1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.567 2009/10/14 13:43:40 nm4 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TK/01 Added patch by Johannes Berg that expands the main option
10 "spamd_servers" if it starts with a dollar sign.
12 TK/02 Write list of recipients to X-Envelope-Sender header when building
13 the mbox-format spool file for content scanning (suggested by Jakob
16 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
17 (http://www.dcc-servers.net/) support via dccifd. Activated by
18 setting EXPERIMENTAL_DCC=yes in Local/Makefile. Check out
19 experimental_spec.txt for more documentation.
21 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
22 by Mark Daniel Reidel <mr@df.eu>.
24 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
25 When building exim an external PCRE library is now needed -
26 PCRE is a system library on the majority of modern systems.
27 See entry on PCRE_LIBS in EDITME file.
29 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
30 conversation. Added nologin parameter to request.
31 Patch contributed by Kirill Miazine
33 TF/01 Do not log submission mode rewrites if they do not change the address.
35 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
37 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
38 log files in place. Contributed by Roberto Lima
40 NM/04 Bugzilla 667: close socket used by dovecot authenticator
42 TF/03 Bugzilla 615: When checking the local_parts router precondition
43 after a local_part_suffix or local_part_prefix option, Exim now
44 does not use the address's named list lookup cache, since this
45 contains cached lookups for the whole local part.
47 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
48 Robert Millan. Documentation is in experimental-spec.txt
50 TF/04 Bugzilla 668: Fix parallel build (make -j).
52 NM/05 Bugzilla 437: Prevent Maildix aux files being created with mode 000
54 NM/05 Bugzilla 598: Improvement to Dovecot authenticator handling.
55 Patch provided by Jan Srzednicki
57 TF/05 Leading white space used to be stripped from $spam_report which
58 wrecked the formatting. Now it is preserved.
60 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
61 that they are available at delivery time.
63 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
65 TF/08 TLS error reporting now respects the incoming_interface and
66 incoming_port log selectors.
68 TF/09 Produce a more useful error message if an SMTP transport's hosts
69 setting expands to an empty string.
71 NM/06 Bugzilla 744: EXPN did not work under TLS.
72 Patch provided by Phil Pennock
74 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
75 Patch provided by Richard Godbee
77 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
78 acl_smtp_notquit, added index entry.
80 NM/09 Bugzilla 787: Potential buffer overflow in string_format
81 Patch provided by Eugene Bujak
83 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to accept()
84 Patch provided by Maxim Dounin
86 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
87 Patch provided by Phil Pennock
89 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
91 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
92 Patch provided by Brad "anomie" Jorsch
94 NM/14 Bugzilla 614: Added timeout setting to transport filter.
95 Patch provided by Dean Brooks
97 TK/05 Add native DKIM support (does not depend on external libraries).
99 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
100 Patch provided by Graeme Fowler
102 NM/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
112 TK/01 Add preliminary DKIM support. Currently requires a forked version of
113 ALT-N's libdkim that I have put here:
114 http://duncanthrax.net/exim-experimental/
116 Note to Michael Haardt: I had to rename some vars in sieve.c. They
117 were called 'true' and it seems that C99 defines that as a reserved
118 keyword to be used with 'bool' variable types. That means you could
119 not include C99-style headers which use bools without triggering
120 build errors in sieve.c.
122 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
123 as mailq or other aliases. Changed the --help handling significantly
124 to do whats expected. exim_usage() emits usage/help information.
126 SC/01 Added the -bylocaldomain option to eximstats.
128 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
130 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
132 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
138 PH/01 Another patch from the Sieve maintainer.
140 PH/02 When an IPv6 address is converted to a string for single-key lookup
141 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
142 dots are used instead of colons so that keys in lsearch files need not
143 contain colons. This was done some time before quoting was made available
144 in lsearch files. However, iplsearch files do require colons in IPv6 keys
145 (notated using the quote facility) so as to distinguish them from IPv4
146 keys. This meant that lookups for IP addresses in host lists did not work
147 for iplsearch lookups.
149 This has been fixed by arranging for IPv6 addresses to be expressed with
150 colons if the lookup type is iplsearch. This is not incompatible, because
151 previously such lookups could never work.
153 The situation is now rather anomolous, since one *can* have colons in
154 ordinary lsearch keys. However, making the change in all cases is
155 incompatible and would probably break a number of configurations.
157 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
160 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
161 conversion specification without a maximum field width, thereby enabling
162 a rogue spamd server to cause a buffer overflow. While nobody in their
163 right mind would setup Exim to query an untrusted spamd server, an
164 attacker that gains access to a server running spamd could potentially
165 exploit this vulnerability to run arbitrary code as the Exim user.
167 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
168 $primary_hostname instead of what libspf2 thinks the hosts name is.
170 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
171 a directory entry by the name of the lookup key. Previously, if a
172 symlink pointed to a non-existing file or a file in a directory that
173 Exim lacked permissions to read, a lookup for a key matching that
174 symlink would fail. Now it is enough that a matching directory entry
175 exists, symlink or not. (Bugzilla 503.)
177 PH/03 The body_linecount and body_zerocount variables are now exported in the
180 PH/04 Added the $dnslist_matched variable.
182 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
183 This means they are set thereafter only if the connection becomes
186 PH/06 Added the client_condition to authenticators so that some can be skipped
187 by clients under certain conditions.
189 PH/07 The error message for a badly-placed control=no_multiline_responses left
190 "_responses" off the end of the name.
192 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
194 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
195 (without spaces) instead of just copying the configuration text.
197 PH/10 Added the /noupdate option to the ratelimit ACL condition.
199 PH/11 Added $max_received_linelength.
201 PH/12 Added +ignore_defer and +include_defer to host lists.
203 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
204 way in which PCRE > 7.0 is built.
206 PH/14 Implemented queue_only_load_latch.
208 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
209 MAIL command. The effect was to mangle the value on 64-bit systems.
211 PH/16 Another patch from the Sieve maintainer.
213 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
215 PH/18 If a system quota error occurred while trying to create the file for
216 a maildir delivery, the message "Mailbox is full" was not appended to the
217 bounce if the delivery eventually timed out. Change 4.67/27 below applied
218 only to a quota excession during the actual writing of the file.
220 PH/19 It seems that peer DN values may contain newlines (and other non-printing
221 characters?) which causes problems in log lines. The DN values are now
222 passed through string_printing() before being added to log lines.
224 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
225 and InterBase are left for another time.)
227 PH/21 Added message_body_newlines option.
229 PH/22 Guard against possible overflow in moan_check_errorcopy().
231 PH/23 POSIX allows open() to be a macro; guard against that.
233 PH/24 If the recipient of an error message contained an @ in the local part
234 (suitably quoted, of course), incorrect values were put in $domain and
235 $local_part during the evaluation of errors_copy.
241 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
242 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
245 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
246 issue a MAIL command.
248 PH/02 In an ACL statement such as
250 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
252 if a client was not listed at all, or was listed with a value other than
253 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
254 the condition was not true (as it should be), so access was not denied.
255 The bug was that the ! inversion was incorrectly passed on to the second
256 item. This has been fixed.
258 PH/03 Added additional dnslists conditions == and =& which are different from
259 = and & when the dns lookup returns more than one IP address.
261 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
262 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
264 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
265 FSYNC, which compiles an option called disable_fsync that allows for
266 bypassing fsync(). The documentation is heavily laced with warnings.
268 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
270 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
271 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
272 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
273 including adding "make clean"; (3) Added -fPIC when compiling the test
274 dynamically loaded module, to get rid of a warning.
276 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
277 message fails, move_frozen_messages = true and ignore_bounce_errors_after
278 = 0s. The bug is otherwise harmless.
280 PH/07 There was a bug in the dovecot authenticator such that the value of
281 $auth1 could be overwritten, and so not correctly preserved, after a
282 successful authentication. This usually meant that the value preserved by
283 the server_setid option was incorrect.
285 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
287 PH/09 Installed PCRE release 7.0.
289 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
290 run for batched SMTP input. It is now run at the start of every message
291 in the batch. While fixing this I discovered that the process information
292 (output by running exiwhat) was not always getting set for -bs and -bS
293 input. This is fixed, and it now also says "batched" for BSMTP.
295 PH/11 Added control=no_pipelining.
297 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
298 patch, slightly modified), and move the expansion of helo_data till after
299 the connection is made in the smtp transport (so it can use these
302 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
304 PH/14 Added log_selector = +pid.
306 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
308 PH/16 Add ${if forany and ${if forall.
310 PH/17 Added dsn_from option to vary the From: line in DSNs.
312 PH/18 Flush SMTP output before performing a callout, unless control =
313 no_callout_flush is set.
315 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
316 was true (the default) a successful delivery failed to delete the retry
317 item, thus causing premature timeout of the address. The bug is now
320 PH/20 Added hosts_avoid_pipelining to the smtp transport.
322 PH/21 Long custom messages for fakedefer and fakereject are now split up
323 into multiline reponses in the same way that messages for "deny" and
324 other ACL rejections are.
326 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
327 with slight modification.
329 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
330 draft, changing the syntax and factoring some duplicate code".
332 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
333 for deliveries of the second and subsequent messages over the same SMTP
336 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
337 ${reduce, with only minor "tidies".
339 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
341 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
342 expansion side effects.
344 PH/27 When a message times out after an over-quota error from an Exim-imposed
345 quota, the bounce message says "mailbox is full". This message was not
346 being given when it was a system quota that was exceeded. It now should
349 MH/03 Made $recipients available in local_scan(). local_scan() already has
350 better access to the recipient list through recipients_list[], but
351 $recipients can be useful in postmaster-provided expansion strings.
353 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
354 in the case of a MAIL command with additional options following the
355 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
356 were accidentally chopped off.
358 PH/29 SMTP synchronization checks are implemented when a command is read -
359 there is a check that no more input is waiting when there shouldn't be
360 any. However, for some commands, a delay in an ACL can mean that it is
361 some time before the response is written. In this time, more input might
362 arrive, invalidly. So now there are extra checks after an ACL has run for
363 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
364 pipelining has not been advertised.
366 PH/30 MH's patch to allow iscntrl() characters to be list separators.
368 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
369 returned in the SMTP response when smtp_return_error_details was false.
372 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
373 instead of the C I/O that was originally supplied, because problems were
376 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
377 Exim which did not show up earlier: it was assuming that a call to
378 SSL_CTX_set_info_callback() might give an error value. In fact, there is
379 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
380 was a macro that became an assignment, so it seemed to work. This has
381 changed to a proper function call with a void return, hence the compile
382 error. Exim's code has been fixed.
384 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
387 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
389 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
391 PH/37 If a message is not accepted after it has had an id assigned (e.g.
392 because it turns out to be too big or there is a timeout) there is no
393 "Completed" line in the log. When some messages of this type were
394 selected by exigrep, they were listed as "not completed". Others were
395 picked up by some special patterns. I have improved the selection
396 criteria to be more general.
398 PH/38 The host_find_failed option in the manualroute router can now be set
399 to "ignore", to completely ignore a host whose IP address cannot be
400 found. If all hosts are ignored, the behaviour is controlled by the new
401 host_all_ignored option.
403 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
404 homing or because of multiple MX records with /mx) generated more than
405 one IP address, and the following item turned out to be the local host,
406 all the secondary addresses of the first item were incorrectly removed
407 from the list, along with the local host and any following hosts (which
408 is what is supposed to happen).
410 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
411 whoever called Exim into the -H file. In the case of the daemon it was
412 behaving confusingly. When first started, it used values for whoever
413 started the daemon, but after a SIGHUP it used the Exim user (because it
414 calls itself on a restart). I have changed the code so that it now always
417 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
418 message are rejected with the same error (e.g. no authentication or bad
419 sender address), and a DATA command is nevertheless sent (as can happen
420 with PIPELINING or a stupid MUA), the error message that was given to the
421 RCPT commands is included in the rejection of the DATA command. This is
422 intended to be helpful for MUAs that show only the final error to their
425 PH/42 Another patch from the Sieve maintainer.
427 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
428 Eximstats - Fixed some broken HTML links and added missing column headers
430 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
431 columns for Rejects, Temp Rejects, Ham, and Spam rows.
433 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
435 PH/43 Yet another patch from the Sieve maintainer.
437 PH/44 I found a way to check for a TCP/IP connection going away before sending
438 the response to the final '.' that terminates a message, but only in the
439 case where the client has not sent further data following the '.'
440 (unfortunately, this is allowed). However, in many cases there won't be
441 any further data because there won't be any more messages to send. A call
442 to select() can be used: if it shows that the input is "ready", there is
443 either input waiting, or the socket has been closed. An attempt to read
444 the next input character can distinguish the two cases. Previously, Exim
445 would have sent an OK response which the client would never have see.
446 This could lead to message repetition. This fix should cure that, at
447 least in a lot of common cases.
449 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
450 advertised in response to EHLO.
456 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
457 fixed by 4.65/MH/01 (is this a record?) are fixed:
459 (i) An empty string was always treated as zero by the numeric comparison
460 operators. This behaviour has been restored.
462 (ii) It is documented that the numeric comparison operators always treat
463 their arguments as decimal numbers. This was broken in that numbers
464 starting with 0 were being interpreted as octal.
466 While fixing these problems I realized that there was another issue that
467 hadn't been noticed. Values of message_size_limit (both the global option
468 and the transport option) were treated as octal if they started with 0.
469 The documentation was vague. These values are now always treated as
470 decimal, and I will make that clear in the documentation.
476 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
477 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
480 MH/01 Don't check that the operands of numeric comparison operators are
481 integers when their expansion is in "skipping" mode (fixes bug
482 introduced by 4.64-PH/07).
484 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
485 child addresses, Exim now panics and dies. Previously, because the count
486 is held in a short int, deliveries were likely to be lost. As such a
487 large number of recipients for a single message is ridiculous
488 (performance will be very, very poor), I have chosen to impose a limit
489 rather than extend the field.
495 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
496 leftover -K file (the existence of which was triggered by #402).
497 While we were at it, introduced process PID as part of the -K
498 filename. This should rule out race conditions when creating
501 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
502 processing considerably. Previous code took too long for large mails,
503 triggering a timeout which in turn triggers #401.
505 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
506 in the DK code in transports.c. sendfile() is not really portable,
507 hence the _LINUX specificness.
509 TF/01 In the add_headers option to the mail command in an Exim filter,
510 there was a bug that Exim would claim a syntax error in any
511 header after the first one which had an odd number of characters
514 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
515 callout verification, Exim cached a "reject" for the entire domain. This
516 is correct for most verifications, but it is not correct for a recipient
517 verification with use_sender or use_postmaster set, because in that case
518 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
519 case of MAIL FROM:<> rejection from other early rejections (e.g.
520 rejection of HELO). When verifying a recipient using a non-null MAIL
521 address, the cache is ignored if it shows MAIL FROM:<> rejection.
