountered.
2789 7. $recipients is now available in the predata ACL (oversight).
2791 8. Tidy the search cache before the fork to do a delivery from a message
2792 received from the command line. Otherwise the child will trigger a lookup
2793 failure and thereby defer the delivery if it tries to use (for example) a
2794 cached ldap connection that the parent has called unbind on.
2796 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2797 of $address_data from the recipient verification was clobbered by the
2798 sender verification.
2800 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2801 was its contents. (It was OK if the option was not defined at all.)
2803 11. A "Completed" log line is now written for messages that are removed from
2804 the spool by the -Mrm option.
2806 12. $host_address is now set to the target address during the checking of
2807 ignore_target_hosts.
2809 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2810 being passed; this would have caused $sender_host_name to have been used if
2811 matching the list had actually called for a host name (not very likely,
2812 since this list is usually IP addresses). A host name is now passed as
2815 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2816 code that specifies a non-restarting handler (typically sigaction() in
2817 modern systems) in an attempt to fix a rare and obscure crash bug.
2819 15. Narrowed the window for a race in the daemon that could cause it to ignore
2820 SIGCHLD signals. This is not a major problem, because they are used only to
2821 wake it up if nothing else does.
2823 16. A malformed maildirsize file could cause Exim to calculate negative values
2824 for the mailbox size or file count. Odd effects could occur as a result.
2825 The maildirsize information is now recalculated if the size or filecount
2828 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2829 support for a long time. Removed HAVE_SYS_VFS_H.
2831 18. Updated exipick to current release from John Jetmore.
2833 19. Allow an empty sender to be matched against a lookup in an address list.
2834 Previously the only cases considered were a regular expression, or an
2837 20. Exim went into a mad DNS lookup loop when doing a callout where the
2838 host was specified on the transport, if the DNS lookup yielded more than
2841 21. The RFC2047 encoding function was originally intended for short strings
2842 such as real names; it was not keeping to the 75-character limit for
2843 encoded words that the RFC imposes. It now respects the limit, and
2844 generates multiple encoded words if necessary. To be on the safe side, I
2845 have increased the buffer size for the ${rfc2047: expansion operator from
2848 22. Failure to deliver a bounce message always caused it to be frozen, even if
2849 there was an errors_to setting on the router. The errors_to setting is now
2852 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2853 canonical form (fully expanded) before being placed in
2854 $sender_host_address.
2856 24. Updated eximstats to version 1.33
2858 25. Include certificate and key file names in error message when GnuTLS fails
2859 to set them up, because the GnuTLS error message doesn't include the name
2860 of the failing file when there is a problem reading it.
2862 26. Expand error message when OpenSSL has problems setting up cert/key files.
2865 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2866 (this can affect the format of dates).
2868 28. exim_tidydb, when checking for the continued existence of a message for
2869 which it has found a message-specific retry record, was not finding
2870 messages that were in split spool directories. Consequently, it was
2871 deleting retry records that should have stayed in existence.
2873 29. eximstats updated to version 1.35
2874 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2875 1.35 - bugfix such that pie charts by volume are generated correctly
2877 30. The SPA authentication driver was not abandoning authentication and moving
2878 on to the next authenticator when an expansion was forced to fail,
2879 contradicting the general specification for all authenticators. Instead it
2880 was generating a temporary error. It now behaves as specified.
2882 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2883 (the order specifies the preference for clients). The order is now AES256,
2884 AES128, 3DES, ARCFOUR128.
2886 31. Small patch to Sieve code - explicitly set From: when generating an
2889 32. Exim crashed if a remote delivery caused a very long error message to be
2890 recorded - for instance if somebody sent an entire SpamAssassin report back
2891 as a large number of 550 error lines. This bug was coincidentally fixed by
2892 increasing the size of one of Exim's internal buffers (big_buffer) that
2893 happened as part of the Exiscan merge. However, to be on the safe side, I
2894 have made the code more robust (and fixed the comments that describe what
2897 33. Some experimental protocols are using DNS PTR records for new purposes. The
2898 keys for these records are domain names, not reversed IP addresses. The
2899 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2900 leaves it alone. Component reversal etc. now happens only for IP addresses.
2903 34. The host_aton() function is supposed to be passed a string that is known
2904 to be a valid IP address. However, in the case of IPv6 addresses, it was
2905 not checking this. This is a hostage to fortune. Exim now panics and dies
2906 if the condition is not met. A case was found where this could be provoked
2907 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2908 components; fortuitously, this particular loophole had already been fixed
2909 by change 4.50/55 or 4.44/33 above.
2911 If there are any other similar loopholes, the new check in host_aton()
2912 itself should stop them being exploited. The report I received stated that
2913 data on the command line could provoke the exploit when Exim was running as
2914 exim, but did not say which command line option was involved. All I could
2915 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2916 running as the user.
2919 35. There was a buffer overflow vulnerability in the SPA authentication code
2920 (which came originally from the Samba project). I have added a test to the
2921 spa_base64_to_bits() function which I hope fixes it.
2924 36. The daemon start-up calls getloadavg() while still root for those OS that
2925 need the first call to be done as root, but it missed one case: when
2926 deliver_queue_load_max is set with deliver_drop_privilege. This is
2927 necessary for the benefit of the queue runner, because there is no re-exec
2928 when deliver_drop_privilege is set.
2930 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2931 was in use, was not putting the data itself into the right store pool;
2932 consequently, it could be overwritten for a subsequent message in the same
2933 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2936 38. Sometimes the final signoff response after QUIT could fail to get
2937 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2938 < 0 before doing a fflush(). This bug looks as though it goes back to the
2939 introduction of TLS in release 3.20, but "sometimes" must have been rare
2940 because the tests only now provoked it.
2946 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2947 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2948 Now it returns an integer. A number of calls were still expecting a T/F
2949 return. Fortuitously, in all cases, the tests worked in OK situations,
2950 which is the norm. However, things would have gone wrong on any write
2951 failures on the smtp file descriptor. This function is used when sending
2952 messages over SMTP and also when doing verify callouts.
2954 2. When Exim is called to do synchronous delivery of a locally submitted
2955 message (the -odf or -odi options), it no longer closes stderr before doing
2958 3. Implemented the mua_wrapper option.
2960 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2962 5. Implemented the functions header_remove(), header_testname(),
2963 header_add_at_position(), and receive_remove_recipient(), and exported them
2966 6. If an ACL "warn" statement specified the addition of headers, Exim already
2967 inserted X-ACL-Warn: at the start if there was no header name. However, it
2968 was not making this test for the second and subsequent header lines if
2969 there were newlines in the string. This meant that an invalid header could
2970 be inserted if Exim was badly configured.
2972 7. Allow an ACL "warn" statement to add header lines at the start or after all
2973 the Received: headers, as well as at the end.
2975 8. Added the rcpt_4xx retry error code.
2977 9. Added postmaster_mailfrom=xxx to callout verification option.
2979 10. Added mailfrom=xxxx to the callout verification option, for verify=
2982 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2983 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2985 12. Inserted some casts to stop certain compilers warning when using pointer
2986 differences as field lengths or precisions in printf-type calls (mostly
2987 affecting debugging statements).
2989 13. Added optional readline() support for -be (dynamically loaded).
2991 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2992 same clock tick as a message's arrival, so that its received time was the
2993 same as the "first fail" time on the retry record, and that message
2994 remained on the queue past the ultimate address timeout, every queue runner
2995 would try a delivery (because it was past the ultimate address timeout) but
2996 after another failure, the ultimate address timeout, which should have then
2997 bounced the address, did not kick in. This was a "< instead of <=" error;
2998 in most cases the first failure would have been in the next clock tick
2999 after the received time, and all would be well.
3001 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3002 being recognized when the domain list was tested by the match_domain
3003 condition in an expansion string.
3005 16. Added the ${str2b64: operator.
3007 17. Exim was always calling setrlimit() to set a large limit for the number of
3008 processes, without checking whether the existing limit was already
3009 adequate. (It did check for the limit on file descriptors.) Furthermore,
3010 errors from getrlimit() and setrlimit() were being ignored. Now they are
3011 logged to the main and panic logs, but Exim does carry on, to try to do its
3012 job under whatever limits there are.
3014 18. Imported PCRE 5.0.
3016 19. Trivial typo in log message " temporarily refused connection" (the leading
3019 20. If the log selector return_path_on_delivery was set and an address was
3020 redirected to /dev/null, the delivery process crashed because it assumed
3021 that a return path would always be set for a "successful" delivery. In this
3022 case, the whole delivery is bypassed as an optimization, and therefore no
3025 21. Internal re-arrangement: the function for sending a challenge and reading
3026 a response while authentication was assuming a zero-terminated challenge
3027 string. It's now changed to take a pointer and a length, to allow for
3028 binary data in such strings.
3030 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3032 23. Exim was not respecting finduser_retries when seeking the login of the
3033 uid under which it was called; it was always trying 10 times. (The default
3034 setting of finduser_retries is zero.) Also, it was sleeping after the final
3035 failure, which is pointless.
3037 24. Implemented tls_on_connect_ports.
3039 25. Implemented acl_smtp_predata.
3041 26. If the domain in control=submission is set empty, Exim assumes that the
3042 authenticated id is a complete email address when it generates From: or
3043 Sender: header lines.
3045 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3046 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3047 chown and chgrp in /bin and hostname in /usr/bin.