522 Whatever the result of the callout, the value of the domain cache is
523 left unchanged (for any other kind of callout, getting as far as trying
524 RCPT means that the domain itself is ok).
526 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
529 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
530 manifest itself as EPIPE rather than ECONNECT. When tidying away a
531 session, the daemon ignores ECONNECT errors and logs others; it now
532 ignores EPIPE as well.
534 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
535 (quoted-printable decoding).
537 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
538 later the small subsequent patch to fix an introduced bug.
540 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
542 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
544 PH/08 An error is now given if message_size_limit is specified negative.
546 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
547 to be given (somewhat) arbitrary names.
549 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
552 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
553 miscellaneous code fixes
555 PH/10 Added the log_reject_target ACL modifier to specify where to log
558 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
559 hostname. This is wrong, because it relates to the incoming message (and
560 probably the interface on which it is arriving) and not to the outgoing
561 callout (which could be using a different interface). This has been
562 changed to use the value of the helo_data option from the smtp transport
563 instead - this is what is used when a message is actually being sent. If
564 there is no remote transport (possible with a router that sets up host
565 addresses), $smtp_active_hostname is used.
567 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
568 tweaks were necessary in order to get it to work (see also 21 below):
569 (a) The code assumed that strncpy() returns a negative number on buffer
570 overflow, which isn't the case. Replaced with Exim's string_format()
572 (b) There were several signed/unsigned issues. I just did the minimum
573 hacking in of casts. There is scope for a larger refactoring.
574 (c) The code used strcasecmp() which is not a standard C function.
575 Replaced with Exim's strcmpic() function.
576 (d) The code set only $1; it now sets $auth1 as well.
577 (e) A simple test gave the error "authentication client didn't specify
578 service in request". It would seem that Dovecot has changed its
579 interface. Fortunately there's a specification; I followed it and
580 changed what the client sends and it appears to be working now.
582 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
585 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
586 address A is aliased to B and C, where B exists and C does not. Without
587 -v the output is "A verified" because verification stops after a
588 successful redirection if more than one address is generated. However,
589 with -v the child addresses are also verified. Exim was outputting "A
590 failed to verify" and then showing the successful verification for C,
591 with its parentage. It now outputs "B failed to verify", showing B's
592 parentage before showing the successful verification of C.
594 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
595 look up a TXT record in a specific list after matching in a combined
598 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
599 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
600 they consult the DNS. I had assumed they would set it the way they
601 wanted; and indeed my experiments on Linux seem to show that in some
602 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
603 To be on the safe side, however, I have now made the interface to
604 host_find_byname() similar to host_find_bydns(), with an argument
605 containing the DNS resolver options. The host_find_byname() function now
606 sets these options at its start, just as host_find_bydns() does. The smtp
607 transport options dns_qualify_single and dns_search_parents are passed to
608 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
609 of host_find_byname() use the default settings of RES_DEFNAMES
610 (qualify_single) but not RES_DNSRCH (search_parents).
612 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
613 spool_read_header() do less string testing, by means of a preliminary
614 switch on the second character of optional "-foo" lines. (This is
615 overdue, caused by the large number of possibilities that now exist.
616 Originally there were few.) While I was there, I also converted the
617 str(n)cmp tests so they don't re-test the leading "-" and the first
618 character, in the hope this might squeeze out yet more improvement.
620 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
621 flag allowing group syntax was set by the header_syntax check but not
622 turned off, possible causing trouble later; (2) The flag was not being
623 set at all for the header_verify test, causing "group"-style headers to
624 be rejected. I have now set it in this case, and also caused header_
625 verify to ignore an empty address taken from a group. While doing this, I
626 came across some other cases where the code for allowing group syntax
627 while scanning a header line wasn't quite right (mostly, not resetting
628 the flag correctly in the right place). These bugs could have caused
629 trouble for malformed header lines. I hope it is now all correct.
631 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
632 with the "reply" argument non-NULL. The code, however (which originally
633 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
634 but it didn't always do it. This confused somebody who was copying the
635 code for some other use. I have removed all the tests.
637 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
638 feature that was used to support insecure browsers during the U.S. crypto
639 embargo. It requires special client support, and Exim is probably the
640 only MTA that supported it -- and would never use it because real RSA is
641 always available. This code has been removed, because it had the bad
642 effect of slowing Exim down by computing (never used) parameters for the
643 RSA_EXPORT functionality.
645 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
646 authenticator to fail if there's a tab character in the incoming data
647 (there should never be unless someone is messing about, as it's supposed
648 to be base64-encoded). Also added, on Timo's advice, the "secured" option
649 if the connection is using TLS or if the remote IP is the same as the
650 local IP, and the "valid-client-cert option" if a client certificate has
653 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
654 authenticators. This can be used for authorization after authentication
655 succeeds. (In the case of plaintext, it servers for both authentication
658 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
659 if any retry times were supplied.
661 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
662 connection, where there is no client IP address to check. In this
663 situation, the verify now always succeeds.
665 PH/25 Applied John Jetmore's -Mset patch.
667 PH/26 Added -bem to be like -Mset, but loading a message from a file.
669 PH/27 In a string expansion for a processed (not raw) header when multiple
670 headers of the same name were present, leading whitespace was being
671 removed from all of them, but trailing whitespace was being removed only
672 from the last one. Now trailing whitespace is removed from each header
673 before concatenation. Completely empty headers in a concatenation (as
676 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
677 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
679 PH/29 [Removed. This was a change that I later backed out, and forgot to
680 correct the ChangeLog entry (that I had efficiently created) before
681 committing the later change.]
683 PH/30 Exim was sometimes attempting to deliver messages that had suffered
684 address errors (4xx response to RCPT) over the same connection as other
685 messages routed to the same hosts. Such deliveries are always "forced",
686 so retry times are not inspected. This resulted in far too many retries
687 for the affected addresses. The effect occurred only when there were more
688 hosts than the hosts_max_try setting in the smtp transport when it had
689 the 4xx errors. Those hosts that it had tried were not added to the list
690 of hosts for which the message was waiting, so if all were tried, there
691 was no problem. Two fixes have been applied:
693 (i) If there are any address or message errors in an SMTP delivery, none
694 of the hosts (tried or untried) are now added to the list of hosts
695 for which the message is waiting, so the message should not be a
696 candidate for sending over the same connection that was used for a
697 successful delivery of some other message. This seems entirely
698 reasonable: after all the message is NOT "waiting for some host".
699 This is so "obvious" that I'm not sure why it wasn't done
700 previously. Hope I haven't missed anything, but it can't do any
701 harm, as the worst effect is to miss an optimization.
703 (ii) If, despite (i), such a delivery is accidentally attempted, the
704 routing retry time is respected, so at least it doesn't keep
705 hammering the server.
707 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
708 in ${readsocket because some servers need this prod.
710 PH/32 Added some extra debug output when updating a wait-xxx database.
712 PH/33 The hint "could be header name not terminated by colon", which has been
713 given for certain expansion errors for a long time, was not being given
714 for the ${if def:h_colon_omitted{... case.
716 PH/34 The spec says: "With one important exception, whenever a domain list is
717 being scanned, $domain contains the subject domain." There was at least
718 one case where this was not true.
720 PH/35 The error "getsockname() failed: connection reset by peer" was being
721 written to the panic log as well as the main log, but it isn't really
722 panic-worthy as it just means the connection died rather early on. I have
723 removed the panic log writing for the ECONNRESET error when getsockname()
726 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
727 runs only) independently of the message's sender address. This meant
728 that, if the 4xx error was in fact related to the sender, a different
729 message to the same recipient with a different sender could confuse
730 things. In particualar, this can happen when sending to a greylisting
731 server, but other circumstances could also provoke similar problems.
732 I have changed the default so that the retry time for these errors is now
733 based a combination of the sender and recipient addresses. This change
734 can be overridden by setting address_retry_include_sender=false in the
737 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
738 remote server are returned as part of bounce messages. This was not
739 happening for LMTP over a pipe (the lmtp transport), but now it is the
740 same for both kinds of LMTP.
742 PH/38 Despite being documented as not happening, Exim was rewriting addresses
743 in header lines that were in fact CNAMEs. This is no longer the case.
745 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
746 and queue runs started by the daemon processed all messages. This has
747 been fixed so that -R and -S can now usefully be given with -q<time>.
749 PH/40 Import PCRE release 6.7 (fixes some bugs).
751 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
753 PH/42 Give an error if -q is specified more than once.
755 PH/43 Renamed the variables $interface_address and $interface_port as
756 $received_ip_address and $received_port, to make it clear that these
757 values apply to message reception, and not to the outgoing interface when
758 a message is delivered. (The old names remain recognized, of course.)
760 PH/44 There was no timeout on the connect() call when using a Unix domain
761 socket in the ${readsocket expansion. There now is.
763 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
764 be meaningful with "accept".
766 SC/01 Eximstats V1.43
767 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
769 SC/02 Eximstats V1.44
770 Use a glob alias rather than an array ref in the generated
771 parser. This improves both readability and performance.
773 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
774 Collect SpamAssassin and rejection statistics.
775 Don't display local sender or destination tables unless
776 there is data to show.
777 Added average volumes into the top table text output.
779 SC/04 Eximstats V1.46
780 Collect data on the number of addresses (recipients)
781 as well as the number of messages.
783 SC/05 Eximstats V1.47
784 Added 'Message too big' to the list of mail rejection
785 reasons (thanks to Marco Gaiarin).
787 SC/06 Eximstats V1.48
788 Mainlog lines which have GMT offsets and are too short to
789 have a flag are now skipped.
791 SC/07 Eximstats V1.49 (Alain Williams)
792 Added the -emptyok flag.
794 SC/08 Eximstats V1.50
795 Fixes for obtaining the IP address from reject messages.
797 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
798 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
799 whitesspace changes from 4.64-PH/27
801 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
804 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
805 are found, allow negative numbers in numeric criteria)
807 JJ/06 exipick.20061117.2, added new $message_body_missing variable
809 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
810 to match changes made in 4.64-PH/43
812 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
814 PH/47 Put in an explicit test for a DNS lookup of an address record where the
815 "domain" is actually an IP address, and force a failure. This locks out
816 those revolvers/nameservers that support "A-for-A" lookups, in
817 contravention of the specifications.
819 PH/48 When a host name was looked up from an IP address, and the subsequent
820 forward lookup of the name timed out, the host name was left in
821 $sender_host_name, contrary to the specification.
823 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
824 restricted to single-key lookups, Exim was not diagnosing an error if
825 * or *@ was used with a query-style lookup.
827 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
829 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
830 long ago, but noone interested enough thought of it. Let's just say that
831 the "1.1" means that there are some new functions that weren't there at
832 some point in the past.
834 PH/51 Error processing for expansion failure of helo_data from an smtp
835 transport during callout processing was broken.
837 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
838 tested/used via the -bh/-bhc/-bs options.
840 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
841 bug, fixed in subsequent PCRE releases).
843 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
844 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
846 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
852 SC/01 Use a glob alias rather than an array ref in eximstats generated
853 parser. This improves both readability and performance.
855 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
856 Don't display local sender or destination tables in eximstats unless
857 there is data to show.
858 Added average volumes into the eximstats top table text output.
860 SC/03 Collect data on the number of addresses (recipients) as well
861 as the number of messages in eximstats.
863 TF/01 Correct an error in the documentation for the redirect router. Exim
864 does (usually) call initgroups() when daemonizing.
866 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
867 with consistent privilege compared to when running as a daemon.
869 TF/03 Note in the spec that $authenticated_id is not set for local
870 submissions from trusted users.
872 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
873 Thanks to Dean Brooks <dean@iglou.com> for the patch.
875 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
876 by adding some example configuration directives to the default
877 configuration file. A little bit of work is required to uncomment the
878 directives and define how usernames and passwords are checked, but
879 there is now a framework to start from.
881 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
882 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
883 without this. I don't know how relevant this is to other LDAP libraries.
885 PH/02 Add the verb name to the "unknown ACL verb" error.
887 PH/03 Magnus Holmgren's patch for filter_prepend_home.
889 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
891 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
892 directory not expanded when it should be if an expanded home directory
893 was set for the address (which is overridden by the transport).
895 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
898 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
899 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
900 because it is too late at that time, and has no effect.
902 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
903 security issue with \' (bugzilla #107). I could not use the
904 PQescapeStringConn() function, because it needs a PGconn value as one of
907 PH/08 When testing addresses using -bt, indicate those final addresses that
908 are duplicates that would not cause an additional delivery. At least one
909 person was confused, thinking that -bt output corresponded to deliveries.
910 (Suppressing duplicates isn't a good idea as you lose the information
911 about possibly different redirections that led to the duplicates.)
913 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
914 systems where poll() doesn't work, in particular OS X.
916 PH/10 Added more information to debugging output for retry time not reached.
918 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
919 operations in malware.c.
921 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
924 PH/13 If write_rejectlog was set false when logging was sent to syslog with
925 syslog_duplication set false, log lines that would normally be written
926 both the the main log and to the reject log were not written to syslog at
929 PH/14 In the default configuration, change the use of "message" in ACL warn
930 statements to "add_header".
932 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
933 not followed by a command (e.g. "seen endif").
935 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
936 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
939 PH/17 Added extra conditions to the default value of delay_warning_condition
943 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
944 { match{$h_precedence:}{(?i)bulk|list|junk} } \
945 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
948 The Auto-Submitted: and various List- headers are standardised, whereas I
949 don't think Precedence: ever was.
951 PH/18 Refactored debugging code in route_finduser() to show more information,
952 in particular, the error code if getpwnam() issues one.
954 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
955 This is apparently needed in addition to the PH/07 change above to avoid
956 any possible encoding problems.
958 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
959 but not after initializing Perl.
961 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
962 output them only if debugging. By default they are written stderr,
963 apparently, which is not desirable.
965 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
968 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
971 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
973 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
974 authenticated or an ident call has been made. Suppress the default
975 values for $authenticated_id and $authenticated_sender (but permit -oMai
976 and -oMas) when testing with -bh.
978 PH/24 Re-jigged the order of the tests in the default configuration so that the
979 tests for valid domains and recipients precede the DNS black list and CSA
980 tests, on the grounds that those ones are more expensive.
982 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
983 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
984 This bug exists in every version of Exim that I still have, right back to
987 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
988 However, an attempt to turn on case-sensitivity in a regex key by
989 including (?-i) didn't work because the subject string was already
990 lowercased, and the effects were non-intuitive. It turns out that a
991 one-line patch can be used to allow (?-i) to work as expected.
997 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
998 other effects) broke the use of negated acl sub-conditions.
1000 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1003 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1004 "Deny" causes Exim to reject the incoming connection with a 554 error.
1005 Unfortunately, if there is a major crisis, such as a disk failure,
1006 tcp-wrappers gives "deny", whereas what one would like would be some
1007 kind of temporary error. A kludge has been added to help with this.
1008 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1009 554 error is used if errno is still zero or contains ENOENT (which occurs
1010 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1013 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1015 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1016 errors. Otherwise a message that provokes a temporary error (when other
1017 messages do not) can cause a whole host to time out.