3049 28. Exim was keeping the "process log" file open after each use, just as it
3050 does for the main log. This opens the possibility of it remaining open for
3051 long periods when the USR1 signal hits a daemon. Occasional processlog
3052 errors were reported, that could have been caused by this. Anyway, it seems
3053 much more sensible not to leave this file open at all, so that is what now
3056 29. The long-running daemon process does not normally write to the log once it
3057 has entered its main loop, and it closes the log before doing so. This is
3058 so that log files can straightforwardly be renamed and moved. However,
3059 there are a couple of unusual error situations where the daemon does write
3060 log entries, and I had neglected to close the log afterwards.
3062 30. The text of an SMTP error response that was received during a remote
3063 delivery was being truncated at 512 bytes. This is too short for some of
3064 the long messages that one sometimes sees. I've increased the limit to
3067 31. It is now possible to make retry rules that apply only when a message has a
3068 specific sender, in particular, an empty sender.
3070 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3071 it possible to be selective about when SMTP synchronization is enforced.
3073 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3075 32. Implemented hosts_connection_nolog.
3077 33. Added an ACL for QUIT.
3079 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3082 35. Added mailbox_size and mailbox_filecount to appendfile.
3084 36. Added control = no_multiline_responses to ACLs.
3086 37. There was a bug in the logic of the code that waits for the clock to tick
3087 in the case where the clock went backwards by a substantial amount such
3088 that the microsecond fraction of "now" was more than the microsecond
3089 fraction of "then" (but the whole seconds number was less).
3091 38. Added support for the libradius Radius client library this is found on
3092 FreeBSD (previously only the radiusclient library was supported).
3098 1. When certain lookups returned multiple values in the form name=value, the
3099 quoting of the values was not always being done properly. Specifically:
3100 (a) If the value started with a double quote, but contained no whitespace,
3102 (b) If the value contained whitespace other than a space character (i.e.
3103 tabs or newlines or carriage returns) it was not quoted.
3104 This fix has been applied to the mysql and pgsql lookups by writing a
3105 separate quoting function and calling it from the lookup code. The fix
3106 should probably also be applied to nisplus, ibase and oracle lookups, but
3107 since I cannot test any of those, I have not disturbed their existing code.
3109 2. A hit in the callout cache for a specific address caused a log line with no
3110 reason for rejecting RCPT. Now it says "Previous (cached) callout
3111 verification failure".
3113 3. There was an off-by-one bug in the queryprogram router. An over-long
3114 return line was truncated at 256 instead of 255 characters, thereby
3115 overflowing its buffer with the terminating zero. As well as fixing this, I
3116 have increased the buffer size to 1024 (and made a note to document this).
3118 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3119 when Exim is waiting for an SMTP response from a remote server, Exim
3120 restarts its select() call on the socket, thereby resetting its timeout.
3121 This is not a problem when such interrupts are rare. Somebody set up a cron
3122 job to run exiwhat every 2 minutes, which is less than the normal select()
3123 timeout (5 or 10 minutes). This meant that the select() timeout never
3124 kicked in because it was always reset. I have fixed this by comparing the
3125 time when an interrupt arrives with the time at the start of the first call
3126 to select(). If more time than the timeout has elapsed, the interrupt is
3127 treated as a timeout.
3129 5. Some internal re-factoring in preparation for the addition of Sieve
3130 extensions (by MH). In particular, the "personal" test is moved to a
3131 separate function, and given an option for scanning Cc: and Bcc: (which is
3132 not set for Exim filters).
3134 6. When Exim created an email address using the login of the caller as the
3135 local part (e.g. when creating a From: or Sender: header line), it was not
3136 quoting the local part when it contained special characters such as @.
3138 7. Installed new OpenBSD configuration files.
3140 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3141 try to make them clearer.
3143 9. Callout options, other than the timeout value, were being ignored when
3144 verifying sender addresses in header lines. For example, when using
3146 verify = header_sender/callout=no_cache
3148 the cache was (incorrectly) being used.
3150 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3151 only the Cygwin environment.
3153 11. When return_path_on_delivery was set as a log selector, if different remote
3154 addresses in the same message used different return paths and parallel
3155 remote delivery occurred, the wrong values would sometimes be logged.
3156 (Whenever a remote delivery process finished, the return path value from
3157 the most recently started remote delivery process was logged.)
3159 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3160 lines when AUTH and/or TLS are in use. This is the "received protocol"
3161 field. Exim used to use "asmtp" for authenticated SMTP, without any
3162 indication (in the protocol name) for TLS use. Now it follows the RFC and
3163 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3164 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3165 names appear in log lines as well as in Received: header lines.
3167 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3168 extensions, and comparison tests, and to fix some bugs.
3170 14. Changes to the "personal" filter test:
3172 (1) The test was buggy in that it was just doing the equivalent of
3173 "contains" tests on header lines. For example, if a user's address was
3174 anne@some.where, the "personal" test would incorrectly be true for
3176 To: susanne@some.where
3178 This test is now done by extracting each address from the header in turn,
3179 and checking the entire address. Other tests that are part of "personal"
3180 are now done using regular expressions (for example, to check local parts
3181 of addresses in From: header lines).
3183 (2) The list of non-personal local parts in From: addresses has been
3184 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3185 taken from the Sieve specification recommendations.
3187 (3) If the message contains any header line starting with "List-" it is
3188 treated as non-personal.
3190 (4) The test for "circular" in the Subject: header line has been removed
3191 because it now seems ill-conceived.
3193 15. Minor typos in src/EDITME comments corrected.
3195 16. Installed latest exipick from John Jetmore.
3197 17. If headers_add on a router specified a text string that was too long for
3198 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3199 of string_sprintf() is now avoided.
3201 18. $message_body_size was not set (it was always zero) when running the DATA
3202 ACL and the local_scan() function.
3204 19. For the "mail" command in an Exim filter, no default was being set for
3205 the once_repeat time, causing a random time value to be used if "once" was
3206 specified. (If the value happened to be <= 0, no repeat happened.) The
3207 default is now 0s, meaning "never repeat". The "vacation" command was OK
3208 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3209 (I found it when inspecting the code).
3211 20. There is now an overall timeout for performing a callout verification. It
3212 defaults to 4 times the callout timeout, which applies to individual SMTP
3213 commands during the callout. The overall timeout applies when there is more
3214 than one host that can be tried. The timeout is checked before trying the
3215 next host. This prevents very long delays if there are a large number of
3216 hosts and all are timing out (e.g. when the network connections are timing
3217 out). The value of the overall timeout can be changed by specifying an
3218 additional sub-option for "callout", called "maxwait". For example:
3220 verify = sender/callout=5s,maxwait=20s
3222 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3223 to the end before writing, but this should make it even safer).
3225 22. Exim was forgetting that it had advertised PIPELINING for the second and
3226 subsequent messages on an SMTP connection. It was also not resetting its
3227 memory on STARTTLS and an internal HELO.
3229 23. When Exim logs an SMTP synchronization error within a session, it now
3230 records whether PIPELINING has been advertised or not.
3232 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3233 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3234 rather than long int.
3236 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3238 26. Added the never_mail option to autoreply.
3244 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3245 crash if the getsockname() call failed; this can happen if a connection is
3246 closed very soon after it is established. The problem was simply in the
3247 order in which certain operations were done, causing Exim to try to write
3248 to the SMTP stream before it had set up the file descriptor. The bug has
3249 been fixed by making things happen in the correct order.
3255 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3256 before the connection was closed, thus losing the rejection response.
3258 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3259 some early Solaris releases, but causes trouble in current releases where
3260 socklen_t is defined.
3262 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3265 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3268 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3269 that releasing the top of it at the end releases what was used for sub-
3270 expansions (unless the block got too big). However, discard this block if
3271 the first thing is a variable or header, so that we can use its block when
3272 it is dynamic (useful for very large $message_headers, for example).
3274 6. Lookups now cache *every* query, not just the most recent. A new, separate
3275 store pool is used for this. It can be recovered when all lookup caches are
3276 flushed. Lookups now release memory at the end of their result strings.
3277 This has involved some general refactoring of the lookup sources.
3279 7. Some code has been added to the store_xxx() functions to reduce the amount
3280 of flapping under certain conditions.
3282 8. log_incoming_interface used to affect only the <= reception log lines. Now
3283 it causes the local interface and port to be added to several more SMTP log
3284 lines, for example "SMTP connection from", and rejection lines.
3286 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3288 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3290 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3291 could be overwritten at the end of the current message (or the start of a
3292 new message if it was set in a HELO ACL). The value is now preserved for
3293 the duration of the SMTP connection.
3295 12. If a transport had a headers_rewrite setting, and a matching header line
3296 contained an unqualified address, that address was qualified, even if it
3297 did not match any rewriting rules. The underlying bug was that the values
3298 of the flags that permit the existence of unqualified sender and recipient
3299 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3300 non-local messages, and by -bnq for local messages) were not being
3301 preserved with the message after it was received.
3303 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3304 "next input=" as part of the text comprising the host identity instead of
3305 the correct text. The code was using the same buffer for two different
3306 strings. However, depending on which order the printing function evaluated
3307 its arguments, the bug did not always show up. Under Linux, for example, my
3308 test suite worked just fine.
3310 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3311 doesn't work with some older versions of Perl. It has been changed to "my",
3312 which in any case is probably the better facility to use.
3314 15. A really picky compiler found some instances of statements for creating
3315 error messages that either had too many or two few arguments for the format
3318 16. The size of the buffer for calls to the DNS resolver has been increased
3319 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3320 for addresses that have a lot of PTR records. This alleviates a problem; it
3321 does not fully solve it.
3323 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3324 buffer now truncates the list and logs the incident, which is the same
3325 action as happens when Exim is looking up a host name and its aliases.
3326 Previously in this situation something unpredictable would happen;
3327 sometimes it was "internal error: store_reset failed".