1019 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1020 addresses were routed directly to files or pipes from a redirect router.
1021 File deliveries just didn't batch; pipe deliveries might have suffered
1024 PH/06 A failure to get a lock for a hints database would erroneously always say
1025 "Failed to get write lock", even when it was really a read lock.
1027 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1028 of 0600. This has been changed to use the value of the lockfile_mode
1029 option (which defaults to 0600).
1031 PH/08 Applied small patch from the Sieve maintainer.
1033 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1034 folder from quota calculations, a direct delivery into this folder messed
1035 up the contents of the maildirsize file. This was because the regex was
1036 used only to exclude .Trash (or whatever) when the size of the mailbox
1037 was calculated. There was no check that a delivery was happening into an
1038 excluded directory. This bug has been fixed by ignoring all quota
1039 processing for deliveries into excluded directories.
1041 PH/10 Added the maildirfolder_create_regex option to appendfile.
1047 PH/01 The code for finding all the local interface addresses on a FreeBSD
1048 system running IPv6 was broken. This may well have applied to all BSD
1049 systems, as well as to others that have similar system calls. The broken
1050 code found IPv4 interfaces correctly, but gave incorrect values for the
1051 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1052 that it would not match correctly against @[] and not recognize the IPv6
1055 PH/02 The ipliteral router was not recognizing addresses of the form user@
1056 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1058 PH/03 Added disable_ipv6.
1060 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1061 decoded form, because it is most often used to construct To: headers
1062 lines in autoreplies, and the decoded form may well be syntactically
1063 invalid. However, $reply_address has leading white space removed, and all
1064 newlines turned into spaces so that the autoreply transport does not
1067 PH/05 If group was specified without a user on a router, and no group or user
1068 was specified on a transport, the group from the router was ignored.
1070 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1071 for visible compile-time settings that can be used to change these
1072 numbers, for those that want even more. Backwards compatibility with old
1073 spool files has been maintained. However, going back to a previous Exim
1074 release will lost any variables that are in spool files.
1076 PH/07 Two small changes when running in the test harness: increase delay when
1077 passing a TCP/IP connection to a new process, in case the original
1078 process has to generate a bounce, and remove special handling of
1079 127.0.0.2 (sic), which is no longer necessary.
1081 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1082 be the same on different OS.
1084 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1087 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1088 whether --show-vars was specified or not
1090 JJ/02 exipick: Added support for new ACL variable spool format introduced
1093 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1094 syntactically invalid From: or Reply-to: line, and a filter used this to
1095 generate an autoreply, and therefore failed to obtain an address for the
1096 autoreply, Exim could try to deliver to a non-existent relative file
1097 name, causing unrelated and misleading errors. What now happens is that
1098 it logs this as a hard delivery error, but does not attempt to create a
1101 PH/11 The exinext utility has a -C option for testing purposes, but although
1102 the given file was scanned by exinext itself; it wasn't being passed on
1103 when Exim was called.
1105 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1106 an end-of-file indication when reading a command response.
1108 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1109 compiled. In many other places in Exim, IPv6 addresses are always
1110 recognized, so I have changed this. It also means that IPv4 domain
1111 literals of the form [IPV4:n.n.n.n] are now always recognized.
1113 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1114 used if the router is not running as root, for example, when verifying at
1115 ACL time, or when using -bh. The debugging output from this situation was
1116 non-existent - all you got was a failure to exec. I have made two
1119 (a) Failures to set uid/gid, the current directory, or a process leader
1120 in a subprocess such as that created by queryprogram now generate
1121 suitable debugging ouput when -d is set.
1123 (b) The queryprogram router detects when it is not running as root,
1124 outputs suitable debugging information if -d is set, and then runs
1125 the subprocess without attempting to change uid/gid.
1127 PH/15 Minor change to Makefile for building test_host (undocumented testing
1130 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1131 additional section of a DNS packet that returns MX or SRV records.
1132 Instead, it always explicitly searches for A/AAAA records. This avoids
1133 major problems that occur when a DNS server includes only records of one
1134 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1135 fixed another bug: if SRV records were looked up and the corresponding
1136 address records were *not* found in the additional section, the port
1137 values from the SRV records were lost.
1139 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1140 using the correct key (the original address) when searching the retry
1141 rules in order to find which one to use for generating the retry hint.
1143 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1144 adding the default one. Similarly, if it contains a Reply-To: header, the
1145 errors_reply_to option, if set, is not used.
1147 PH/19 When calculating a retry time, Exim used to measure the "time since
1148 failure" by looking at the "first failed" field in the retry record. Now
1149 it does not use this if it is later than than the arrival time of the
1150 message. Instead it uses the arrival time. This makes for better
1151 behaviour in cases where some deliveries succeed, thus re-setting the
1152 "first failed" field. An example is a quota failure for a huge message
1153 when small messages continue to be delivered. Without this change, the
1154 "time since failure" will always be short, possible causing more frequent
1155 delivery attempts for the huge message than are intended.
1156 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1158 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1159 $1, $2, $3) because the numerical variables can be reset during some
1160 expansion items (e.g. "match"), thereby losing the authentication data.
1162 PH/21 Make -bV show the size of off_t variables so that the test suite can
1163 decide whether to run tests for quotas > 2G.
1165 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1166 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1167 filecount value is greater than 2G or if a quota value is greater than 2G
1168 on a system where the size of off_t is not greater than 4, a panic error
1171 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1172 never match. The debug and -bh output now contains an explicit error
1173 message indicating a malformed IPv4 address or mask.
1175 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1176 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1177 PH/23 above applies.
1179 PH/25 Do not write to syslog when running in the test harness. The only
1180 occasion when this arises is a failure to open the main or panic logs
1181 (for which there is an explicit test).
1183 PH/26 Added the /no_tell option to "control=freeze".
1185 PH/27 If a host name lookup failed very early in a connection, for example, if
1186 the IP address matched host_lookup and the reverse lookup yielded a name
1187 that did not have a forward lookup, an error message of the form "no IP
1188 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1189 could be logged. Now it outputs the IP address instead of "NULL".
1191 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1192 allows the sender and the authenticated sender to be set when
1193 submitting a message from within Exim. Since child_open_exim() is
1194 documented for local_scan(), the new function should be too.
1196 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1197 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1198 results in an empty string is now treated as unset.
1200 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1202 PH/31 Added sender_verify_fail logging option.
1204 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1205 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1206 tidied the source and removed it altogether.
1208 PH/33 When a queue run was abandoned because the load average was too high, a
1209 log line was always written; now it is written only if the queue_run log
1210 selector is set. In addition, the log line for abandonment now contains
1211 information about the queue run such as the pid. This is always present
1212 in "start" and "stop" lines but was omitted from the "abandon" line.
1214 PH/34 Omit spaces between a header name and the colon in the error message that
1215 is given when verify = headers_syntax fails (if there are lots of them,
1216 the message gets confusing).
1218 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1219 names, as there are now some PTR records that contain slashes. This check
1220 is only to protect against broken name servers that fall over on strange
1221 characters, so the fact that it applies to all lookups doesn't matter.
1223 PH/36 Now that the new test suite is complete, we can remove some of the
1224 special code in Exim that was needed for the old test suite. For example,
1225 sorting DNS records because real resolvers return them in an arbitrary
1226 order. The new test suite's fake resolver always returns records in the
1229 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1230 bounces) except when queue_only is set, to avoid logging races between
1231 the different processes.
1233 PH/38 Panic-die if .include specifies a non-absolute path.
1235 PH/39 A tweak to the "H" retry rule from its user.
1237 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1238 a label. They prevented compilation on older perls.
1240 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1241 a warning to be raised on newish perls.
1243 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1244 on queue. Changes to match documented behaviour of showing count of
1245 messages matching specified criteria.
1247 PH/40 Changed the default ident timeout from 30s to 5s.
1249 PH/41 Added support for the use of login_cap features, on those BSD systems
1250 that have them, for controlling the resources used by pipe deliveries.
1252 PH/42 The content-scanning code uses fopen() to create files in which to put
1253 message data. Previously it was not paying any attention to the mode of
1254 the files. Exim runs with umask(0) because the rest of the code creates
1255 files with open(), and sets the required mode explicitly. Thus, these
1256 files were ending up world-writeable. This was not a big issue, because,
1257 being within the spool directory, they were not world-accessible. I have
1258 created a function called modefopen, which takes an additional mode
1259 argument. It sets umask(777), creates the file, chmods it to the required
1260 mode, then resets the umask. All the relevant calls to fopen() in the
1261 content scanning code have been changed to use this function.
1263 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1264 to 24 hours. This avoids potential overflow problems when processing G
1265 and H retry rules. I suspect nobody ever tinkers with this value.
1267 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1269 PH/45 When the plaintext authenticator is running as a client, the server's
1270 challenges are checked to ensure they are valid base64 strings. By
1271 default, the authentication attempt is cancelled if an invalid string is
1272 received. Setting client_ignore_invalid_base64 true ignores these errors.
1273 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1274 they are received. Thus, the responses can be made to depend on the
1275 challenges. If an invalid string is ignored, an empty string is placed in
1278 PH/46 Messages that are created by the autoreply transport now contains a
1279 References: header, in accordance with RFCs 2822 and 3834.
1281 PH/47 Added authenticated_sender_force to the smtp transport.
1283 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1285 PH/49 Installed latest patch from the Sieve maintainer.
1287 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1288 was also set, the appendfile transport was unnecessarily scanning a
1289 directory of message files (e.g. for maildir delivery) to find the count
1290 of files (along with the size), even though it did not need this
1291 information. It now does the scan only if it needs to find either the
1292 size of the count of files.
1294 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1296 PH/52 Two bugs concerned with error handling when the smtp transport is
1299 (i) Exim was not creating retry information for temporary errors given
1300 for individual recipients after the DATA command when the smtp transport
1301 was used in LMTP mode. This meant that they could be retried too
1302 frequently, and not timed out correctly.
1304 (ii) Exim was setting the flag that allows error details to be returned
1305 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1306 recipients that were returned after the DATA command.
1308 PH/53 This is related to PH/52, but is more general: for any failing address,
1309 when detailed error information was permitted to be returned to the
1310 sender, but the error was temporary, then after the final timeout, only
1311 "retry timeout exceeded" was returned. Now it returns the full error as
1312 well as "retry timeout exceeded".
1314 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1315 do this, and (what is worse) MTAs that accept it.
1317 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1318 will now be deprecated.
1320 PH/56 New os.c-cygwin from the Cygwin maintainer.
1322 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1323 formats (previously only available in exim formats via -bpr, -bpru,
1324 and -bpra. Now also available in native and exiqgrep formats)
1326 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1327 with very large, slow to parse queues
1329 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1331 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1333 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1334 responses to authentication challenges, though it was showing the
1335 challenges; (ii) I've removed the CR characters from the debug output for
1338 PH/58 Allow for the insertion of a newline as well as a space when a string
1339 is turned into more than one encoded-word during RFC 2047 encoding. The
1340 Sieve code now uses this.
1342 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1343 data_4xx, lost_connection, tls_required.
1345 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1346 message was being sent as an SMTP response.
1348 PH/61 Add -l and -k options to exicyclog.
1350 PH/62 When verifying, if an address was redirected to one new address, so that
1351 verification continued, and the new address failed or deferred after
1352 having set something in $address_data, the value of $address_data was not
1353 passed back to the ACL. This was different to the case when no
1354 redirection occurred. The value is now passed back in both cases.
1356 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1357 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1358 use login_cap.h, so on its own it isn't the distinguishing feature. The
1359 new name refers directly to the setclassresources() function.
1361 PH/65 Added configuration files for NetBSD3.
1363 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1365 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1366 is preferred over IPv4.
1368 PH/68 The bounce_return_message and bounce_return_body options were not being
1369 honoured for bounces generated during the reception of non-SMTP messages.
1370 In particular, this applied to messages rejected by the ACL. This bug has
1371 been fixed. However, if bounce_return_message is true and bounce_return_
1372 body is false, the headers that are returned for a non-SMTP message
1373 include only those that have been read before the error was detected.
1374 (In the case of an ACL rejection, they have all been read.)
1376 PH/69 The HTML version of the specification is now built in a directory called
1377 spec_html instead of spec.html, because the latter looks like a path with
1378 a MIME-type, and this confuses some software.
1380 PH/70 Catch two compiler warnings in sieve.c.
1382 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1383 function verify_get_ident() calls ip_connect() to connect a socket, but
1384 if the "connect()" function timed out, ip_connect() used to close the
1385 socket. However, verify_get_ident() also closes the socket later, and in
1386 between Exim writes to the log, which may get opened at this point. When
1387 the socket was closed in ip_connect(), the log could get the same file
1388 descriptor number as the socket. This naturally causes chaos. The fix is
1389 not to close the socket in ip_connect(); the socket should be closed by
1390 the function that creates it. There was only one place in the code where
1391 this was missing, in the iplookup router, which I don't think anybody now
1392 uses, but I've fixed it anyway.
1394 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1395 well as to direct DNS lookups. Otherwise the handling of names in host
1396 lists is inconsistent and therefore confusing.
1402 PH/01 Two changes to the default runtime configuration:
1404 (1) Move the checks for relay_from_hosts and authenticated clients from
1405 after to before the (commented out) DNS black list checks.
1407 (2) Add control=submission to the relay_from_hosts and authenticated
1408 clients checks, on the grounds that messages accepted by these
1409 statements are most likely to be submissions.
1411 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1413 (1) Generate an error if the third argument for the ${prvs expansion is
1416 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1419 (3) Reset the variables that are obtained from the first argument of
1420 ${prvscheck and used in the second argument before leaving the code,
1421 because their memory is reclaimed, so using them afterwards may do
1424 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1425 one (it's much easier than Tom thought :-).
1427 (5) Because of (4), we can now allow for the use of $prvscheck_result
1428 inside the third argument.
1430 PH/03 For some reason, the default setting of PATH when running a command from
1431 a pipe transport was just "/usr/bin". I have changed it to
1434 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1435 anything to be listed in the output from -bV.
1437 PH/05 When a filter generated an autoreply, the entire To: header line was
1438 quoted in the delivery log line, like this:
1440 => >A.N.Other <ano@some.domain> <original@ddress> ...
1442 This has been changed so that it extracts the operative address. There
1443 may be more than one such address. If so, they are comma-separated, like
1446 => >ano@some.domain,ona@other.domain <original@ddress> ...
1448 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1449 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1450 not being looked up in the rDNS to get a host name, Exim was showing the
1451 IP address twice in Received: lines, even though the IP addresses were
1452 identical. For example:
1454 Received: from [1.2.3.4] (helo=[1.2.3.4])
1456 However, if the real host name was known, it was omitting the HELO data
1457 if it matched the actual IP address. This has been tidied up so that it
1458 doesn't show the same IP address twice.
1460 PH/07 When both +timestamp and +memory debugging was on, the value given by
1461 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1462 called by the debug printing, thereby overwriting the timestamp buffer.