3329 18. If a server dropped the connection unexpectedly when an Exim client was
3330 using GnuTLS and trying to read a response, the client delivery process
3331 crashed while trying to generate an error log message.
3333 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3334 string, for example:
3336 warn message = H1: something\nH2: something
3338 the text was added as a single header line from Exim's point of view
3339 though it ended up OK in the delivered message. However, searching for the
3340 second and subsequent header lines using $h_h2: did not work. This has been
3341 fixed. Similarly, if a system filter added multiple headers in this way,
3342 the routers could not see them.
3344 20. Expanded the error message when iplsearch is called with an invalid key to
3345 suggest using net-iplsearch in a host list.
3347 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3348 ACLs are no longer actually imposed (and a message to that effect is
3351 22. If a "gecos" field in a passwd entry contained escaped characters, in
3352 particular, if it contained a \" sequence, Exim got it wrong when building
3353 a From: or a Sender: header from that name. A second bug also caused
3354 incorrect handling when an unquoted " was present following a character
3355 that needed quoting.
3357 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3358 was not being matched caselessly.
3360 24. Arranged for all hyphens in the exim.8 source to be escaped with
3363 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3364 properly. Recipient callout cache records were still being keyed to include
3365 the sender, even when use_sender was set false. This led to far more
3366 callouts that were necessary. The sender is no longer included in the key
3367 when use_sender is false.
3369 26. Added "control = submission" modifier to ACLs.
3371 27. Added the ${base62d: operator to decode base 62 numbers.
3373 28. dnsdb lookups can now access SRV records.
3375 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3376 the configuration file.
3378 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3379 (-v) mode. This makes the output for a verbose queue run more intelligible.
3381 31. Added a use_postmaster feature to recipient callouts.
3383 32. Added the $body_zerocount variable, containing the number of binary zero
3384 bytes in the message body.
3386 33. The time of last modification of the "new" subdirectory is now used as the
3387 "mailbox time last read" when there is a quota error for a maildir
3390 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3392 35. Added +ignore_unknown as a special item in host lists.
3394 36. Code for decoding IPv6 addresses in host lists is now included, even if
3395 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3396 address was recognized as an IP address, but was then not correctly decoded
3397 into binary, causing unexpected and incorrect effects when compared with
3404 1. Very minor rewording of debugging text in manualroute to say "list of
3405 hosts" instead of "hostlist".
3407 2. If verify=header_syntax was set, and a header line with an unqualified
3408 address (no domain) and a large number of spaces between the end of the
3409 name and the colon was received, the reception process suffered a buffer
3410 overflow, and (when I tested it) crashed. This was caused by some obsolete
3411 code that should have been removed. The fix is to remove it!
3413 3. When running in the test harness, delay a bit after writing a bounce
3414 message to get a bit more predictability in the log output.
3416 4. Added a call to search_tidyup() just before forking a reception process. In
3417 theory, someone could use a lookup in the expansion of smtp_accept_max_
3418 per_host which, without the tidyup, could leave open a database connection.
3420 5. Added the variables $recipient_data and $sender_data which get set from a
3421 lookup success in an ACL "recipients" or "senders" condition, or a router
3422 "senders" option, similar to $domain_data and $local_part_data.
3424 6. Moved the writing of debug_print from before to after the "senders" test
3427 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3428 problems for message scanning, either using a data ACL, or using
3429 local_scan() because the Received: header was not generated till after they
3430 were called (in order to set the time as the time of reception completion).
3431 I have revised the way this works. The header is now generated after the
3432 body is received, but before the ACL or local_scan() are called. After they
3433 are run, the timestamp in the header is updated.
3439 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3440 before starting a queue runner without re-exec. This happened only when
3441 deliver_drop_privilege was set or when the Exim user was set to root. The
3442 effect of the bug was that timeouts during subsequent deliveries caused
3443 crashes instead of being properly handled. The handler is now left at its
3444 default (and expected) setting.
3446 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3447 message, again when deliver_drop_privilege is set or Exim is run as root.
3448 The bug described in (1) was not present in this case, but the tidying up
3449 of the other signals was missing. I have made the two cases consistent.
3451 3. The ignore_target_hosts setting on a manualroute router was being ignored
3452 for hosts that were looked up using the /MX notation.
3454 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3457 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3458 operated on the sender address. After changing the $sender_address to <>
3459 for the sender address verify, Exim was re-instated it as the original
3460 (before rewriting) address, but remembering that it had rewritten it, so it
3461 wasn't rewriting it again. This bug also had the effect of breaking the
3462 sender address verification caching when the sender address was rewritten.
3464 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3465 This has been changed so that if the ip literal address matches
3466 ignore_target_hosts, the router declines.
3468 7. Added expansion conditions match_domain, match_address, and match_local_
3469 part (NOT match_host).
3471 8. The placeholder for the Received: header didn't have a length field set.
3473 9. Added code to Exim itself and to exim_lock to test for a specific race
3474 condition that could lead to file corruption when using MBX delivery. The
3475 issue is with the lockfile that is created in /tmp. If this file is removed
3476 after a process has opened it but before that process has acquired a lock,
3477 there is the potential for a second process to recreate the file and also
3478 acquire a lock. This could lead to two Exim processes writing to the file
3479 at the same time. The added code performs the same test as UW imapd; it
3480 checks after acquiring the lock that its file descriptor still refers to
3481 the same named file.
3483 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3484 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3487 11. Added the smtp_active_hostname option. If used, this will typically be made
3488 to depend on the incoming interface address. Because $interface_address is
3489 not set up until the daemon has forked a reception process, error responses
3490 that can happen earlier (such as "too many connections") no longer contain
3493 12. If an expansion in a condition on a "warn" statement fails because a lookup
3494 defers, the "warn" statement is abandoned, and the next ACL statement is
3495 processed. Previously this caused the whole ACL to be aborted.
3497 13. Added the iplsearch lookup type.
3499 14. Added ident_timeout as a log selector.
3501 15. Added tls_certificate_verified as a log selector.
3503 16. Added a global option tls_require_ciphers (compare the smtp transport
3504 option of the same name). This controls incoming TLS connections.
3506 17. I finally figured out how to make tls_require_ciphers do a similar thing
3507 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3508 before starting the TLS session.
3510 18. Tabs are now shown as \t in -bP output.
3512 19. If the log selector return_path_on_delivery was set, Exim crashed when
3513 bouncing a message because it had too many Received: header lines.
3515 20. If two routers both had headers_remove settings, and the first one included
3516 a superfluous trailing colon, the final name in the first list and the
3517 first name in the second list were incorrectly joined into one item (with a
3518 colon in the middle).
3524 1. Added -C and -D options to the exinext utility, mainly to make it easier
3525 to include in the automated testing, but these could be helpful when
3526 multiple configurations are in use.
3528 2. The exinext utility was not formatting the output nicely when there was
3529 an alternate port involved in the retry record key, nor when there was a
3530 message id as well (for retries that were specific to a specific message
3531 and a specific host). It was also confused by IPv6 addresses, because of
3532 the additional colons they contain. I have fixed the IPv4 problem, and
3533 patched it up to do a reasonable job for IPv6.
3535 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3536 delivery, the log line now contains "pipelined" if PIPELINING was used.
3538 4. An SMTP transport process used to panic and die if the bind() call to set
3539 an explicit outgoing interface failed. This has been changed; it is now
3540 treated in the same way as a connect() failure.
3542 5. A reference to $sender_host_name in the part of a conditional expansion
3543 that was being skipped was still causing a DNS lookup. This no longer
3546 6. The def: expansion condition was not recognizing references to header lines
3547 that used bh_ and bheader_.
3549 7. Added the _cache feature to named lists.
3551 8. The code for checking quota_filecount in the appendfile transport was
3552 allowing one more file than it should have been.
3554 9. For compatibility with Sendmail, the command line option
3562 and sets the incoming protocol and host name (for trusted callers). The
3563 host name and its colon can be omitted when only the protocol is to be set.
3564 Note the Exim already has two private options, -pd and -ps, that refer to
3565 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3566 "s", but I don't think that's a major issue.
3568 10. A number of refactoring changes to the code, none of which should affect
3571 (a) The number of logging options was getting close to filling up the
3572 32-bit word that was used as a bit map. I have split them into two classes:
3573 those that are passed in the argument to log_write(), and those that are
3574 only ever tested independently outside of that function. These are now in
3575 separate 32-bit words, so there is plenty of room for expansion again.
3576 There is no change in the user interface or the logging behaviour.
3578 (b) When building, for example, log lines, the code previously used a
3579 macro that called string_cat() twice, in order to add two strings. This is
3580 not really sufficiently general. Furthermore, there was one instance where
3581 it was actually wrong because one of the argument was used twice, and in
3582 one call a function was used. (As it happened, calling the function twice
3583 did not affect the overall behaviour.) The macro has been replaced by a
3584 function that can join an arbitrary number of extra strings onto a growing
3587 (c) The code for expansion conditions now uses a table and a binary chop
3588 instead of a serial search (which was left over from when there were very
3589 few conditions). Also, it now recognizes conditions like "pam" even when
3590 the relevant support is not compiled in: a suitably worded error message is
3591 given if an attempt is made to use such a condition.
3593 11. Added ${time_interval:xxxxx}.
3595 12. A bug was causing one of the ddress fields not to be passed back correctly
3596 from remote delivery subprocesses. The field in question was not being
3597 subsequently used, so this caused to problems in practice.
3599 13. Added new log selectors queue_time and deliver_time.
3601 14. Might have fixed a bug in maildirsizefile handling that threw up
3602 "unexpected character" debug warnings, and recalculated the data
3603 unnecessarily. In any case, I expanded the warning message to give more
3606 15. Added the message "Restricted characters in address" to the statements in
3607 the default ACL that block characters like @ and % in local parts.
3609 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3610 Three changes have been made:
3612 (a) There was a serious bug; a negative response to MAIL caused the whole
3613 recipient domain to be cached as invalid, thereby blocking all messages
3614 to all local parts at the same domain, from all senders. This bug has
3615 been fixed. The domain is no longer cached after a negative response to
3616 MAIL if the sender used is not empty.
3618 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3621 (c) A new callout option, "use_sender" has been added for people who want
3622 the modified behaviour.
3628 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3631 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3632 indeed breaks things for older releases.
3634 3. Added additional logging to the case where there is a problem reading data
3635 from a filter that is running in a subprocess using a pipe, in order to
3636 try to track down a specific problem.
3638 4. Testing facility fudge: when running in the test harness and attempting
3639 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3640 getting "No route to host". Convert this to a timeout.
3642 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3645 6. Some OS don't have socklen_t but use size_t instead. This affects the
3646 fifth argument of getsockopt() amongst other things. This is now
3647 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3648 can be set for individual OS. I have set it for SunOS5, OSF1, and
3649 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3650 some earlier ones do not.
3652 7. Change 4.30/15 was not doing the test caselessly.
3654 8. The standard form for an IPv6 address literal was being rejected by address
3655 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3656 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3657 this, as well as the form without the "IPv6" on the front (but only when
3658 address literals are enabled, of course).
3660 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3662 10. Exim crashed if a message with an empty sender address specified by -f
3663 encountered a router with an errors_to setting. This could be provoked only
3664 by a command such as
3668 where an empty string was supplied; "<>" did not hit this bug.
3670 11. Installed PCRE release 4.5.
3672 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3673 remained set. It is now erased.
3675 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3676 times from message ids (which are base 36 rather than the normal 62).
3678 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3679 were being counted as actual protocol errors, and logged if the log
3680 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3681 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3682 and DATA following a set of rejected RCPTs do not count as protocol errors.
3683 In other words, Exim assumes they were pipelined, though this may not
3684 actually be the case. Of course, in all cases the client gets an
3685 appropriate error code.
3687 15. If a lookup fails in an ACL condition, a message about the failure may
3688 be available; it is used if testing the ACL cannot continue, because most
3689 such messages specify what the cause of the deferral is. However, some
3690 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3691 that caused an old message to be retained and used if a later statement
3692 caused a defer, replacing the real cause of the deferral.
3694 16. If an IP address had so many PTR records that the DNS lookup buffer
3695 was not large enough to hold them, Exim could crash while trying to process
3696 the truncated data. It now detects and logs this case.
3698 17. Further to 4.21/58, another change has been made: if (and only if) the
3699 first line of a message (the first header line) ends with CRLF, a bare LF
3700 in a subsequent header line has a space inserted after it, so as not to
3701 terminate the header.
3703 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3704 unnecessarily, used atoi() instead of strtol(), and didn't check the
3705 termination when getting file sizes from file names by regex.
3707 19. Completely re-implemented the support for maildirsize files, in the light
3708 of a number of problems with the previous contributed implementation
3709 (4.30/29). In particular:
3711 . If the quota is zero, the maildirsize file is maintained, but no quota is
3714 . If the maildir directory does not exist, it is created before any attempt
3715 to write a maildirsize file.
3717 . The quota value in the file is just a cache; if the quota is changed in
3718 the transport, the new value overrides.
3720 . A regular expression is available for excluding directories from the
3723 20. The autoreply transport checks the characters in options that define the
3724 message's headers; it allows continued headers, but it was checking with
3725 isspace() after an embedded newline instead of explicitly looking for a
3728 21. If all the "regular" hosts to which an address was routed had passed their
3729 expiry times, and had not reached their retry times, the address was
3730 bounced, even if fallback hosts were defined. Now Exim should go on to try
3733 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3734 equivalent code in the SMTP transport. Some hosts send humungous responses
3735 to HELO/EHLO, more than 1024 it seems.
3737 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3738 gives compiler warnings in some environments. I've now done it "properly",
3741 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3742 (because of problems with the built-in one) was declared to return uschar *
3743 instead of char *, causing compiler failure.
3745 25. Fixed a file descriptor leak when processing alias/forward files.
3747 26. Fixed a minor format string issue in dbfn.c.
3749 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3751 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3752 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3755 29. When a sender address is verified, it is cached, to save repeating the test
3756 when there is more than one recipient in a message. However, when the
3757 verification involves a callout, it is possible for different callout
3758 options to be set for different recipients. It is too complicated to keep
3759 track of this in the cache, so now Exim always runs a verification when a
3760 callout is required, relying on the callout cache for the optimization.
3761 The overhead is duplication of the address routing, but this should not be
3764 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3765 to be verified with callout=postmaster, and the main callout worked but the
3766 postmaster check failed, the verification correctly failed. However, if a
3767 subsequent RCPT command asked for sender verification *without* the
3768 postmaster check, incorrect caching caused this verification also to fail,
3771 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3772 it was not caching the DNS options (qualify_single, search_parents) that
3773 were used when the lookup failed. A subsequent lookup with different
3774 options therefore always gave the same answer, though there were cases
3775 where it should not have. (Example: a "domains = !$mx_any" option on a
3776 dnslookup router: the "domains" option is always processed without any
3777 widening, but the router might have qualify_single set.) Now Exim uses the
3778 cached value only when the same options are set.
3780 32. Added John Jetmore's "exipick" utility to the distribution.
3782 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3783 than a timeout (e.g. a certificate is required, and is not provided), an
3784 Exim server now closes the connection immediately. Previously it waited for
3785 the client to close - but if the client is SSL, it seems that they each
3786 wait for each other, leading to a delay before one of them times out.
3788 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3789 maintained 0.8.x compatibility because I don't think many are using it, and
3790 it is clearly obsolete.
3792 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3795 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3796 client certificate was expired. A simple patch fixes this, though I don't
3797 understand the full logic of why the verify callback is called multiple
3800 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3801 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3802 which causes problems with some clients (such as the Certicom SSL Plus
3803 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3804 disables the coutermeasure allowing Eudora to connect."
3806 38. Exim was not checking that a write() to a log file succeeded. This could
3807 lead to Bad Things if a log got too big, in particular if it hit a file
3808 size limit. Exim now panics and dies if it cannot write to a log file, just
3809 as it does if it cannot open a log file.
3811 39. Modified OS/Makefile-Linux so that it now contains
3813 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3815 The two -D definitions ensure that Exim is compiled with large file
3816 support, which makes it possible to handle log files that are bigger than
3819 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3820 instance) a domain was checked against a named list that involved a lookup,
3821 causing $domain_data to be set, then another domain was checked against the
3822 same list, then the first domain was re-checked, the value of $domain_data
3823 after the final check could be wrong. In particular, if the second check
3824 failed, it could be set empty. This bug probably also applied to
3827 41. The strip_trailing_dot option was not being applied to the address given
3828 with the -f command-line option.
3830 42. The code for reading a message's header from the spool was incrementing
3831 $received_count, but never initializing it. This meant that the value was
3832 incorrect (doubled) while delivering a message in the same process in which
3833 it was received. In the most common configuration of Exim, this never
3834 happens - a fresh exec is done - but it can happen when
3835 deliver_drop_privilege is set.
3837 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3838 it now includes up to 150 characters of the unexpected data in the log
3841 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3842 and building data strings. The size of both of these buffers was 10 000
3843 bytes - far larger than anybody would *ever* want, thought I. Needless to
3844 say, somebody hit the limit. I have increased the maximum line length to
3845 20 000 and the maximum data length of concatenated lines to 100 000. I have
3846 also fixed two bugs, because there was no checking on these buffers. Tsk,
3847 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3848 buffer is too small.
3850 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3851 lsearch lookups. Now it does.
3853 46. When parsing a route_list item in a manualroute router, a fixed-length
3854 buffer was used for the list of hosts. I made this 1024 bytes long,
3855 thinking that nobody would ever have a list of hosts that long. Wrong.
3856 Somebody had a whole pile of complicated expansion conditions, and the
3857 string was silently truncated, leading to an expansion error. It turns out
3858 that it is easier to change to an unlimited length (owing to other changes
3859 that have happened since this code was originally written) than to build
3860 structure for giving a limitation error. The length of the item that
3861 expands into the list of hosts is now unlimited.
3863 47. The lsearch lookup could not handle data where the length of text line was
3864 more than 4095 characters. Such lines were truncated, leading to shortened
3865 data being returned. It should now handle lines of any length.
3867 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3868 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3871 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3872 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3873 no longer happens in comment lines. A list of replacements is now placed
3874 at the head of all of the source files, except those whose only change is
3875 to replace PERL_COMMAND in the very first #! line.
3877 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3878 messages on the queue, with a bottom-up merge sort, using code contributed
3879 by Michael Haardt. This should make operations like -bp somewhat faster on
3880 large queues. It won't affect queue runners, except when queue_run_in_order
3883 51. Installed eximstats 1.31 in the distribution.
3885 52. Added support for SRV lookups to the dnslookup router.
3887 53. If an ACL referred to $message_body or $message_body_end, the value was not
3888 reset for any messages that followed in the same SMTP session.
3890 54. The store-handling optimization for building very long strings was not
3891 differentiating between the different store pools. I don't think this
3892 actually made any difference in practice, but I've tidied it.
3894 55. While running the routers to verify a sender address, $sender_address
3895 was still set to the sender address. This is wrong, because when routing to
3896 send a bounce to the sender, it would be empty. Therefore, I have changed
3897 it so that, while verifying a sender address, $sender_address is set to <>.