1463 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1465 PH/08 When the original message was included in an autoreply transport, it
1466 always said "this is a copy of the message, including all the headers",
1467 even if body_only or headers_only was set. It now gives an appropriate
1470 PH/09 Applied a patch from the Sieve maintainer which:
1472 o fixes some comments
1473 o adds the (disabled) notify extension core
1474 o adds some debug output for the result of if/elsif tests
1475 o points to the current vacation draft in the documentation
1476 and documents the missing references header update
1480 o fixes a bug in processing the envelope test (when testing
1481 multiple envelope elements, the last element determinted the
1484 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1485 Electronic Mail") by including:
1487 Auto-submitted: auto-generated
1489 in the messages that it generates (bounce messages and others, such as
1490 warnings). In the case of bounce messages for non-SMTP mesages, there was
1491 also a typo: it was using "Auto_submitted" (underscore instead of
1492 hyphen). Since every message generated by Exim is necessarily in response
1493 to another message, thes have all been changed to:
1495 Auto-Submitted: auto-replied
1497 in accordance with these statements in the RFC:
1499 The auto-replied keyword:
1501 - SHOULD be used on messages sent in direct response to another
1502 message by an automatic process,
1504 - MUST NOT be used on manually-generated messages,
1506 - MAY be used on Delivery Status Notifications (DSNs) and Message
1507 Disposition Notifications (MDNs),
1509 - MUST NOT be used on messages generated by automatic or periodic
1510 processes, except for messages which are automatic responses to
1513 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1514 to the default Received: header definition.
1516 PH/12 Added log selector acl_warn_skipped (default on).
1518 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1519 variables because (a) they are in the wrong storage pool and (b) even if
1520 they were copied, it wouldn't work properly because of the caching.
1522 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1523 checking when decoding. Apparently there are clients that generate
1524 overlong encoded strings. Why am I not surprised?
1526 PH/15 If the first argument of "${if match_address" was not empty, but did not
1527 contain an "@" character, Exim crashed. Now it writes a panic log message
1528 and treats the condition as false.
1530 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1532 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1533 extension "envelope-auth". The code is finished and in agreement with
1534 other implementations, but there is no documentation so far and in fact,
1535 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1536 not changing the active code.
1538 Print executed "if" and "elsif" statements when debugging is used. This
1539 helps a great deal to understand what a filter does.
1541 Document more things not specified clearly in RFC3028. I had all this
1542 sorted out, when out of a sudden new issues came to my mind. Oops."
1544 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1547 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1548 canonical form (as documented). However, after a host name lookup from
1549 the IP address, check_host() was doing a simple string comparison with
1550 addresses acquired from the DNS when checking that the found name did
1551 have the original IP as one of its addresses. Since any found IPv6
1552 addresses are likely to be in abbreviated form, the comparison could
1553 fail. Luckily, there already exists a function for doing the comparison
1554 by converting both addresses to binary, so now that is used instead of
1555 the text comparison.
1557 PH/20 There was another similar case to PH/19, when a complete host name was
1558 given in a host list; looking up its IP address could give an abbreviated
1559 form, whereas the current host's name might or might not be abbreviated.
1560 The same fix has been applied.
1566 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1567 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1570 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1571 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1573 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1575 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1576 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1577 Also, for "Auto-Submitted", treat anything other than "no" as
1578 non-personal, in accordance with RFC 3834. (Previously it treated
1579 anything starting "auto-" as non-personal.)
1581 TF/01 The control=submission/name=... option had a problem with syntax
1582 errors if the name included a slash character. The /name= option
1583 now slurps the rest of the string, so it can include any characters
1584 but it must come last in the list of options (after /sender_retain
1587 PH/05 Some modifications to the interface to the fake nameserver for the new
1595 TK/01 Added the "success_on_redirect" address verification option. See
1596 NewStuff for rationale and an example.
1598 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1600 PH/02 Patch to exigrep to allow it to work on syslog lines.
1602 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1603 fread() to skip over the body file's header line, because in Cygwin the
1604 header line is locked and is inaccessible.
1606 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1607 co-exist for some time) to make it clear that it is the Exim ID that is
1608 referenced, not the Message-ID: header line.
1610 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1611 string_format() function, because snprintf() does not exist on all
1614 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1615 ${stat: expansion item.
1617 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1618 protocol synchronization error", to keep the pedants happy.
1620 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1621 well as for IRIX systems, when gcc is being used. See the host.c source
1624 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1626 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1629 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1630 transport and to the smtp transport in LMTP mode.
1632 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1634 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1635 run a filter in a subprocess. This could lead to confusion in subsequent
1636 lookups in the parent process. There should also be a search_tidyup() at
1637 the end of the subprocess.
1639 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1640 only if the host matched helo_try_verify_hosts, which caused the
1641 verification to occur when the EHLO/HELO command was issued. The ACL just
1642 tested the remembered result. Now, if a previous verification attempt has
1643 not happened, "verify = helo" does it there and then.
1645 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1647 TK/03 Fix log output including CR from clamd.
1649 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1650 exist provoked a memory error which could cause a segfault.
1652 PH/15 Installed PCRE 6.2
1654 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1656 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1657 of the problem. Specifically, suggested +O2 rather than +O1 for the
1660 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1662 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1663 record, the port was not correctly logged when the outgoing_port log
1664 selector was set (it logged the transort's default port).
1666 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1667 fallback_hosts, and "hosts" in the smtp transport.
1669 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1670 host errors such as "Connection refused".
1672 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1673 authentication with radiusclient 0.4.9:
1675 - Error returned from rc_read_config was caught wrongly
1676 - Username/password not passed on to radius server due to wrong length.
1678 The presumption is that some radiusclient API changes for 4.51/PH/17
1679 were not taken care of correctly. The code is still untested by me (my
1680 Linux distribution still has 0.3.2 of radiusclient), but it was
1681 contributed by a Radius user.
1683 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1684 expanding the "port" option of the smtp transport.
1686 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1687 while reading a MIME header. Thanks to Tom Hughes for a patch.
1689 PH/24 Include config.h inside local_scan.h so that configuration settings are
1692 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1693 that in an ACL for RCPT (for example), you can examine exactly what was
1696 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1697 commands, but it was not correctly comparing the address with the actual
1698 client host address. Thus, it would show the EHLO address in Received:
1699 header lines when this was not necessary.
1701 PH/27 Added the % operator to ${eval:}.
1703 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1704 it should be ignoring failures (because with -C, for example, it has lost
1705 privilege). It wasn't ignoring creation failures other than "already
1708 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1711 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1712 because an input line was too long, either on its own, or by virtue of
1713 too many continuations, the temporary file was not being removed, and the
1714 return code was incorrect.
1716 PH/31 Missing "BOOL" in function definition in filtertest.c.
1718 PH/32 Applied Sieve patches from the maintainer.
1720 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1722 PH/33 Added "verify = not_blind".
1724 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1725 Local/Makefile (with some defaults set). These are used in built scripts
1726 such as exicyclog, but they have never been used in the exim_install
1727 script (though there are many overriding facilities there). I have
1728 arranged that the exim_install script now takes note of these two
1731 PH/35 Installed configuration files for Dragonfly.
1733 PH/36 When a locally submitted message by a trusted user did not contain a
1734 From: header, and the sender address was obtained from -f or from an SMTP
1735 MAIL command, and the trusted user did not use -F to supply a sender
1736 name, $originator_name was incorrectly used when constructing a From:
1737 header. Furthermore, $originator_name was used for submission mode
1738 messages from external hosts without From: headers in a similar way,
1739 which is clearly wrong.
1741 PH/37 Added control=suppress_local_fixups.
1743 PH/38 When log_selector = +received_sender was set, and the addition of the
1744 sender made the log line's construction buffer exactly full, or one byte
1745 less than full, an overflow happened when the terminating "\n" was
1748 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1749 when the result of a list match is failure because a DNS lookup failed.
1751 PH/40 RM_COMMAND is now used in the building process.
1753 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1754 the "build-* directories that it finds.
1756 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1757 address in a domain literal was a prefix of an interface address.
1759 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1760 when verifying a sender address, unless rewrite_headers is false.
1762 PH/44 Wrote a long comment about why errors_to addresses are verified as
1763 recipients, not senders.
1765 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1766 the ratelimit ACL was added.
1768 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1770 PH/46 Added extra information about PostgreSQL errors to the error string.
1772 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1773 avoiding the need to install special zones in a real server. This is
1774 backwards compatible; if it can't find the fake resolver, it drops back.
1775 Thus, both old and new test suites can be run.
1777 TF/02 Added util/ratelimit.pl
1779 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1780 clock is set back in time.
1782 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1783 Candler <B.Candler@pobox.com>.
1785 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1786 OK for addresses that are the result of redirections.
1788 PH/48 A number of further additions for the benefit of the new test suite,
1789 including a fake gethostbyname() that interfaces to the fake DNS resolver
1792 TF/06 The fix for widen_domains has also been applied to qualify_single and
1793 search_parents which are the other dnslookup options that can cause
1796 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1799 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1801 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1802 service" if it encountered an MX record with an empty target hostname.
1803 The message is now "an MX or SRV record indicated no SMTP service".
1805 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1806 if the DNS of the sending site is misconfigured. This is quite a
1807 common situation. This change restores the behaviour of treating a
1808 helo verification defer as a failure.
1810 PH/51 If self=fail was set on a router, the bounce message did not include the
1811 actual error message.
1817 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1819 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1820 command itself ended in error, the underlying message about the transport
1821 filter timeout was being overwritten with the pipe command error. Now the
1822 underlying error message should be appended to the second error message.
1824 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1826 PH/02 Reduce the amount of output that "make" produces by default. Full output
1827 can still be requested.
1829 PH/03 The warning log line about a condition test deferring for a "warn" verb
1830 was being output only once per connection, rather than after each
1831 occurrence (because it was using the same function as for successful
1832 "warn" verbs). This seems wrong, so I have changed it.
1834 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1835 it should not have, which might have caused a crash in the right
1836 circumstances, but probably never did.
1838 PH/04 Installed a modified version of Tony Finch's patch to make submission
1839 mode fix the return path as well as the Sender: header line, and to
1840 add a /name= option so that you can make the user's friendly name appear
1843 TF/03 Added the control = fakedefer ACL modifier.
1845 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1846 Mark Lowes for thorough testing.
1848 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1850 TK/03 Merged latest SRS patch from Miles Wilton.
1852 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1853 with the definition in sysexits.h (which is #included earlier).
1854 Fortunately, Exim does not actually use EX_OK. The code used to try to
1855 preserve the sysexits.h value, by assumimg that macro definitions were
1856 scanned for macro replacements. I have been disabused of this notion,
1857 so now the code just undefines EX_OK before #including unistd.h.
1859 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1860 in the smtp transport. When a block could not be written in a single
1861 write() function, the timeout was being re-applied to each part-write.
1862 This seems wrong - if the receiver was accepting one byte at a time it
1863 would take for ever. The timeout is now adjusted when this happens. It
1864 doesn't have to be particularly precise.
1866 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1867 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1869 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1870 without a domain if the check to <postmaster@domain> fails.
1872 SC/01 Eximstats: added -xls and the ability to specify output files
1873 (patch written by Frank Heydlauf).
1875 SC/02 Eximstats: use FileHandles for outputing results.
1877 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1879 SC/04 Eximstats: fixed display of large numbers with -nvr option
1881 SC/05 Eximstats: fixed merging of reports with empty tables.
1883 SC/06 Eximstats: added the -include_original_destination flag
1885 SC/07 Eximstats: removed tabs and trailing whitespace.
1887 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1889 TK/06 MBOX spool code: Add real "From " MBOX separator line
1890 so the .eml file is really in mbox format (even though
1891 most programs do not really care). Patch from Alex Miller.
1893 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1894 The latter is generated from $received_to and is only set if the
1895 message has one envelope recipient. SA can use these headers,
1896 obviously out-of-the-box. Patch from Alex Miller.
1898 PH/08 The ${def test on a variable was returning false if the variable's
1899 value was "0", contrary to what the specification has always said!
1900 The result should be true unless the variable is empty.
1902 PH/09 The syntax error of a character other than { following "${if
1903 def:variable_name" (after optional whitespace) was not being diagnosed.
1904 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1905 accidental colon was present, for example, could give incorrect results.
1907 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1908 result is used (not including appendfile, where other changes are about
1911 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1912 This involved changing a lot of size variables from int to off_t. It
1913 should work with maildirs and everything.
1915 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1916 spamd dying while we are connected to it.
1918 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1921 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1922 function for BDB 4.3.
1924 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1926 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1927 the sources for PCRE have been re-arranged and more files are now
1930 PH/15 The code I had for printing potentially long long variables in PH/11
1931 above was not the best (it lost precision). The length of off_t variables
1932 is now inspected at build time, and an appropriate printing format (%ld
1933 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1934 to be "long long int" or "long int". This is needed for the internal
1935 formatting function string_vformat().
1937 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1938 the configuration file to be ":syslog", then the script "guesses" where
1939 the logs files are, rather than using the compiled in default. In our
1940 case the guess is not the same as the compiled default, so the script
1941 suddenly stopped working when I started to use syslog. The patch checks
1942 to see if log_file_path is "". If so, it attempts to read it from exim
1943 with no configuration file to get the compiled in version, before it
1944 falls back to the previous guessing code."
1946 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1947 implementing BATV in an Exim configuration. See NewStuff for the gory
1950 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1951 Makefile that are specific to HP-UX.
1953 PH/18 If the "use_postmaster" option was set for a recipient callout together
1954 with the "random" option, the postmaster address was used as the MAIL
1955 FROM address for the random test, but not for the subsequent recipient
1956 test. It is now used for both.
1958 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1959 patch removes a few documentation additions to RFC 3028, because the
1960 latest draft now contains them. It adds the new en;ascii-case comparator
1961 and a new error check for 8bit text in MIME parts. Comparator and
1962 require names are now matched exactly. I enabled the subaddress
1963 extension, but it is not well tested yet (read: it works for me)."
1965 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1966 rework some of the code of TK/09 above to avoid the hardwired use of
1967 "%lld" and "long long". Replaced the call to snprintf() with a call to
1970 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1971 records point to non-existent hosts", "retry timeout exceeded", and
1972 "retry time not reached for any host after a long failure period".
1974 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1975 experimental DomainKeys support:
1977 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1978 (2) On an error such as an illegally used "control", the wrong name for
1979 the control was given.
1981 These problems did NOT occur unless DomainKeys support was compiled.
1983 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1985 PH/24 Added ${if match_ip condition.
1987 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1988 databases so that it will be absolutely obvious if a crash occurs in the
1989 DB library. This is a regular occurrence (often caused by mis-matched
1992 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1993 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1994 picked up on a user's system that detects such things. There doesn't seem
1995 to be a gcc warning option for this - only an attribute that has to be
1996 put on the function's prototype. It seems that in Fedora Core 4 they have
1997 set this on a number of new functions. No doubt there will be more in due
2000 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2001 specify a transport. However, if an address that was verified by such a
2002 router was the subject of a callout, Exim crashed because it tried to
2003 read the rcpt_include_affixes from the non-existent transport. Now it
2004 just assumes that the setting of that option is false. This bug was
2005 introduced by 4.51/PH/31.
2007 PH/28 Changed -d+all to exclude +memory, because that information is very
2008 rarely of interest, but it makes the output a lot bigger. People tend to
2009 do -d+all out of habit.
2011 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2012 code in os-type was giving problems when libc.so lives in lib64, like on
2015 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2016 aren't the modern standard, and it seems that some systems' include files
2017 don't always have them. Exim was already checking for some of the newer
2018 ones like T_AAAA, and defining it itself. I've added checks for all the
2019 record types that Exim uses.
2021 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2022 not automatically generating a new one, as it is supposed to. This
2023 prevented TLS from working. If the file did exist, but contained invalid
2024 data, a new version was generated, as expected. It was only the case of a
2025 non-existent file that was broken.
2027 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2028 with a change in libdomainkeys > 0.64.
2030 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2031 from DNS. If the selector record carries the flag, it now has
2032 precedence over the domain-wide flag.
2034 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2036 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2037 the use of a port name that isn't defined in /etc/services) occurred, the
2038 message was deferred as in a normal delivery, and thus remained on the
2039 spool, instead of being failed because of the mua_wrapper setting. This
2040 is now fixed, and I tidied up some of the mua_wrapper messages at the
2043 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2044 the messages in a hash of arrays rather than using individual hashes.
2045 This is a bit cleaner and results in dramatic memory savings, albeit
2046 at a slight CPU cost.
2048 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2049 as requested by Marc Sherman.
2051 SC/10 Eximstats: added histograms for user specified patterns as requested
2054 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2056 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2057 fopen() in the content-scanning modules that did not already have it.
2063 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2064 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2066 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2068 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2070 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2071 to test Sieve filters that use "vacation".
2073 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2074 that changes the way the GnuTLS parameters are stored in the cache file.
2075 The new format can be generated externally. For backward compatibility,
2076 if the data in the cache doesn't make sense, Exim assumes it has read an
2077 old-format file, and it generates new data and writes a new file. This
2078 means that you can't go back to an older release without removing the
2081 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2082 work if there are any delivery delays because "one_time" forces the
2083 parent to be marked "delivered", so its unseen clone is never tried
2084 again. For this reason, Exim now forbids the simultaneous setting of
2087 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2088 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2089 entry if you want to know the details. The fix, however, neglected to
2090 consider the case when local delivery batching is involved. The test for
2091 "previously delivered" was not happening when checking to see if an
2092 address could be batched with a previous (undelivered) one; under
2093 certain circumstances this could lead to multiple deliveries to the same
2096 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2097 in its include files, and this causes problems building Exim.
2099 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2100 header_syntax) but Exim was just ignoring anything given after a slash.
2101 In particular, this caused confusion with an attempt to use "verify =
2102 reverse_host_lookup/defer_ok". An error is now given when options are
2103 supplied for verify items that do not have them. (Maybe reverse_host_
2104 lookup should have a defer_ok option, but that's a different point.)
2106 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2107 defined by RFC 821) to 2048, because there were problems with some AUTH
2108 commands, and RFC 1869 says the size should be increased for extended
2109 SMTP commands that take arguments.
2111 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2114 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2115 "unknown" error; now it says that the functionality isn't in the binary.
2117 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2118 an address' error message when a string expansion fails (syntax or
2119 whatever). Otherwise the password may appear in the log. Following change
2120 PH/42 below, there is no longer a chance of it appearing in a bounce
2123 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2125 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2126 of its addresses was ever tried. (Bugzilla bug #2.)
2128 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2129 the result incorrectly in the debug output. (It correctly added a newline
2130 to what was transported.)
2132 TF/01 Added $received_time.
2134 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2135 commented out examples of how to interface to a virus scanner and to
2136 SpamAssassin. Also added commented examples of av_scanner and
2137 spamd_address settings.
2139 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2140 and controls are allowed in which ACLs. There were a couple of minor
2141 errors. Some of the entries in the conditions table (which is a table of
2142 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2143 negation of where the condition IS allowed.
2145 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2147 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2148 header file does not have a version number, so I've had to invent a new
2149 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2150 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2151 radiusclient), but it was contributed by a Radius user.
2153 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2154 files or directories, for OpenSSL.
2156 PH/19 When an Exim process that is running as root has to create an Exim log
2157 file, it does so in a subprocess that runs as exim:exim so as to get the
2158 ownership right at creation (otherwise, other Exim processes might see
2159 the file with the wrong ownership). There was no test for failure of this
2160 fork() call, which would lead to the process getting stuck as it waited
2161 for a non-existent subprocess. Forks do occasionally fail when resources
2162 run out. I reviewed all the other calls to fork(); they all seem to check
2165 PH/20 When checking for unexpected SMTP input at connect time (before writing
2166 the banner), Exim was not dealing correctly with a non-positive return
2167 from the read() function. If the client had disconnected by this time,
2168 the result was a log entry for a synchronization error with an empty
2169 string after "input=" when read() returned zero. If read() returned -1
2170 (an event I could not check), uninitialized data bytes were printed.
2171 There were reports of junk text (parts of files, etc) appearing after
2174 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2176 PH/22 Added support for macro redefinition, and (re)definition in between
2177 driver and ACL definitions.
2179 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2180 forgetting to use the resulting value; it was using the unexpanded value.
2182 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2183 hadn't been configured. The fix is from Juergen Kreileder, who
2184 understands it better than I do:
2186 "Here's what I see happening with three configured cyrus_sasl
2187 authenticators configured (plain, login, cram-md5):
2189 On startup auth_cyrus_sasl_init() gets called for each of these.
2190 This means three calls to sasl_listmech() without a specified mech_list.
2191 => SASL tests which mechs of all available mechs actually work
2192 => three warnings about OTP not working
2193 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2195 With the patch, sasl_listmech() also gets called three times. But now
2196 SASL's mech_list option is set to the server_mech specified in the the
2197 authenticator. Or in other words, the answer from sasl_listmech()
2198 gets limited to just the mech you're testing for (which is different
2200 => the return list contains just 'plain' or 'login', 'cram-md5' or
2201 nothing depending on the value of ob->server_mech.
2203 I've just tested the patch: Authentication still works fine,
2204 unavailable mechs specified in the exim configuration are still
2205 caught, and the auth.log warnings about OTP are gone."
2207 PH/25 When debugging is enabled, the contents of the command line are added
2208 to the debugging output, even when log_selector=+arguments is not
2211 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2212 answer is "GNU", and only if the return is "GNU/something" is the answer
2215 PH/27 $acl_verify_message is now set immediately after the failure of a
2216 verification in an ACL, and so is available in subsequent modifiers. In
2217 particular, the message can be preserved by coding like this:
2219 warn !verify = sender
2220 set acl_m0 = $acl_verify_message
2222 Previously, $acl_verify_message was set only while expanding "message"
2223 and "log_message" when a very denied access.
2225 PH/28 Modified OS/os.c-Linux with
2227 -#ifndef OS_LOAD_AVERAGE
2228 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2230 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2231 nomenclature these days.)
2233 PH/29 Installed patch from the Sieve maintainer that adds the options
2234 sieve_useraddress and sieve_subaddress to the redirect router.
2236 PH/30 In these circumstances:
2237 . Two addresses routed to the same list of hosts;
2238 . First host does not offer TLS;
2239 . First host accepts first address;
2240 . First host gives temporary error to second address;
2241 . Second host offers TLS and a TLS session is established;
2242 . Second host accepts second address.
2243 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2244 and peerdn, if requested) that were in fact used only for the second
2247 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2248 attention to any local part prefix or suffix that was matched by the
2249 router that accepted the address. It now behaves in the same way as it
2250 does for delivery: the affixes are removed from the local part unless
2251 rcpt_include_affixes is set on the transport.
2253 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2254 timeout during the DATA phase of an incoming message.
2256 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2257 applied a patch sanctioned by the Sieve maintainer.
2259 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2260 the uid or gid is negative. A case of a negative gid caused this to be
2261 noticed. The fix allows for either to be negative.
2263 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2264 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2267 PH/36 The change PH/12 above was broken. Fixed it.
2269 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2270 the grounds that routing the same address twice would always produce the
2271 same answer. This might have been true once, but it is certainly no
2272 longer true now. Routing a child address may depend on the previous
2273 routing that produced that child. Some complicated redirection strategies
2274 went wrong when messages had multiple recipients, and made Exim's
2275 behaviour dependent on the order in which the addresses were given.
2277 I have moved the duplicate checking until after the routing is complete.
2278 Exim scans the addresses that are assigned to local and remote
2279 transports, and removes any duplicates. This means that more work will be
2280 done, as duplicates will always all be routed, but duplicates are
2281 presumably rare, so I don't expect this is of any significance.
2283 For deliveries to pipes, files, and autoreplies, the duplicate checking
2284 still happens during the routing process, since they are not going to be
2287 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2288 It corrects a timeout issue with spamd. This is Ian's comment: "The
2289 background is that sometimes spamd either never reads data from a
2290 connection it has accepted, or it never writes response data. The exiscan
2291 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2292 blindly assumes that writes won't block so it may never time out."
2294 PH/39 Allow G after quota size as well as K and M.
2296 PH/40 The value set for $authenticated_id in an authenticator may not contain
2297 binary zeroes or newlines because the value is written to log lines and
2298 to spool files. There was no check on this. Now the value is run through
2299 the string_printing() function so that such characters are converted to
2300 printable escape sequences.
2302 PH/41 $message_linecount is a new variable that contains the total number of
2303 lines in the message. Compare $body_linecount, which is the count for the
2306 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2307 bounce and delay warning messages, except in certain special cases, which
2310 (a) An SMTP error message from a remote host;
2311 (b) A message specified in a :fail: redirection;
2312 (c) A message specified in a "fail" command in a system filter;
2313 (d) A message specified in a FAIL return from the queryprogram router;
2314 (e) A message specified by the cannot_route_message router option.
2316 In these cases only, Exim does include the error details in bounce and
2317 warning messages. There are also a few cases where bland messages such
2318 as "unrouteable address" or "local delivery error" are given.
2320 PH/43 $value is now also set for the "else" part of a ${run expansion.
2322 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2323 being worked on, but at least Exim now implements the latest version to
2326 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2327 process to complete was treated as a delivery failure, a timeout while
2328 writing the message to the pipe was logged, but erroneously treated as a
2329 successful delivery. Such timeouts include transport filter timeouts. For
2330 consistency with the overall process timeout, these timeouts are now
2331 treated as errors, giving rise to delivery failures by default. However,
2332 there is now a new Boolean option for the pipe transport called
2333 timeout_defer, which, if set TRUE, converts the failures into defers for
2334 both kinds of timeout. A transport filter timeout is now identified in
2337 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2338 systems where "make" and "gmake" are different, calling "gmake" at top
2339 level broke things. I've arranged for the value of $(MAKE) to be passed
2340 from the Makefile to this script so that it can call the same version of
2344 A note about Exim versions 4.44 and 4.50
2345 ----------------------------------------
2347 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2348 changes of various kinds. As a consequence, a big documentation update was
2349 needed. This delayed the release for rather longer than seemed good, especially
2350 in the light of a couple of (minor) security issues. Therefore, the changes
2351 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2352 release. So 4.44 and 4.50 are in effect two different branches that both start
2355 I have left the 4.50 change log unchanged; it contains all the changes since
2356 4.43. The change log for 4.44 is below; many of its items are identical to
2357 those for 4.50. This seems to be the most sensible way to preserve the
2358 historical information.
2364 1. Minor wording change to the doc/README.SIEVE file.
2366 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2367 computation of the current number of files was incorrect.
2369 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2370 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2373 4. Give more explanation in the error message when the command for a transport
2374 filter fails to execute.
2376 5. There are several places where Exim runs a non-Exim command in a
2377 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2378 was being done only for the command run by the queryprogram router. It is
2379 now done for all such subprocesses. The other cases are: ${run, transport
2380 filters, and the commands run by the lmtp and pipe transports.
2382 6. Added CONFIGURE_GROUP build-time option.
2384 7. Some older OS have a limit of 256 on the maximum number of file
2385 descriptors. Exim was using setrlimit() to set 1000 as a large value
2386 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2387 systems. I've change it so that if it can't get 1000, it tries for 256.
2389 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2390 was an oversight, and furthermore, ever since the addition of extra
2391 controls (e.g. 4.43/32), the checks on when to allow different forms of
2392 "control" were broken. There should now be diagnostics for all cases when a
2393 control that does not make sense is encountered.
2395 9. Added the /retain_sender option to "control=submission".
2397 10. $recipients is now available in the predata ACL (oversight).
2399 11. Tidy the search cache before the fork to do a delivery from a message
2400 received from the command line. Otherwise the child will trigger a lookup
2401 failure and thereby defer the delivery if it tries to use (for example) a
2402 cached ldap connection that the parent has called unbind on.
2404 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2405 of $address_data from the recipient verification was clobbered by the
2406 sender verification.
2408 13. The value of address_data from a sender verification is now available in
2409 $sender_address_data in subsequent conditions in the ACL statement.
2411 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2413 15. Added a new option "connect=<time>" to callout options, to set a different
2416 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2417 was its contents. (It was OK if the option was not defined at all.)
2419 17. A "Completed" log line is now written for messages that are removed from
2420 the spool by the -Mrm option.
2422 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2423 information about exactly what failed.
2425 19. Added -dd to debug only the daemon process.
2427 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2428 handles timeouts, both on the server side and network timeouts. Renamed the
2429 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2431 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2432 It is now set to "smtps".
2434 22. $host_address is now set to the target address during the checking of
2435 ignore_target_hosts.
2437 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2438 being passed; this would have caused $sender_host_name to have been used if
2439 matching the list had actually called for a host name (not very likely,
2440 since this list is usually IP addresses). A host name is now passed as
2443 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2444 code that specifies a non-restarting handler (typically sigaction() in
2445 modern systems) in an attempt to fix a rare and obscure crash bug.
2447 25. Narrowed the window for a race in the daemon that could cause it to ignore
2448 SIGCHLD signals. This is not a major problem, because they are used only to
2449 wake it up if nothing else does.
2451 26. A malformed maildirsize file could cause Exim to calculate negative values
2452 for the mailbox size or file count. Odd effects could occur as a result.
2453 The maildirsize information is now recalculated if the size or filecount
2456 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2457 support for a long time. Removed HAVE_SYS_VFS_H.
2459 28. Installed the latest version of exipick from John Jetmore.
2461 29. In an address list, if the pattern was not a regular expression, an empty
2462 subject address (from a bounce message) matched only if the pattern was an
2463 empty string. Non-empty patterns were not even tested. This was the wrong
2464 because it is perfectly reasonable to use an empty address as part of a
2465 database query. An empty address is now tested by patterns that are
2466 lookups. However, all the other forms of pattern expect the subject to
2467 contain a local part and a domain, and therefore, for them, an empty
2468 address still always fails if the pattern is not itself empty.