3898 (There is no change to what happens when verifying a recipient address.)
3900 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3901 target A or AAAA records (if not already returned) without resetting the
3902 qualify_single or search_parents options of the DNS resolver. These are
3903 inappropriate in this case because the targets of MX and SRV records must
3904 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3905 target that, when qualified, matched something in the local domain. These
3906 two options are now turned off when doing these lookups.
3908 57. It seems that at least some releases of Reiserfs (which does not have the
3909 concept of a fixed number of inodes) returns zero and not -1 for the
3910 number of available inodes. This interacted badly with check_spool_inodes,
3911 which assumed that -1 was the "no such thing" setting. What I have done is
3912 to check that the total number of inodes is greater than zero before doing
3913 the test of how many are available.
3915 58. When a "warn" ACL statement has a log_message modifier, the message is
3916 remembered, and not repeated. This is to avoid a lot of repetition when a
3917 message has many recipients that cause the same warning to be written.
3918 Howewer, Exim was preserving the list of already written lines for an
3919 entire SMTP session, which doesn't seem right. The memory is now reset if a
3920 new message is started.
3922 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3923 debugging output unless log_rewrite was also set.
3925 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3926 of (int)(handle) when we know that handle contains (void *)(-1).
3928 61. The Exim daemon panic-logs an error return when it closes the incoming
3929 connection. However "connection reset by peer" seems to be common, and
3930 isn't really an error worthy of noting specially, so that particular error
3933 62. When Exim is trying to find all the local interfaces, it used to panic and
3934 die if the ioctl to get the interface flags failed. However, it seems that
3935 on at least one OS (Solaris 9) it is possible to have an interface that is
3936 included in the list of interfaces, but for which you get a failure error
3937 for this call. This happens when the interface is not "plumbed" into a
3938 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3939 failure of the "get flags" call assumes that the interface is down.
3941 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3942 makes life easier for people who are doing arithmetic on fields extracted
3943 from dates, where you often get leading zeros that should not be
3944 interpreted as octal.
3946 64. Added qualify_domain to the redirect router, to override the global
3949 65. If a pathologically long header line contained very many addresses (the
3950 report of this problem mentioned 10 000) and each of them was rewritten,
3951 Exim could use up a very large amount of memory. (It kept on making new
3952 copies of the header line as it rewrote, and never released the old ones.)
3953 At the expense of a bit more processing, the header rewriting function has
3954 been changed so that it no longer eats memory in this way.
3956 66. The generation of the Received: header has been moved from the time that a
3957 message starts to be received, to the time that it finishes. The timestamp
3958 in the Received: header should now be very close to that of the <= log
3959 line. There are two side-effects of this change:
3961 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3962 logged header lines no longer include the local Received: line, because
3963 it has not yet been created. The same applies to a copy of the message
3964 that is returned to a non-SMTP sender when a message is rejected.
3966 (b) When a filter file is tested using -bf, no additional Received: header
3967 is added to the test message. After some thought, I decided that this
3970 This change does not affect the value of $received_for. It is still set
3971 after address rewriting, but before local_scan() is called.
3973 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3975 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3976 gave an unhelpful panic error message, and a defer error. I have managed to
3977 change this behaviour so that it now rejects any supplied certificate,
3978 which seems right, as the list of acceptable certificates is empty.
3980 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3981 gave an unhelpful defer error. I have not managed to make this reject any
3982 supplied certificates, but the error message it gives is "no certificate
3983 supplied", which is not helpful.
3985 70. exigrep's output now also includes lines that are not associated with any
3986 message, but which match the given pattern. Implemented by a patch from
3987 Martin Sluka, which also tidied up the Perl a bit.
3989 71. Recipient callout verification, like sender verification, was using <> in
3990 the MAIL FROM command. This isn't really the right thing, since the actual
3991 sender may affect whether the remote host accepts the recipient or not. I
3992 have changed it to use the actual sender in the callout; this means that
3993 the cache record is now keyed on a recipient/sender pair, not just the
3994 recipient address. There doesn't seem to be a real danger of callout loops,
3995 since a callout by the remote host to check the sender would use <>.
3996 [SEE ABOVE: changed after hitting problems.]
3998 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3999 temporary errors. However, in the case of such a code being given after
4000 the end of a data transmission (i.e. after ".") Exim was failing to write
4001 a retry record for the message. (Yes, there was some broken host that was
4002 actually sending 8xx at this point.)
4004 73. An unknown lookup type in a host list could cause Exim to panic-die when
4005 the list was checked. (An example that provoked this was putting <; in the
4006 middle of a list instead of at the start.) If this happened during a DATA
4007 ACL check, a -D file could be left lying around. This kind of configuration
4008 error no longer causes Exim to die; instead it causes a defer errror. The
4009 incident is still logged to the main and panic logs.
4011 74. Buglet left over from Exim 3 conversion. The message "too many messages
4012 in one connection" was written to the rejectlog but not the mainlog, except
4013 when address rewriting (yes!) was being logged.
4015 75. Added write_rejectlog option.
4017 76. When a system filter was run not as root (that is, when system_filter_user
4018 was set), the values of the $n variables were not being returned to the
4019 main process; thus, they were not subsequently available in the $sn
4022 77. Added +return_path_on_delivery log selector.
4024 78. A connection timeout was being treated differently from recipients deferred
4025 when testing hosts_max_try with a message that was older than the host's
4026 retry timeout. (The host should not be counted, thus allowing all hosts to
4027 be tried at least once before bouncing.) This may have been the cause of an
4028 occasionally reported bug whereby a message would remain on the queue
4029 longer than the retry timeout, but would be bounced if a delivery was
4030 forced. I say "may" because I never totally pinned down the problem;
4031 setting up timeout/retry tests is difficult. See also the next item.
4033 79. The ultimate address timeout was not being applied to errors that involved
4034 a combination of host plus message (for example, a timeout on a MAIL
4035 command). When an address resolved to a number of possible hosts, and they
4036 were not all tried for each delivery (e.g. because of hosts_max_try), a
4037 message could remain on the queue longer than the retry timeout.
4039 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4042 81. Fixed an obscure SMTP outgoing bug which required at least the following
4043 conditions: (a) there was another message waiting for the same server;
4044 (b) the server returned 5xx to all RCPT commands in the first message so
4045 that the message was not completed; (c) the server dropped the connection
4046 or gave a negative response to the RSET that Exim sends to abort the
4047 transaction. The observed case was a dropped connection after DATA that had
4048 been sent in pipelining mode. That is, the server had advertised PIPELINING
4049 but was not implementing it correctly. The effect of the bug was incorrect
4050 behaviour, such as trying another host, and this could lead to a crash.
4056 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4057 and daemon.c were passed as pointers to ints; they should have been
4058 pointers to socklen_t variables (which are typically unsigned ints).
4060 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4063 3. Fixed a really odd bug that affected only the testing scheme; patching a
4064 certain fixed string in the binary changed the value of another string that
4065 happened to be identical to the end of the original first string.
4067 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4068 name", it returns that address as the IP address. On some operating
4069 systems (e.g. Solaris), it also passes back the IP address string as the
4070 "host name". However, on others (e.g. Linux), it passes back an empty
4071 string. Exim wasn't checking for this, and was changing the host name to an
4072 empty string, assuming it had been canonicized.
4074 5. Although rare, it is permitted to have more than one PTR record for a given
4075 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4076 all the names associated with an address, because they do in Solaris.
4077 However, it seems that they do not in Linux for data that comes from the
4078 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4079 I found this out when I moved to a new Linux workstation and tried to run
4080 the Exim test suite.
4082 To get round this problem I have changed the code so that it now does its
4083 own call to the DNS to look up PTR records when searching for a host name.
4084 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4085 addresses that are only in /etc/hosts are still found.
4087 This behaviour is, however, controlled by an option called host_lookup_
4088 order, which defaults to "bydns:byaddr". If people want to use the other
4089 order, or indeed, just use one or the other means of lookup, they can
4090 specify it in this variable.
4092 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4093 some operating systems, this comes back from gethostbyaddr() as an empty
4094 string, and this is what Exim used to test for. However, it seems that in
4095 other systems, "." is yielded. Exim now tests for this case too.
4097 7. The values of check_spool_space and check_log_space are now held internally
4098 as a number of kilobytes instead of an absolute number of bytes. If a
4099 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4100 kilobyte. This means that much larger values can be stored.
4102 8. Exim monitor: an attempt to get the action menu when not actually pointing
4103 at a message produces an empty menu entitled "No message selected". This
4104 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4105 no entries in it ("Shell widget menu has zero width and/or height"). So I
4106 have added a single, blank menu entry in this case.
4108 9. Added ${quote_local_part.
4110 10. MIME decoding is now applied to the contents of Subject: header lines when
4113 11. Now that a reference to $sender_host_address automatically causes a reverse
4114 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4115 host lookup before query-style lookups in lists that might use this
4116 variable. This has therefore been abolished, and the "net-" prefix is no
4117 longer necessary for query-style lookups.
4119 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4120 have been a typo for LFLAGS, so it has been changed.
4122 13. The install script calls Exim with "-C /dev/null" in order to find the
4123 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4124 to be output. Howeve, since Exim outputs its version number before the
4125 error, it didn't break the script. It just looked ugly. I fixed this by
4126 always allowing "-C /dev/null" if the caller is root.
4128 14. Ignore overlarge ACL variable number when reading spool file - insurance
4129 against a later release with more variables having written the file.
4131 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4132 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4133 this, as well as the form without the "IPv6" on the front.
4135 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4136 OS/Makefile-Darwin file.
4138 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4139 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4141 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4142 variable. However, some OS use ints for inodes. I've added cast to long int
4143 to get rid of the compiler warning.