2470 30. Exim went into a mad DNS loop when attempting to do a callout where the
2471 host was specified on an smtp transport, and looking it up yielded more
2472 than one IP address.
2474 31. Re-factored the code for checking spool and log partition space into a
2475 function that finds that data and another that does the check. The former
2476 is then used to implement four new variables: $spool_space, $log_space,
2477 $spool_inodes, and $log_inodes.
2479 32. The RFC2047 encoding function was originally intended for short strings
2480 such as real names; it was not keeping to the 75-character limit for
2481 encoded words that the RFC imposes. It now respects the limit, and
2482 generates multiple encoded words if necessary. To be on the safe side, I
2483 have increased the buffer size for the ${rfc2047: expansion operator from
2486 33. It is now permitted to omit both strings after an "if" condition; if the
2487 condition is true, the result is "true". As before, when the second string
2488 is omitted, a false condition yields an empty string. This makes it less
2489 cumbersome to write custom ACL and router conditions.
2491 34. Failure to deliver a bounce message always caused it to be frozen, even if
2492 there was an errors_to setting on the router. The errors_to setting is now
2495 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2496 canonical form (fully expanded) before being placed in
2497 $sender_host_address.
2499 36. The table in the code that translates DNS record types into text (T_A to
2500 "A" for instance) was missing entries for NS and CNAME. It is just possible
2501 that this could have caused confusion if both these types were looked up
2502 for the same domain, because the text type is used as part of Exim's
2503 per-process caching. But the chance of anyone hitting this buglet seems
2506 37. The dnsdb lookup has been extended in a number of ways.
2508 (1) There is a new type, "zns", which walks up the domain tree until it
2509 finds some nameserver records. It should be used with care.
2511 (2) There is a new type, "mxh", which is like "mx" except that it returns
2512 just the host names, not the priorities.
2514 (3) It is now possible to give a list of domains (or IP addresses) to be
2515 looked up. The behaviour when one of the lookups defers can be
2516 controlled by a keyword.
2518 (4) It is now possible to specify the separator character for use when
2519 multiple records are returned.
2521 38. The dnslists ACL condition has been extended: it is now possible to supply
2522 a list of IP addresses and/or domains to be looked up in a particular DNS
2525 39. Added log_selector=+queue_time_overall.
2527 40. When running the queue in the test harness, wait just a tad after forking a
2528 delivery process, to get repeatability of debugging output.
2530 41. Include certificate and key file names in error message when GnuTLS fails
2531 to set them up, because the GnuTLS error message doesn't include the name
2532 of the failing file when there is a problem reading it.
2534 42. Allow both -bf and -bF in the same test run.
2536 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2538 44. The "Exiscan patch" is now merged into the mainline Exim source.
2540 45. Sometimes the final signoff response after QUIT could fail to get
2541 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2542 < 0 before doing a fflush(). This bug looks as though it goes back to the
2543 introduction of TLS in release 3.20, but "sometimes" must have been rare
2544 because the tests only now provoked it.
2546 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2547 (this can affect the format of dates).
2549 47. exim_tidydb, when checking for the continued existence of a message for
2550 which it has found a message-specific retry record, was not finding
2551 messages that were in split spool directories. Consequently, it was
2552 deleting retry records that should have stayed in existence.
2554 48. Steve fixed some bugs in eximstats.
2556 49. The SPA authentication driver was not abandoning authentication and moving
2557 on to the next authenticator when an expansion was forced to fail,
2558 contradicting the general specification for all authenticators. Instead it
2559 was generating a temporary error. It now behaves as specified.
2561 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2562 (the order specifies the preference for clients). The order is now AES256,
2563 AES128, 3DES, ARCFOUR128.
2565 51. Small patch to Sieve code - explicitly set From: when generating an
2568 52. Exim crashed if a remote delivery caused a very long error message to be
2569 recorded - for instance if somebody sent an entire SpamAssassin report back
2570 as a large number of 550 error lines. This bug was coincidentally fixed by
2571 increasing the size of one of Exim's internal buffers (big_buffer) that
2572 happened as part of the Exiscan merge. However, to be on the safe side, I
2573 have made the code more robust (and fixed the comments that describe what
2576 53. Now that there can be additional text after "Completed" in log lines (if
2577 the queue_time_overall log selector is set), a one-byte patch to exigrep
2578 was needed to allow it to recognize "Completed" as not the last thing in
2581 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2582 patch that reportedly fixes this has been added. I am not expert enough to
2583 create a test for it. This is what the patch creator wrote:
2585 "I found a little strange behaviour of ldap code when working with
2586 Windows 2003 AD Domain, where users was placed in more than one
2587 Organization Units. When I tried to give exim partial DN, the exit code
2588 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2589 But simultaneously result of request was absolutely normal ldap result,
2590 so I produce this patch..."
2592 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2593 REFERENCE, so I have modified the code to exclude the patch when that macro
2596 55. Some experimental protocols are using DNS PTR records for new purposes. The
2597 keys for these records are domain names, not reversed IP addresses. The
2598 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2599 leaves it alone. Component reversal etc. now happens only for IP addresses.
2602 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2604 57. Double the size of the debug message buffer (to 2048) so that more of very
2605 long debug lines gets shown.
2607 58. The exicyclog utility now does better if the number of log files to keep
2608 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2610 59. Two changes related to the smtp_active_hostname option:
2612 (1) $smtp_active_hostname is now available as a variable.
2613 (2) The default for smtp_banner uses $smtp_active_hostname instead
2614 of $primary_hostname.
2616 60. The host_aton() function is supposed to be passed a string that is known
2617 to be a valid IP address. However, in the case of IPv6 addresses, it was
2618 not checking this. This is a hostage to fortune. Exim now panics and dies
2619 if the condition is not met. A case was found where this could be provoked
2620 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2621 components; fortuitously, this particular loophole had already been fixed
2622 by change 4.50/55 above.
2624 If there are any other similar loopholes, the new check in host_aton()
2625 itself should stop them being exploited. The report I received stated that
2626 data on the command line could provoke the exploit when Exim was running as
2627 exim, but did not say which command line option was involved. All I could
2628 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2629 running as the user.
2632 61. There was a buffer overflow vulnerability in the SPA authentication code
2633 (which came originally from the Samba project). I have added a test to the
2634 spa_base64_to_bits() function which I hope fixes it.
2637 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2638 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2640 63. The daemon start-up calls getloadavg() while still root for those OS that
2641 need the first call to be done as root, but it missed one case: when
2642 deliver_queue_load_max is set with deliver_drop_privilege. This is
2643 necessary for the benefit of the queue runner, because there is no re-exec
2644 when deliver_drop_privilege is set.
2646 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2647 This has been fixed.
2649 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2650 was in use, was not putting the data itself into the right store pool;
2651 consequently, it could be overwritten for a subsequent message in the same
2652 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2655 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2657 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2658 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2659 the function were treating the return as a boolean value, which happened to
2660 work because 0=false and not-0=true, but is not correct code.
2662 68. The host_aton() function was not handling scoped IPv6 addresses (those
2663 with, for example, "%eth0" on the end) correctly.
2665 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2666 negated items (that is, ~something) in unsigned ints. Some compilers
2667 apparently mutter when there is no cast.
2669 70. If an address verification called from an ACL failed, and did not produce a
2670 user-specific message (i.e. there was only a "system" message), nothing was
2671 put in $acl_verify_message. In this situation, it now puts the system
2674 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2675 session; change 4.31/43 added the unwanted input to the log line - except
2676 that it did not do this in the start of session case. It now does.
2678 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2679 This is wrong and can cause the other end to generate a synchronization
2680 error if it is another Exim or anything else that does the synchronization
2681 check. A QUIT command is no longer sent after a timeout.
2683 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2684 during host lookups.
2686 74. The defer_ok option of callout verification was not working if it was used
2687 when verifying addresses in header lines, that is, for this case:
2689 verify = header_sender/callout=defer_ok
2691 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2692 those file descriptors could be used for SMTP connections. If anything
2693 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2694 could be sent to the SMTP client, causing chaos. The daemon now opens
2695 stdin, stdout, and stderr to /dev/null when it puts itself into the
2698 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2699 log by default. The user can override this with suitable Perl magic.
2701 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2702 the log message when discard triggers, was not working for the DATA ACL or
2703 for the non-SMTP ACL.
2705 78. Error message wording change in sieve.c.
2707 79. If smtp_accept_max_per_host was set, the number of connections could be
2708 restricted to fewer than expected, because the daemon was trying to set up
2709 a new connection before checking whether the processes handling previous
2710 connections had finished. The check for completed processes is now done
2711 earlier. On busy systems, this bug wouldn't be noticed because something
2712 else would have woken the daemon, and it would have reaped the completed
2715 80. If a message was submitted locally by a user whose login name contained one
2716 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2717 It caused a spool format error. I have fixed the spool reading code. A
2718 related problem was that the "from" clause in the Received: line became
2719 illegal because of the space(s). It is now covered by ${quote_local_part.
2721 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2724 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2726 83. Patch from Sieve maintainer to fix unterminated string problem in
2727 "vacation" handling.
2729 84. Some minor changes to the Linux configuration files to help with other
2730 OS variants using glibc.
2732 85. One more patch for Sieve to update vacation handling to latest spec.
2735 ----------------------------------------------------
2736 See the note above about the 4.44 and 4.50 releases.
2737 ----------------------------------------------------
2743 1. Change 4.43/35 introduced a bug that caused file counts to be
2744 incorrectly computed when quota_filecount was set in an appendfile
2747 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2748 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2751 3. Give more explanation in the error message when the command for a transport
2752 filter fails to execute.
2754 4. There are several places where Exim runs a non-Exim command in a
2755 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2756 was being done only for the command run by the queryprogram router. It is
2757 now done for all such subprocesses. The other cases are: ${run, transport
2758 filters, and the commands run by the lmtp and pipe transports.
2760 5. Some older OS have a limit of 256 on the maximum number of file
2761 descriptors. Exim was using setrlimit() to set 1000 as a large value
2762 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2763 systems. I've change it so that if it can't get 1000, it tries for 256.
2765 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2766 was an oversight, and furthermore, ever since the addition of extra
2767 controls (e.g. 4.43/32), the checks on when to allow different forms of
2768 "control" were broken. There should now be diagnostics for all cases when a
2769 control that does not make sense is encountered.
2771 7. $recipients is now available in the predata ACL (oversight).
2773 8. Tidy the search cache before the fork to do a delivery from a message
2774 received from the command line. Otherwise the child will trigger a lookup
2775 failure and thereby defer the delivery if it tries to use (for example) a
2776 cached ldap connection that the parent has called unbind on.
2778 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2779 of $address_data from the recipient verification was clobbered by the
2780 sender verification.
2782 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2783 was its contents. (It was OK if the option was not defined at all.)
2785 11. A "Completed" log line is now written for messages that are removed from
2786 the spool by the -Mrm option.
2788 12. $host_address is now set to the target address during the checking of
2789 ignore_target_hosts.
2791 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2792 being passed; this would have caused $sender_host_name to have been used if
2793 matching the list had actually called for a host name (not very likely,
2794 since this list is usually IP addresses). A host name is now passed as
2797 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2798 code that specifies a non-restarting handler (typically sigaction() in
2799 modern systems) in an attempt to fix a rare and obscure crash bug.
2801 15. Narrowed the window for a race in the daemon that could cause it to ignore
2802 SIGCHLD signals. This is not a major problem, because they are used only to
2803 wake it up if nothing else does.
2805 16. A malformed maildirsize file could cause Exim to calculate negative values
2806 for the mailbox size or file count. Odd effects could occur as a result.
2807 The maildirsize information is now recalculated if the size or filecount
2810 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2811 support for a long time. Removed HAVE_SYS_VFS_H.
2813 18. Updated exipick to current release from John Jetmore.
2815 19. Allow an empty sender to be matched against a lookup in an address list.
2816 Previously the only cases considered were a regular expression, or an
2819 20. Exim went into a mad DNS lookup loop when doing a callout where the
2820 host was specified on the transport, if the DNS lookup yielded more than
2823 21. The RFC2047 encoding function was originally intended for short strings
2824 such as real names; it was not keeping to the 75-character limit for
2825 encoded words that the RFC imposes. It now respects the limit, and
2826 generates multiple encoded words if necessary. To be on the safe side, I
2827 have increased the buffer size for the ${rfc2047: expansion operator from
2830 22. Failure to deliver a bounce message always caused it to be frozen, even if
2831 there was an errors_to setting on the router. The errors_to setting is now
2834 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2835 canonical form (fully expanded) before being placed in
2836 $sender_host_address.
2838 24. Updated eximstats to version 1.33
2840 25. Include certificate and key file names in error message when GnuTLS fails
2841 to set them up, because the GnuTLS error message doesn't include the name
2842 of the failing file when there is a problem reading it.
2844 26. Expand error message when OpenSSL has problems setting up cert/key files.
2847 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2848 (this can affect the format of dates).
2850 28. exim_tidydb, when checking for the continued existence of a message for
2851 which it has found a message-specific retry record, was not finding
2852 messages that were in split spool directories. Consequently, it was
2853 deleting retry records that should have stayed in existence.
2855 29. eximstats updated to version 1.35
2856 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2857 1.35 - bugfix such that pie charts by volume are generated correctly
2859 30. The SPA authentication driver was not abandoning authentication and moving
2860 on to the next authenticator when an expansion was forced to fail,
2861 contradicting the general specification for all authenticators. Instead it
2862 was generating a temporary error. It now behaves as specified.
2864 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2865 (the order specifies the preference for clients). The order is now AES256,
2866 AES128, 3DES, ARCFOUR128.
2868 31. Small patch to Sieve code - explicitly set From: when generating an
2871 32. Exim crashed if a remote delivery caused a very long error message to be
2872 recorded - for instance if somebody sent an entire SpamAssassin report back
2873 as a large number of 550 error lines. This bug was coincidentally fixed by
2874 increasing the size of one of Exim's internal buffers (big_buffer) that
2875 happened as part of the Exiscan merge. However, to be on the safe side, I
2876 have made the code more robust (and fixed the comments that describe what
2879 33. Some experimental protocols are using DNS PTR records for new purposes. The
2880 keys for these records are domain names, not reversed IP addresses. The
2881 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2882 leaves it alone. Component reversal etc. now happens only for IP addresses.
2885 34. The host_aton() function is supposed to be passed a string that is known
2886 to be a valid IP address. However, in the case of IPv6 addresses, it was
2887 not checking this. This is a hostage to fortune. Exim now panics and dies
2888 if the condition is not met. A case was found where this could be provoked
2889 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2890 components; fortuitously, this particular loophole had already been fixed
2891 by change 4.50/55 or 4.44/33 above.