4145 19. I had forgotten to lock out "/../" in configuration file names when
4146 ALT_CONFIG_PREFIX was set.
4148 20. Routers used for verification do not need to specify transports. However,
4149 if such a router generated a host list, and callout was configured, Exim
4150 crashed, because it could not find a port number from the (non-existent)
4151 transport. It now assumes port 25 in this circumstance.
4153 21. Added the -t option to exigrep.
4155 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4156 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4157 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4158 them is set, LOOKUP_LSEARCH is forced.
4160 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4162 24. Added sender and host information to the "rejected by local_scan()" log
4163 line; previously there was no indication of these.
4165 25. Added .include_if_exists.
4167 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4168 Linux. It turns out that not all file systems support this. Apparently some
4169 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4170 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4171 not supported on the file descriptor, is now ignored when Exim is trying to
4172 sync a directory. This applies only to Linux.
4174 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4176 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4177 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4180 29. Added support for maildirsize files from supplied patch (modified a bit).
4182 30. The use of :fail: followed by an empty string could lead Exim to respond to
4183 sender verification failures with (e.g.):
4185 550 Verification failed for <xxx>
4186 550 Sender verify failed
4188 where the first response line was missing the '-' that indicates it is not
4189 the final line of the response.
4191 31. The loop for finding the name of the user that called Exim had a hardwired
4192 limit of 10; it now uses the value of finduser_retries, which is used for
4193 all other user lookups.
4195 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4198 33. Exim was neglecting to zero errno before one call of strtol() when
4199 expanding a string and expecting an integer value. On some systems this
4200 resulted in spurious "integer overflow" errors. Also, it was casting the
4201 result into an int without checking.
4203 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4204 did not work. The code looks as if it has *never* worked, though it appears
4205 to have been documented since at least releast 1.62. I have made it work.
4207 35. The "timeout_DNS" error in retry rules, also documented since at least
4208 1.62, also never worked. As it isn't clear exactly what this means, and
4209 clearly it isn't a major issue, I have abolished the feature by treating it
4210 as "timeout", and writing a warning to the main and panic logs.
4212 36. The display of retry rules for -brt wasn't always showing the error code
4215 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4216 timeout_connect_A, timeout_connect_MX.
4218 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4219 to the empty sender.
4221 39. The daemon was not analysing the content of -oX till after it had closed
4222 stderr and disconnected from the controlling terminal. This meant that any
4223 syntax errors were only noted on the panic log, and the return code from
4224 the command was 0. By re-arranging the code a little, I've made the
4225 decoding happen first, so such errors now appear on stderr, and the return
4226 code is 1. However, the actual setting up of the sockets still happens in
4227 the disconnected process, so errors there are still only recorded on the
4230 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4231 connections (as happens on some IP stacks) was logged at start up time as
4232 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4233 it from "IPv6 and IPv4", which means that two separate sockets are being
4236 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4237 says whether AF_INET or AF_INET6 was passed as an argument.
4239 42. Exiwhat output was messed up when time zones were included in log
4242 43. Exiwhat now gives more information about the daemon's listening ports,
4243 and whether -tls-on-connect was used.
4245 44. The "port" option of the smtp transport is now expanded.
4247 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4248 silently ignored. Now an error message is written to the main and panic
4251 46. There's a new ACL modifier called "logwrite" which writes to a log file
4252 as soon as it is encountered.
4254 47. Added $local_user_uid and $local_user_gid at routing time.
4256 48. Exim crashed when trying to verify a sender address that was being
4259 49. Exim was recognizing only a space character after ".include". It now also
4260 recognizes a tab character.
4262 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4263 extracting the relevant information from the specification. The man page no
4264 longer contains scrambled data for the -d option, and I've added a section
4265 at the front about calling Exim under different names.
4267 51. Added "extra_headers" argument to the "mail" command in filter files.
4269 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4272 53. Installed eximstats 1.29.
4274 54. Added transport_filter_timeout as a generic transport option.
4276 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4277 Cc: header lines. This was required by RFC 822, but it not required by RFC
4280 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4281 incoming messages that did not have them. Now it does so only if the
4282 message originates locally, that is, if there is no associated remote host
4283 address. When Resent- header lines are present, this applies to the Resent-
4284 lines rather than the non-Resent- lines.
4286 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4287 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4289 58. Messages for configuration errors now include the name of the main
4290 configuration file - useful now that there may be more than one file in a
4291 list (.included file names were always shown).
4293 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4294 for those rare installations that do not start the daemon as root or run it
4295 setuid root. I've cut out the call to initgroups() if the daemon is not
4298 60. The Exim user and group can now be bound into the binary as text strings
4299 that are looked up at the start of Exim's processing.
4301 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4303 62. Added $mailstore_basename variable.
4305 63. Installed patch to sieve.c from Michael Haardt.
4307 64. When Exim failed to open the panic log after failing to open the main log,
4308 the original message it was trying to log was written to stderr and debug
4309 output, but if they were not available (the usual case in production), it
4310 was lost. Now it is written to syslog before the two lines that record the
4311 failures to open the logs.
4313 65. Users' Exim filters run in subprocesses under the user's uid. It is
4314 possible for a "deliver" command or an alias in a "personal" command to
4315 provoke an address rewrite. If logging of address rewriting is configured,
4316 this fails because the process is not running as root or exim. There may be
4317 a better way of dealing with this, but for the moment (because 4.30 needs
4318 to be released), I have disabled address rewrite logging when running a
4319 filter in a non-root, non-exim process.
4325 1. The buildconfig auxiliary program wasn't quoting the value set for
4326 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4327 not defined. This bug was masked in 4.22 by the effect that was fixed in
4330 2. Some messages that were rejected after a message id was allocated were
4331 shown as "incomplete" by exigrep. It no longer does this for messages that
4332 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4334 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4335 have allow_domain_literals set, the ID did not get logged in the <= line.
4336 Domain literals are now always recognized in Message-ID: header lines.
4338 4. The first argument for a ${extract expansion item is the key name or field
4339 number. Leading and trailing spaces in this item were not being ignored,
4340 causing some misleading effects.
4342 5. When deliver_drop_privilege was set, single queue runner processes started
4343 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4344 same command in the process it spins off) were not dropping privilege.
4346 6. When the daemon running as "exim" started a queue runner, it always
4347 re-executed Exim in the spun-off process. This is a waste of effort when
4348 deliver_drop_privilege is set. The new process now just calls the
4349 queue-runner function directly.
4355 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4358 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4359 ignored. Though the use of -oP was forcing the writing of a pid file, it
4360 was always written to the default place.
4362 3. If the message "no IP address found for host xxxx" is generated during
4363 incoming verification, it is now followed by identification of the incoming
4364 connection (so you can more easily find what provoked it).
4366 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4368 5. Added some features to "harden" Exim a bit more against certain attacks:
4370 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4371 be put in Local/Makefile. This is like the never_users runtime option,
4372 but it cannot be overridden. The default setting is "root".
4374 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4375 prefix string with which any file named in a -C command line option
4378 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4379 is retained for -C and -D only if the caller of Exim is root. Without
4380 it, the exim user may also use -C and -D and retain privilege.
4382 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4383 command line option is disabled.
4385 6. Macro names set by the -D option must start with an upper case letter, just
4386 like macro names defined in the configuration file.
4388 7. Added "dereference=" facility to LDAP.
4390 8. Two instances of the typo "uknown" in the source files are fixed.
4392 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4393 the Configure-Makefile script screwed up while processing it.
4395 10. Incorporated PCRE 4.4.
4397 11. The SMTP synchronization check was not operating right at the start of an
4398 SMTP session. For example, it could not catch a HELO sent before the client
4399 waited for the greeting. There is now a check for outstanding input at the
4400 point when the greeting is written. Because of the duplex, asynchronous
4401 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4402 way, but not yet received, when the check is performed.
4404 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4405 on TCP/IP sockets, because this apparently causes some broken clients to
4408 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4409 unchanged) from the Cygwin maintainer.
4411 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4412 was supported instead of testing for mbx. Effectively a typo.
4414 15. The spa authenticator server code was not checking that the input it
4415 received was valid base64.
4417 16. The debug output line for the "set" modifier in ACLs was not showing the
4418 name of the variable that was being set.
4420 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4422 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4423 lookup on its own. Something else was needed to trigger the lookup. For
4424 example, a match in host_lookup or the need for a host name in a host list.
4425 Now, if $sender_host_name is referenced and the host name has not yet been
4426 looked up, a lookup is performed. If the lookup fails, the variable remains
4427 empty, and $host_lookup_failed is set to "1".
4429 19. Added "eqi" as a case-independent comparison operator.
4431 20. The saslauthd authentication condition could segfault if neither service
4432 nor realm was specified.
4434 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4435 error message that was logged was misleading, and incoming SMTP
4436 connections were dropped. The message is now more accurate, and temporary
4437 errors are given to SMTP connections.
4439 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4440 (except RSET) with a 421 error, until QUIT is received. However, it was
4441 failing to send a response to QUIT.
4443 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4444 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4445 ACL is now run after the helo_try_verify_hosts code.
4447 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4448 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4449 case-independent, so other case variants are also recognized). Apparently
4450 some systems use these upper case variants.
4452 25. If more than two messages were waiting for the same host, and a transport
4453 filter was specified for the transport, Exim sent two messages over the
4454 same TCP/IP connection, and then failed with "socket operation on non-
4455 socket" when it tried to send the third.
4457 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4459 27. The extern definition of crypt16() in expand.c was not being excluded when
4460 the OS had its own crypt16() function.