2893 If there are any other similar loopholes, the new check in host_aton()
2894 itself should stop them being exploited. The report I received stated that
2895 data on the command line could provoke the exploit when Exim was running as
2896 exim, but did not say which command line option was involved. All I could
2897 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2898 running as the user.
2901 35. There was a buffer overflow vulnerability in the SPA authentication code
2902 (which came originally from the Samba project). I have added a test to the
2903 spa_base64_to_bits() function which I hope fixes it.
2906 36. The daemon start-up calls getloadavg() while still root for those OS that
2907 need the first call to be done as root, but it missed one case: when
2908 deliver_queue_load_max is set with deliver_drop_privilege. This is
2909 necessary for the benefit of the queue runner, because there is no re-exec
2910 when deliver_drop_privilege is set.
2912 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2913 was in use, was not putting the data itself into the right store pool;
2914 consequently, it could be overwritten for a subsequent message in the same
2915 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2918 38. Sometimes the final signoff response after QUIT could fail to get
2919 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2920 < 0 before doing a fflush(). This bug looks as though it goes back to the
2921 introduction of TLS in release 3.20, but "sometimes" must have been rare
2922 because the tests only now provoked it.
2928 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2929 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2930 Now it returns an integer. A number of calls were still expecting a T/F
2931 return. Fortuitously, in all cases, the tests worked in OK situations,
2932 which is the norm. However, things would have gone wrong on any write
2933 failures on the smtp file descriptor. This function is used when sending
2934 messages over SMTP and also when doing verify callouts.
2936 2. When Exim is called to do synchronous delivery of a locally submitted
2937 message (the -odf or -odi options), it no longer closes stderr before doing
2940 3. Implemented the mua_wrapper option.
2942 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2944 5. Implemented the functions header_remove(), header_testname(),
2945 header_add_at_position(), and receive_remove_recipient(), and exported them
2948 6. If an ACL "warn" statement specified the addition of headers, Exim already
2949 inserted X-ACL-Warn: at the start if there was no header name. However, it
2950 was not making this test for the second and subsequent header lines if
2951 there were newlines in the string. This meant that an invalid header could
2952 be inserted if Exim was badly configured.
2954 7. Allow an ACL "warn" statement to add header lines at the start or after all
2955 the Received: headers, as well as at the end.
2957 8. Added the rcpt_4xx retry error code.
2959 9. Added postmaster_mailfrom=xxx to callout verification option.
2961 10. Added mailfrom=xxxx to the callout verification option, for verify=
2964 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2965 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2967 12. Inserted some casts to stop certain compilers warning when using pointer
2968 differences as field lengths or precisions in printf-type calls (mostly
2969 affecting debugging statements).
2971 13. Added optional readline() support for -be (dynamically loaded).
2973 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2974 same clock tick as a message's arrival, so that its received time was the
2975 same as the "first fail" time on the retry record, and that message
2976 remained on the queue past the ultimate address timeout, every queue runner
2977 would try a delivery (because it was past the ultimate address timeout) but
2978 after another failure, the ultimate address timeout, which should have then
2979 bounced the address, did not kick in. This was a "< instead of <=" error;
2980 in most cases the first failure would have been in the next clock tick
2981 after the received time, and all would be well.
2983 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2984 being recognized when the domain list was tested by the match_domain
2985 condition in an expansion string.
2987 16. Added the ${str2b64: operator.
2989 17. Exim was always calling setrlimit() to set a large limit for the number of
2990 processes, without checking whether the existing limit was already
2991 adequate. (It did check for the limit on file descriptors.) Furthermore,
2992 errors from getrlimit() and setrlimit() were being ignored. Now they are
2993 logged to the main and panic logs, but Exim does carry on, to try to do its
2994 job under whatever limits there are.
2996 18. Imported PCRE 5.0.
2998 19. Trivial typo in log message " temporarily refused connection" (the leading
3001 20. If the log selector return_path_on_delivery was set and an address was
3002 redirected to /dev/null, the delivery process crashed because it assumed
3003 that a return path would always be set for a "successful" delivery. In this
3004 case, the whole delivery is bypassed as an optimization, and therefore no
3007 21. Internal re-arrangement: the function for sending a challenge and reading
3008 a response while authentication was assuming a zero-terminated challenge
3009 string. It's now changed to take a pointer and a length, to allow for
3010 binary data in such strings.
3012 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3014 23. Exim was not respecting finduser_retries when seeking the login of the
3015 uid under which it was called; it was always trying 10 times. (The default
3016 setting of finduser_retries is zero.) Also, it was sleeping after the final
3017 failure, which is pointless.
3019 24. Implemented tls_on_connect_ports.
3021 25. Implemented acl_smtp_predata.
3023 26. If the domain in control=submission is set empty, Exim assumes that the
3024 authenticated id is a complete email address when it generates From: or
3025 Sender: header lines.
3027 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3028 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3029 chown and chgrp in /bin and hostname in /usr/bin.
3031 28. Exim was keeping the "process log" file open after each use, just as it
3032 does for the main log. This opens the possibility of it remaining open for
3033 long periods when the USR1 signal hits a daemon. Occasional processlog
3034 errors were reported, that could have been caused by this. Anyway, it seems
3035 much more sensible not to leave this file open at all, so that is what now
3038 29. The long-running daemon process does not normally write to the log once it
3039 has entered its main loop, and it closes the log before doing so. This is
3040 so that log files can straightforwardly be renamed and moved. However,
3041 there are a couple of unusual error situations where the daemon does write
3042 log entries, and I had neglected to close the log afterwards.
3044 30. The text of an SMTP error response that was received during a remote
3045 delivery was being truncated at 512 bytes. This is too short for some of
3046 the long messages that one sometimes sees. I've increased the limit to
3049 31. It is now possible to make retry rules that apply only when a message has a
3050 specific sender, in particular, an empty sender.
3052 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3053 it possible to be selective about when SMTP synchronization is enforced.
3055 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3057 32. Implemented hosts_connection_nolog.
3059 33. Added an ACL for QUIT.
3061 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3064 35. Added mailbox_size and mailbox_filecount to appendfile.
3066 36. Added control = no_multiline_responses to ACLs.
3068 37. There was a bug in the logic of the code that waits for the clock to tick
3069 in the case where the clock went backwards by a substantial amount such
3070 that the microsecond fraction of "now" was more than the microsecond
3071 fraction of "then" (but the whole seconds number was less).
3073 38. Added support for the libradius Radius client library this is found on
3074 FreeBSD (previously only the radiusclient library was supported).
3080 1. When certain lookups returned multiple values in the form name=value, the
3081 quoting of the values was not always being done properly. Specifically:
3082 (a) If the value started with a double quote, but contained no whitespace,
3084 (b) If the value contained whitespace other than a space character (i.e.
3085 tabs or newlines or carriage returns) it was not quoted.
3086 This fix has been applied to the mysql and pgsql lookups by writing a
3087 separate quoting function and calling it from the lookup code. The fix
3088 should probably also be applied to nisplus, ibase and oracle lookups, but
3089 since I cannot test any of those, I have not disturbed their existing code.
3091 2. A hit in the callout cache for a specific address caused a log line with no
3092 reason for rejecting RCPT. Now it says "Previous (cached) callout
3093 verification failure".
3095 3. There was an off-by-one bug in the queryprogram router. An over-long
3096 return line was truncated at 256 instead of 255 characters, thereby
3097 overflowing its buffer with the terminating zero. As well as fixing this, I
3098 have increased the buffer size to 1024 (and made a note to document this).
3100 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3101 when Exim is waiting for an SMTP response from a remote server, Exim
3102 restarts its select() call on the socket, thereby resetting its timeout.
3103 This is not a problem when such interrupts are rare. Somebody set up a cron
3104 job to run exiwhat every 2 minutes, which is less than the normal select()
3105 timeout (5 or 10 minutes). This meant that the select() timeout never
3106 kicked in because it was always reset. I have fixed this by comparing the
3107 time when an interrupt arrives with the time at the start of the first call
3108 to select(). If more time than the timeout has elapsed, the interrupt is
3109 treated as a timeout.
3111 5. Some internal re-factoring in preparation for the addition of Sieve
3112 extensions (by MH). In particular, the "personal" test is moved to a
3113 separate function, and given an option for scanning Cc: and Bcc: (which is
3114 not set for Exim filters).
3116 6. When Exim created an email address using the login of the caller as the
3117 local part (e.g. when creating a From: or Sender: header line), it was not
3118 quoting the local part when it contained special characters such as @.
3120 7. Installed new OpenBSD configuration files.
3122 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3123 try to make them clearer.
3125 9. Callout options, other than the timeout value, were being ignored when
3126 verifying sender addresses in header lines. For example, when using
3128 verify = header_sender/callout=no_cache
3130 the cache was (incorrectly) being used.
3132 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3133 only the Cygwin environment.
3135 11. When return_path_on_delivery was set as a log selector, if different remote
3136 addresses in the same message used different return paths and parallel
3137 remote delivery occurred, the wrong values would sometimes be logged.
3138 (Whenever a remote delivery process finished, the return path value from
3139 the most recently started remote delivery process was logged.)
3141 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3142 lines when AUTH and/or TLS are in use. This is the "received protocol"
3143 field. Exim used to use "asmtp" for authenticated SMTP, without any
3144 indication (in the protocol name) for TLS use. Now it follows the RFC and
3145 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3146 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3147 names appear in log lines as well as in Received: header lines.
3149 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3150 extensions, and comparison tests, and to fix some bugs.
3152 14. Changes to the "personal" filter test:
3154 (1) The test was buggy in that it was just doing the equivalent of
3155 "contains" tests on header lines. For example, if a user's address was
3156 anne@some.where, the "personal" test would incorrectly be true for
3158 To: susanne@some.where
3160 This test is now done by extracting each address from the header in turn,
3161 and checking the entire address. Other tests that are part of "personal"
3162 are now done using regular expressions (for example, to check local parts
3163 of addresses in From: header lines).
3165 (2) The list of non-personal local parts in From: addresses has been
3166 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3167 taken from the Sieve specification recommendations.
3169 (3) If the message contains any header line starting with "List-" it is
3170 treated as non-personal.
3172 (4) The test for "circular" in the Subject: header line has been removed
3173 because it now seems ill-conceived.
3175 15. Minor typos in src/EDITME comments corrected.
3177 16. Installed latest exipick from John Jetmore.
3179 17. If headers_add on a router specified a text string that was too long for
3180 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3181 of string_sprintf() is now avoided.
3183 18. $message_body_size was not set (it was always zero) when running the DATA
3184 ACL and the local_scan() function.
3186 19. For the "mail" command in an Exim filter, no default was being set for
3187 the once_repeat time, causing a random time value to be used if "once" was
3188 specified. (If the value happened to be <= 0, no repeat happened.) The
3189 default is now 0s, meaning "never repeat". The "vacation" command was OK
3190 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3191 (I found it when inspecting the code).
3193 20. There is now an overall timeout for performing a callout verification. It
3194 defaults to 4 times the callout timeout, which applies to individual SMTP
3195 commands during the callout. The overall timeout applies when there is more
3196 than one host that can be tried. The timeout is checked before trying the
3197 next host. This prevents very long delays if there are a large number of
3198 hosts and all are timing out (e.g. when the network connections are timing
3199 out). The value of the overall timeout can be changed by specifying an
3200 additional sub-option for "callout", called "maxwait". For example:
3202 verify = sender/callout=5s,maxwait=20s
3204 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3205 to the end before writing, but this should make it even safer).
3207 22. Exim was forgetting that it had advertised PIPELINING for the second and
3208 subsequent messages on an SMTP connection. It was also not resetting its
3209 memory on STARTTLS and an internal HELO.
3211 23. When Exim logs an SMTP synchronization error within a session, it now
3212 records whether PIPELINING has been advertised or not.
3214 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3215 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3216 rather than long int.
3218 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3220 26. Added the never_mail option to autoreply.
3226 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3227 crash if the getsockname() call failed; this can happen if a connection is
3228 closed very soon after it is established. The problem was simply in the
3229 order in which certain operations were done, causing Exim to try to write
3230 to the SMTP stream before it had set up the file descriptor. The bug has
3231 been fixed by making things happen in the correct order.
3237 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3238 before the connection was closed, thus losing the rejection response.
3240 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3241 some early Solaris releases, but causes trouble in current releases where
3242 socklen_t is defined.
3244 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3247 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3250 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3251 that releasing the top of it at the end releases what was used for sub-
3252 expansions (unless the block got too big). However, discard this block if
3253 the first thing is a variable or header, so that we can use its block when
3254 it is dynamic (useful for very large $message_headers, for example).
3256 6. Lookups now cache *every* query, not just the most recent. A new, separate
3257 store pool is used for this. It can be recovered when all lookup caches are
3258 flushed. Lookups now release memory at the end of their result strings.
3259 This has involved some general refactoring of the lookup sources.
3261 7. Some code has been added to the store_xxx() functions to reduce the amount
3262 of flapping under certain conditions.
3264 8. log_incoming_interface used to affect only the <= reception log lines. Now
3265 it causes the local interface and port to be added to several more SMTP log
3266 lines, for example "SMTP connection from", and rejection lines.
3268 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3270 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3272 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3273 could be overwritten at the end of the current message (or the start of a
3274 new message if it was set in a HELO ACL). The value is now preserved for
3275 the duration of the SMTP connection.
3277 12. If a transport had a headers_rewrite setting, and a matching header line
3278 contained an unqualified address, that address was qualified, even if it
3279 did not match any rewriting rules. The underlying bug was that the values
3280 of the flags that permit the existence of unqualified sender and recipient
3281 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3282 non-local messages, and by -bnq for local messages) were not being
3283 preserved with the message after it was received.
3285 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3286 "next input=" as part of the text comprising the host identity instead of
3287 the correct text. The code was using the same buffer for two different
3288 strings. However, depending on which order the printing function evaluated
3289 its arguments, the bug did not always show up. Under Linux, for example, my
3290 test suite worked just fine.
3292 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3293 doesn't work with some older versions of Perl. It has been changed to "my",
3294 which in any case is probably the better facility to use.
3296 15. A really picky compiler found some instances of statements for creating
3297 error messages that either had too many or two few arguments for the format
3300 16. The size of the buffer for calls to the DNS resolver has been increased
3301 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3302 for addresses that have a lot of PTR records. This alleviates a problem; it
3303 does not fully solve it.
3305 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3306 buffer now truncates the list and logs the incident, which is the same
3307 action as happens when Exim is looking up a host name and its aliases.
3308 Previously in this situation something unpredictable would happen;
3309 sometimes it was "internal error: store_reset failed".
3311 18. If a server dropped the connection unexpectedly when an Exim client was
3312 using GnuTLS and trying to read a response, the client delivery process
3313 crashed while trying to generate an error log message.
3315 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3316 string, for example:
3318 warn message = H1: something\nH2: something
3320 the text was added as a single header line from Exim's point of view
3321 though it ended up OK in the delivered message. However, searching for the
3322 second and subsequent header lines using $h_h2: did not work. This has been
3323 fixed. Similarly, if a system filter added multiple headers in this way,
3324 the routers could not see them.
3326 20. Expanded the error message when iplsearch is called with an invalid key to
3327 suggest using net-iplsearch in a host list.
3329 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3330 ACLs are no longer actually imposed (and a message to that effect is
3333 22. If a "gecos" field in a passwd entry contained escaped characters, in
3334 particular, if it contained a \" sequence, Exim got it wrong when building
3335 a From: or a Sender: header from that name. A second bug also caused
3336 incorrect handling when an unquoted " was present following a character
3337 that needed quoting.
3339 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3340 was not being matched caselessly.
3342 24. Arranged for all hyphens in the exim.8 source to be escaped with
3345 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3346 properly. Recipient callout cache records were still being keyed to include
3347 the sender, even when use_sender was set false. This led to far more
3348 callouts that were necessary. The sender is no longer included in the key
3349 when use_sender is false.
3351 26. Added "control = submission" modifier to ACLs.
3353 27. Added the ${base62d: operator to decode base 62 numbers.
3355 28. dnsdb lookups can now access SRV records.
3357 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3358 the configuration file.
3360 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3361 (-v) mode. This makes the output for a verbose queue run more intelligible.
3363 31. Added a use_postmaster feature to recipient callouts.
3365 32. Added the $body_zerocount variable, containing the number of binary zero
3366 bytes in the message body.
3368 33. The time of last modification of the "new" subdirectory is now used as the
3369 "mailbox time last read" when there is a quota error for a maildir
3372 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3374 35. Added +ignore_unknown as a special item in host lists.
3376 36. Code for decoding IPv6 addresses in host lists is now included, even if
3377 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3378 address was recognized as an IP address, but was then not correctly decoded
3379 into binary, causing unexpected and incorrect effects when compared with
3386 1. Very minor rewording of debugging text in manualroute to say "list of
3387 hosts" instead of "hostlist".
3389 2. If verify=header_syntax was set, and a header line with an unqualified
3390 address (no domain) and a large number of spaces between the end of the
3391 name and the colon was received, the reception process suffered a buffer
3392 overflow, and (when I tested it) crashed. This was caused by some obsolete
3393 code that should have been removed. The fix is to remove it!
3395 3. When running in the test harness, delay a bit after writing a bounce
3396 message to get a bit more predictability in the log output.
3398 4. Added a call to search_tidyup() just before forking a reception process. In
3399 theory, someone could use a lookup in the expansion of smtp_accept_max_
3400 per_host which, without the tidyup, could leave open a database connection.
3402 5. Added the variables $recipient_data and $sender_data which get set from a
3403 lookup success in an ACL "recipients" or "senders" condition, or a router
3404 "senders" option, similar to $domain_data and $local_part_data.
3406 6. Moved the writing of debug_print from before to after the "senders" test
3409 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3410 problems for message scanning, either using a data ACL, or using
3411 local_scan() because the Received: header was not generated till after they
3412 were called (in order to set the time as the time of reception completion).
3413 I have revised the way this works. The header is now generated after the
3414 body is received, but before the ACL or local_scan() are called. After they
3415 are run, the timestamp in the header is updated.
3421 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3422 before starting a queue runner without re-exec. This happened only when
3423 deliver_drop_privilege was set or when the Exim user was set to root. The
3424 effect of the bug was that timeouts during subsequent deliveries caused
3425 crashes instead of being properly handled. The handler is now left at its
3426 default (and expected) setting.
3428 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3429 message, again when deliver_drop_privilege is set or Exim is run as root.
3430 The bug described in (1) was not present in this case, but the tidying up
3431 of the other signals was missing. I have made the two cases consistent.
3433 3. The ignore_target_hosts setting on a manualroute router was being ignored
3434 for hosts that were looked up using the /MX notation.
3436 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3439 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3440 operated on the sender address. After changing the $sender_address to <>
3441 for the sender address verify, Exim was re-instated it as the original
3442 (before rewriting) address, but remembering that it had rewritten it, so it
3443 wasn't rewriting it again. This bug also had the effect of breaking the
3444 sender address verification caching when the sender address was rewritten.
3446 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3447 This has been changed so that if the ip literal address matches
3448 ignore_target_hosts, the router declines.
3450 7. Added expansion conditions match_domain, match_address, and match_local_
3451 part (NOT match_host).
3453 8. The placeholder for the Received: header didn't have a length field set.
3455 9. Added code to Exim itself and to exim_lock to test for a specific race
3456 condition that could lead to file corruption when using MBX delivery. The
3457 issue is with the lockfile that is created in /tmp. If this file is removed
3458 after a process has opened it but before that process has acquired a lock,
3459 there is the potential for a second process to recreate the file and also
3460 acquire a lock. This could lead to two Exim processes writing to the file
3461 at the same time. The added code performs the same test as UW imapd; it
3462 checks after acquiring the lock that its file descriptor still refers to
3463 the same named file.
3465 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3466 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3469 11. Added the smtp_active_hostname option. If used, this will typically be made
3470 to depend on the incoming interface address. Because $interface_address is
3471 not set up until the daemon has forked a reception process, error responses
3472 that can happen earlier (such as "too many connections") no longer contain
3475 12. If an expansion in a condition on a "warn" statement fails because a lookup
3476 defers, the "warn" statement is abandoned, and the next ACL statement is
3477 processed. Previously this caused the whole ACL to be aborted.
3479 13. Added the iplsearch lookup type.
3481 14. Added ident_timeout as a log selector.
3483 15. Added tls_certificate_verified as a log selector.
3485 16. Added a global option tls_require_ciphers (compare the smtp transport
3486 option of the same name). This controls incoming TLS connections.
3488 17. I finally figured out how to make tls_require_ciphers do a similar thing
3489 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3490 before starting the TLS session.
3492 18. Tabs are now shown as \t in -bP output.
3494 19. If the log selector return_path_on_delivery was set, Exim crashed when
3495 bouncing a message because it had too many Received: header lines.
3497 20. If two routers both had headers_remove settings, and the first one included
3498 a superfluous trailing colon, the final name in the first list and the
3499 first name in the second list were incorrectly joined into one item (with a
3500 colon in the middle).
3506 1. Added -C and -D options to the exinext utility, mainly to make it easier
3507 to include in the automated testing, but these could be helpful when
3508 multiple configurations are in use.
3510 2. The exinext utility was not formatting the output nicely when there was
3511 an alternate port involved in the retry record key, nor when there was a
3512 message id as well (for retries that were specific to a specific message
3513 and a specific host). It was also confused by IPv6 addresses, because of
3514 the additional colons they contain. I have fixed the IPv4 problem, and
3515 patched it up to do a reasonable job for IPv6.
3517 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3518 delivery, the log line now contains "pipelined" if PIPELINING was used.
3520 4. An SMTP transport process used to panic and die if the bind() call to set
3521 an explicit outgoing interface failed. This has been changed; it is now
3522 treated in the same way as a connect() failure.
3524 5. A reference to $sender_host_name in the part of a conditional expansion
3525 that was being skipped was still causing a DNS lookup. This no longer
3528 6. The def: expansion condition was not recognizing references to header lines
3529 that used bh_ and bheader_.
3531 7. Added the _cache feature to named lists.
3533 8. The code for checking quota_filecount in the appendfile transport was
3534 allowing one more file than it should have been.
3536 9. For compatibility with Sendmail, the command line option
3544 and sets the incoming protocol and host name (for trusted callers). The
3545 host name and its colon can be omitted when only the protocol is to be set.
3546 Note the Exim already has two private options, -pd and -ps, that refer to
3547 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3548 "s", but I don't think that's a major issue.
3550 10. A number of refactoring changes to the code, none of which should affect
3553 (a) The number of logging options was getting close to filling up the
3554 32-bit word that was used as a bit map. I have split them into two classes:
3555 those that are passed in the argument to log_write(), and those that are
3556 only ever tested independently outside of that function. These are now in
3557 separate 32-bit words, so there is plenty of room for expansion again.
3558 There is no change in the user interface or the logging behaviour.
3560 (b) When building, for example, log lines, the code previously used a
3561 macro that called string_cat() twice, in order to add two strings. This is
3562 not really sufficiently general. Furthermore, there was one instance where
3563 it was actually wrong because one of the argument was used twice, and in
3564 one call a function was used. (As it happened, calling the function twice
3565 did not affect the overall behaviour.) The macro has been replaced by a
3566 function that can join an arbitrary number of extra strings onto a growing
3569 (c) The code for expansion conditions now uses a table and a binary chop
3570 instead of a serial search (which was left over from when there were very
3571 few conditions). Also, it now recognizes conditions like "pam" even when
3572 the relevant support is not compiled in: a suitably worded error message is
3573 given if an attempt is made to use such a condition.
3575 11. Added ${time_interval:xxxxx}.
3577 12. A bug was causing one of the ddress fields not to be passed back correctly
3578 from remote delivery subprocesses. The field in question was not being
3579 subsequently used, so this caused to problems in practice.
3581 13. Added new log selectors queue_time and deliver_time.
3583 14. Might have fixed a bug in maildirsizefile handling that threw up
3584 "unexpected character" debug warnings, and recalculated the data
3585 unnecessarily. In any case, I expanded the warning message to give more
3588 15. Added the message "Restricted characters in address" to the statements in
3589 the default ACL that block characters like @ and % in local parts.
3591 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3592 Three changes have been made:
3594 (a) There was a serious bug; a negative response to MAIL caused the whole
3595 recipient domain to be cached as invalid, thereby blocking all messages
3596 to all local parts at the same domain, from all senders. This bug has
3597 been fixed. The domain is no longer cached after a negative response to
3598 MAIL if the sender used is not empty.
3600 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3603 (c) A new callout option, "use_sender" has been added for people who want
3604 the modified behaviour.
3610 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3613 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3614 indeed breaks things for older releases.
3616 3. Added additional logging to the case where there is a problem reading data
3617 from a filter that is running in a subprocess using a pipe, in order to
3618 try to track down a specific problem.
3620 4. Testing facility fudge: when running in the test harness and attempting
3621 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3622 getting "No route to host". Convert this to a timeout.
3624 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3627 6. Some OS don't have socklen_t but use size_t instead. This affects the
3628 fifth argument of getsockopt() amongst other things. This is now
3629 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3630 can be set for individual OS. I have set it for SunOS5, OSF1, and
3631 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3632 some earlier ones do not.
3634 7. Change 4.30/15 was not doing the test caselessly.
3636 8. The standard form for an IPv6 address literal was being rejected by address
3637 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3638 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3639 this, as well as the form without the "IPv6" on the front (but only when
3640 address literals are enabled, of course).
3642 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3644 10. Exim crashed if a message with an empty sender address specified by -f
3645 encountered a router with an errors_to setting. This could be provoked only
3646 by a command such as
3650 where an empty string was supplied; "<>" did not hit this bug.
3652 11. Installed PCRE release 4.5.
3654 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3655 remained set. It is now erased.
3657 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3658 times from message ids (which are base 36 rather than the normal 62).
3660 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3661 were being counted as actual protocol errors, and logged if the log
3662 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3663 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3664 and DATA following a set of rejected RCPTs do not count as protocol errors.
3665 In other words, Exim assumes they were pipelined, though this may not
3666 actually be the case. Of course, in all cases the client gets an
3667 appropriate error code.
3669 15. If a lookup fails in an ACL condition, a message about the failure may
3670 be available; it is used if testing the ACL cannot continue, because most
3671 such messages specify what the cause of the deferral is. However, some
3672 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3673 that caused an old message to be retained and used if a later statement
3674 caused a defer, replacing the real cause of the deferral.
3676 16. If an IP address had so many PTR records that the DNS lookup buffer
3677 was not large enough to hold them, Exim could crash while trying to process
3678 the truncated data. It now detects and logs this case.
3680 17. Further to 4.21/58, another change has been made: if (and only if) the
3681 first line of a message (the first header line) ends with CRLF, a bare LF
3682 in a subsequent header line has a space inserted after it, so as not to
3683 terminate the header.
3685 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3686 unnecessarily, used atoi() instead of strtol(), and didn't check the
3687 termination when getting file sizes from file names by regex.
3689 19. Completely re-implemented the support for maildirsize files, in the light
3690 of a number of problems with the previous contributed implementation
3691 (4.30/29). In particular:
3693 . If the quota is zero, the maildirsize file is maintained, but no quota is
3696 . If the maildir directory does not exist, it is created before any attempt
3697 to write a maildirsize file.
3699 . The quota value in the file is just a cache; if the quota is changed in
3700 the transport, the new value overrides.
3702 . A regular expression is available for excluding directories from the
3705 20. The autoreply transport checks the characters in options that define the
3706 message's headers; it allows continued headers, but it was checking with
3707 isspace() after an embedded newline instead of explicitly looking for a
3710 21. If all the "regular" hosts to which an address was routed had passed their
3711 expiry times, and had not reached their retry times, the address was
3712 bounced, even if fallback hosts were defined. Now Exim should go on to try
3715 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3716 equivalent code in the SMTP transport. Some hosts send humungous responses
3717 to HELO/EHLO, more than 1024 it seems.
3719 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3720 gives compiler warnings in some environments. I've now done it "properly",
3723 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3724 (because of problems with the built-in one) was declared to return uschar *
3725 instead of char *, causing compiler failure.
3727 25. Fixed a file descriptor leak when processing alias/forward files.
3729 26. Fixed a minor format string issue in dbfn.c.
3731 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3733 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3734 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3737 29. When a sender address is verified, it is cached, to save repeating the test
3738 when there is more than one recipient in a message. However, when the
3739 verification involves a callout, it is possible for different callout
3740 options to be set for different recipients. It is too complicated to keep
3741 track of this in the cache, so now Exim always runs a verification when a
3742 callout is required, relying on the callout cache for the optimization.
3743 The overhead is duplication of the address routing, but this should not be
3746 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3747 to be verified with callout=postmaster, and the main callout worked but the
3748 postmaster check failed, the verification correctly failed. However, if a
3749 subsequent RCPT command asked for sender verification *without* the
3750 postmaster check, incorrect caching caused this verification also to fail,
3753 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3754 it was not caching the DNS options (qualify_single, search_parents) that
3755 were used when the lookup failed. A subsequent lookup with different
3756 options therefore always gave the same answer, though there were cases
3757 where it should not have. (Example: a "domains = !$mx_any" option on a
3758 dnslookup router: the "domains" option is always processed without any
3759 widening, but the router might have qualify_single set.) Now Exim uses the
3760 cached value only when the same options are set.
3762 32. Added John Jetmore's "exipick" utility to the distribution.
3764 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3765 than a timeout (e.g. a certificate is required, and is not provided), an
3766 Exim server now closes the connection immediately. Previously it waited for
3767 the client to close - but if the client is SSL, it seems that they each
3768 wait for each other, leading to a delay before one of them times out.
3770 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3771 maintained 0.8.x compatibility because I don't think many are using it, and
3772 it is clearly obsolete.
3774 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3777 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3