4462 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4463 as a preferred synonym for return_size_limit, both as an option and as an
4466 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4468 30. Changed the default configuration ACL to relax the local part checking rule
4469 for addresses that are not in any local domains. For these addresses,
4470 slashes and pipe symbols are allowed within local parts, but the sequence
4471 /../ is explicitly forbidden.
4473 31. SPA server authentication was not clearing the challenge buffer before
4476 32. log_message in a "warn" ACL statement was writing to the reject log as
4477 well as to the main log, which contradicts the documentation and doesn't
4478 seem right (because no rejection is happening). So I have stopped it.
4480 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4481 However, I am unable to do any testing of this.
4483 34. Fixed an infelicity in the appendfile transport. When checking directories
4484 for a mailbox, to see if any needed to be created, it was accidentally
4485 using path names with one or more superfluous leading slashes; tracing
4486 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4488 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4489 contents are added to the log line that is written for every discarded
4490 recipient. (Previously a log_message setting was ignored.)
4492 36. The ${quote: operator now quotes the string if it is empty.
4494 37. The install script runs exim in order to find its version number. If for
4495 some reason other than non-existence or emptiness, which it checks, it
4496 could not run './exim', it was installing it with an empty version number,
4497 i.e. as "exim-". This error state is now caught, and the installation is
4500 38. An argument was missing from the function that creates an error message
4501 when Exim fails to connect to the socket for saslauthd authentication.
4502 This could cause Exim to crash, or give a corrupted message.
4504 39. Added isip, isip4, and isip6 to ${if conditions.
4506 40. The ACL variables $acl_xx are now saved with the message, and can be
4507 accessed later in routers, transports, and filters.
4509 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4510 strings in the file are not string-expanded.
4512 42. If a MAIL command specified a SIZE value that was too large to fit into an
4513 int variable, the check against message_size_limit failed. Such values are
4514 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4515 day this will have to be increased, but I don't think I want to be around
4516 when emails are that large.
4523 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4524 iconv() is not standard in FreeBSD.
4526 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4527 IPv6 enabled. The observed symptom was a segmentation fault on return from
4528 the function os_common_find_running_interfaces() in src/os.c.
4530 3. In the check_special_case() function in daemon.c I had used "errno" as an
4531 argument name, which causes warnings on some systems. This was basically a
4532 typo, since it was named "eno" in the comments!
4534 4. The code that waits for the clock to tick (at a resolution of some fraction
4535 of a second) so as to ensure message-id uniqueness was always waiting for
4536 at least one whole tick, when it could have waited for less. [This is
4537 almost certainly not relevant at current processor speeds, where it is
4538 unlikely to ever wait at all. But we try to future-proof.]
4540 5. The function that sleeps for a time interval that includes fractions of a
4541 second contained a race. It did not block SIGALRM between setting the
4542 timer, and suspending (a couple of lines later). If the interval was short
4543 and the sigsuspend() was delayed until after it had expired, the suspension
4544 never ended. On busy systems this could lead to processes getting stuck for
4547 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4548 process, before it forks any delivery processes. The open lookup caching
4549 mechanism meant that the open file or database connection was passed into
4550 the delivery process. The problem was that delivery processes always tidy
4551 up cached lookup data. This could cause a problem for the next delivery
4552 process started by the queue runner, because the external queue runner
4553 process does not know about the closure. So the next delivery process
4554 still has data in the lookup cache. In the case of a file lookup, there was
4555 no problem because closing a file descriptor in a subprocess doesn't affect
4556 the parent. However, if the lookup was caching a connection to a database,
4557 the connection was closed, and the second delivery process was likely to
4558 see errors such as "PGSQL: query failed: server closed the connection
4559 unexpectedly". The problem has been fixed by closing all cached lookups
4560 in a queue runner before running a delivery process.
4562 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4563 seem to have the "const" qualifier which it has on other OS. I've
4566 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4567 *of the same type* (client or server) with the same public name that an
4568 error should be diagnosed.
4570 9. When Exim looked up a host name for an IP address, but failed to find the
4571 original IP address when looking up the host name (a safety check), it
4572 output the message "<ip address> does not match any IP for NULL", which was
4573 confusing, to say the least. The bug was that the host name should have
4574 appeared instead of "NULL".
4576 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4577 user that is built into the binary, and the -C or -D options is used, root
4578 privilege is dropped before the configuration file is read. In addition,
4579 logging is switched to stderr instead of the normal log files. If the
4580 configuration then re-defines the Exim user, the unprivileged environment
4581 is probably not what is expected, so Exim logs a panic warning message (but
4584 However, if deliver_drop_privilege is set, the unprivileged state may well
4585 be exactly what is intended, so the warning has been cut out in that case,
4586 and Exim is allowed to try to write to its normal log files.
4592 1. smtp_return_error_details was not giving details for temporary sender
4593 or receiver verification errors.
4595 2. Diagnose a configuration error if two authenticators have the same public
4598 3. Exim used not to create the message log file for a message until the first
4599 delivery attempt. This could be confusing when incoming messages were held
4600 for policy or load reasons. The message log file is now created at the time
4601 the message is received, and an initial "Received" line is written to it.
4603 4. The automatically generated man page for command line options had a minor
4604 bug that caused no ill effects; however, a more serious problem was that
4605 the procedure for building the man page automatically didn't always
4606 operate. Consequently, release 4.20 contains an out-of-date version. This
4607 shouldn't happen again.
4609 5. When building Exim with embedded Perl support, the script that builds the
4610 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4611 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4613 6. The freeze_tell option was not being used for messages that were frozen on
4614 arrival, either by an ACL or by local_scan().
4616 7. Added the smtp_incomplete_transaction log selector.
4618 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4619 was accepting AUTH without a new EHLO.
4621 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4622 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4625 10. Logging of TCP/IP connections (when configured) now happens in the main
4626 daemon process instead of the child process, so that the TCP/IP connection
4627 count is more accurate (but it can never be perfect).
4629 11. The use of "drop" in a nested ACL was not being handled correctly in the
4630 outer ACL. Now, if condition failure induced by the nested "drop" causes
4631 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4632 or "require"), the connection is dropped.
4634 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4635 that yield "discard" can now be used with an "accept" or a "discard" verb,
4636 but an error is generated for any others (because I can't see a useful way
4637 to define what should happen).
4639 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4640 are now processed in the same way as lines in the Exim configuration file.
4641 In particular, continuation lines are supported.
4643 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4645 15. Added -ti meaning -t -i.
4647 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4648 domains, and warn by logging if others are found.
4650 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4651 returned from GIFCONF when Exim is trying to find the list of interfaces on
4652 a host. The code in os.c has been modified to copy each ifreq to an aligned
4653 structure in all cases.
4655 Also, in some cases, the returned ifreq's were being copied to a 'struct
4656 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4657 means the last couple of bytes of an IPv6 address could be chopped if the
4658 ifreq contained only a normal sockaddr (14 bytes storage).
4660 18. Named domain lists were not supported in the hosts_treat_as_local option.
4661 An entry such as +xxxx was not recognized, and was treated as a literal
4664 19. Ensure that header lines added by a DATA ACL are included in the reject log
4665 if the ACL subsequently rejects the message.
4667 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4668 MD5 (which is deprecated).
4670 21. When testing a filter file using -bf, Exim was writing a message when it
4671 took the sender from a "From " line in the message, but it was not doing so
4672 when it took $return_path from a Return-Path: header line. It now does.
4674 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4675 with a valid header line field name (a series of printing characters
4676 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4678 23. Changed "disc" in the source to "disk" to conform to the documentation and
4679 the book and for uniformity.
4681 24. Ignore Sendmail's -Ooption=value command line item.
4683 25. When execve() failed while trying to run a command in a pipe transport,
4684 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4685 could be confused with a return value of 69 from the command itself. This
4686 has been changed to 127, the value the shell returns if it is asked to run
4687 a non-existent command. The wording for the related log line suggests a
4688 non-existent command as the problem.
4690 26. If received_header_text expands to an empty string, do not add a Received:
4691 header line to the message. (Well, it adds a token one on the spool, but
4692 marks it "old" so that it doesn't get used or transmitted.)
4694 27. Installed eximstats 1.28 (addition of -nt option).
4696 28. There was no check for failure on the call to getsockname() in the daemon
4697 code. This can fail if there is a shortage of resources on the system, with
4698 ENOMEM, for example. A temporary error is now given on failure.
4700 29. Contrary to the C standard, it seems that in some environments, the
4701 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4702 program. Exim now does this explicitly; it affects the formatting of
4703 timestamps using strftime().
4705 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4706 complaints. I've now initialized all the variables, to avoid this.
4708 32. Header lines added by a system filter were not being "seen" during
4709 transport-time rewrites.
4711 33. The info_callback() function passed to OpenSSL is set up with type void
4712 (*)(SSL *, int, int), as described somewhere. However, when calling the
4713 function (actually a macro) that sets it up, the type void(*)() is
4714 expected. I've put in a cast to prevent warnings from picky compilers.
4716 34. If a DNS black list lookup found a CNAME record, but there were no A
4717 records associated with the domain it pointed at, Exim crashed.
4719 35. If a DNS black list lookup returned more than one A record, Exim ignored
4720 all but the first. It now scans all returned addresses if a particular IP
4721 value is being sought. In this situation, the contents of the
4722 $dnslist_value variable are a list of all the addresses, separated by a
4725 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4726 to accept a host name and all its aliases if the forward lookup for any of
4727 them yielded the IP address of the incoming connection. Now it accepts only
4728 those names whose forward lookup yields the correct IP address. Any other
4729 names are discarded. This closes a loophole whereby a rogue DNS
4730 administrator could create reverse DNS records to break through a
4731 wildcarded host restriction in an ACL.
4733 37. If a user filter or a system filter that ran in a subprocess used any of
4734 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4735 the wrong values were passed to the pipe command ($thisaddress had the
4736 value of $0, $0 had the value of $1, etc). This bug was introduced by
4737 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4739 38. Improved the line breaking for long SMTP error messages from ACLs.
4740 Previously, if there was no break point between 40 and 75 characters, Exim
4741 left the rest of the message alone. Two changes have been made: (a) I've
4742 reduced the minimum length to 35 characters; (b) if it can't find a break
4743 point between 35 and 75 characters, it looks ahead and uses the first one
4744 that it finds. This may give the occasional overlong line, but at least the
4745 remaining text gets split now.
4747 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4748 file descriptors might be low, and that setting 1000 would always raise it.
4749 It turns out that in some environments, the limit is already over 1000 and
4750 that lowering it causes trouble. So now Exim takes care not to decrease it.
4752 40. When delivering a message, the value of $return_path is set to $sender_
4753 address at the start of routing (routers may change the value). By an
4754 oversight, this default was not being set up when an address was tested by
4755 -bt or -bv, which affected the outcome if any router or filter referred to
4758 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4759 log only when "message" or "log_message" are set. However, if one of the
4760 conditions was an address verification, or a call to a nested ACL, the
4761 messages generated by the underlying test were being passed through. This
4762 no longer happens. The underlying message is available in $acl_verify_
4763 message for both "message" and "log_message" expansions, so it can be
4764 passed through if needed.
4766 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4767 new expansion $bh_ to give the encoded byte string without charset
4768 translation. Translation happens only if iconv() is available; HAVE_ICONV
4769 indicates this at build time. HEADERS_CHARSET gives the charset to
4770 translate to; headers_charset can change it in the configuration, and
4771 "headers charset" can change it in an individual filter file.
4773 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4774 that creates RFC 2047 encoded "words" labels them as that charset instead
4775 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4776 expansion operator; (ii) when Exim creates a From: line for a local
4777 message; (iii) when a header line is rewritten to include a "phrase" part.
4779 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4780 buggy, causing it to skip the first lines of messages whose message ID
4781 ended in 'D'. This would not have bitten before Exim release 4.14, because
4782 message IDs were unlikely to end in 'D' before then. The effect was to have
4783 incorrect size information for certain domains.
4785 45. #include "config.h" was missing at the start of the crypt16.c module. This
4786 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4789 46. If there was a timeout during a "random" callout check, Exim treated it as
4790 a failure of the random address, and carried on sending RSET and the real
4791 address. If the delay was just some slowness somewhere, the response to the
4792 original RCPT would be taken as a response to RSET and so on, causing
4793 mayhem of various kinds.
4795 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4796 when I implemented it. It didn't allow for the fact that some option values
4797 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4798 the right test for positive values.
4800 48. Domain names in DNS records are case-independent. Exim always looks them up
4801 in lower case. Some resolvers return domain names in exactly the case they
4802 appear in the zone file, that is, they may contain uppercase letters. Not
4803 all resolvers do this - some return always lower case. Exim was treating a
4804 change of case by a resolver as a change of domain, similar to a widening
4805 of a domain abbreviation. This triggered its re-routing code and so it was
4806 trying to route what was effectively the same domain again. This normally
4807 caused routing to fail (because the router wouldn't handle the domain
4808 twice). Now Exim checks for this case specially, and just changes the
4809 casing of the domain that it ultimately uses when it transmits the message
4812 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4815 50. If a filter generated a file delivery with a non-absolute name (possible if
4816 no home directory exists for the router), the forbid_file option was not
4819 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4820 the existing equality matching.
4822 52. Exim was using ints instead of ino_t variables in some places where it was
4823 dealing with inode numbers.
4825 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4826 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4827 called TMPDIR, and if it finds it is different, it changes its value.
4829 54. The smtp_printf() function is now made available to local_scan() so
4830 additional output lines can be written before returning. There is also an
4831 smtp_fflush() function to enable the detection of a dropped connection.
4832 The variables smtp_input and smtp_batched_input are exported to
4835 55. Changed the default runtime configuration: the message "Unknown user"
4836 has been removed from the ACL, and instead placed on the localuser router,
4837 using the cannot_route_message feature. This means that any verification
4838 failures that generate their own messages won't get overridden. Similarly,
4839 the "Unrouteable address" message that was in the ACL for unverifiable
4840 relay addresses has also been removed.
4842 56. Added hosts_avoid_esmtp to the smtp transport.
4844 57. The exicyclog script was not checking for the esoteric option
4845 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4846 will work only if exicyclog is run under the appropriate euid.
4848 58. Following a discussion on the list, the rules by which Exim recognises line
4849 endings on incoming messages have been changed. The -dropcr and drop_cr
4850 options are now no-ops, retained only for backwards compatibility. The
4851 following line terminators are recognized: LF CRLF CR. However, special
4852 processing applies to CR:
4854 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4855 nor a local message in the state where . is a terminator.
4857 (ii) If a bare CR is encountered in a header line, an extra space is added
4858 after the line terminator so as not to end the header. The reasoning
4859 behind this is that bare CRs in header lines are most likely either
4860 to be mistakes, or people trying to play silly games.
4862 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4863 was being incorrectly given as 18 bytes larger than it should have been.
4864 This is a VOB (very old bug).
4866 60. This may never have affected anything current, but just in case it has:
4867 When the local host is found other than at the start of a list of hosts,
4868 the local host, those with the same MX, and any that follow, are discarded.
4869 When the list in question was part of a longer list of hosts, the following
4870 hosts (not currently being processed) were also being discarded. This no
4871 longer happens. I'm not sure if this situation could ever has previously
4874 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4877 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4878 header. This is something that is recommended in a new Internet Draft, and
4879 is something that is documented as being done by Sendmail. There are two
4880 possible values. For messages generated by the autoreply transport, Exim
4883 Auto-Submitted: auto-replied
4885 whereas for all other generated messages (e.g. bounces) it adds
4887 Auto-Submitted: auto-generated
4889 63. The "personal" condition in filters now includes a test for the
4890 Auto-Submitted: header. If it contains the string "auto-" the message it
4891 not considered personal.
4893 64. Added rcpt_include_affixes as a generic transport option.
4895 65. Added queue_only_override (default true).
4897 66. Added the syslog_duplication option.
4899 67. If what should have been the first header line of a message consisted of
4900 a space followed by a colon, Exim was mis-interpreting it as a header line.
4901 It isn't of course - it is syntactically invalid and should therefore be
4902 treated as the start of the message body. The misbehaviour could have
4903 caused a number of strange effects, including loss of data in subsequent
4904 header lines, and spool format errors.
4906 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4907 client host had authenticated. This control can now be exercised by an ACL
4908 for more flexibility.
4910 69. By default, callouts do not happen when testing with -bh. There is now a
4911 variant, -bhc, which does actually run the callout code, including
4912 consulting and updating the callout cache.
4914 70. Added support for saslauthd authentication, courtesy of Alexander
4917 71. If statvfs() failed on the spool or log directories while checking their
4918 size for availability, Exim confusingly gave the error "space shortage".
4919 Furthermore, in debugging mode it crashed with a floating point exception.
4920 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4921 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4922 serious problem, Exim now writes to the main and panic logs when this
4923 happens, with details of the failure. It then refuses to accept the
4924 incoming message, giving the message "spool directory problem" or "log
4925 directory problem" with a 421 code for SMTP messages.
4927 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4928 0, 1, and 2 exist, because some OS complain for execs without them (see
4929 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4930 descriptors. However, the code omitted to check that the open succeeded,
4931 causing mysterious errors if for some reason the permissions on /dev/null
4932 got screwed. Now Exim writes a message to the main and panic logs, and
4933 bombs out if it can't open /dev/null.
4935 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4936 interact so that it is all more flexible. It is supposed to remain
4937 backwards compatible. Also added extra_local_interfaces.
4939 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4940 to bomb out with an assertion failure - to the client this appears as a
4941 connection drop. This problem occurs in the part of the code that was taken
4942 from the Samba project. Fortunately, the assertion is in a very simple
4943 function, so I have fixed this by reproducing the function inline in the
4944 one place where it is called, and arranging for authentication to fail
4945 instead of killing the process with assert().
4947 75. The SPA client code was not working when the server requested OEM rather
4948 than Unicode encoding.
4950 76. Added code to make require_files with a specific uid setting more usable in
4951 the case where statting the file as root fails - usually a non-root-mounted
4952 NFS file system. When this happens and the failure is EACCES, Exim now
4953 forks a subprocess and does the per-uid checking as the relevant uid.
4955 77. Added process_log_path.
4957 78. If log_file_path was not explicitly set, a setting of check_log_space or
4958 check_log_inodes was ignored.
4960 79. If a space check for the spool or log partitions fails, the incident is now
4961 logged. Of course, in the latter case the data may get lost...
4963 80. Added the %p formatting code to string_format() so that it can be used to
4964 print addresses in debug_print(). Adjusted all the address printing in the
4965 debugging in store.c to use %p rather than %d.
4967 81. There was a concern that a line of code in smtp_in.c could overflow a
4968 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4969 initially expressed, the concern was not well-founded, because trailing
4970 spaces are removed early. However, if the trailing spaces were followed by
4971 a NULL, they did not get removed, so the overflow was possible. Two fixes
4974 (a) I re-wrote the offending code in a cleaner fashion.
4975 (b) If an incoming SMTP command contains a NULL character, it is rejected
4978 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4979 runs initgroups(), so that if the Exim user is in any additional groups,
4980 they will be used during message reception.
4986 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob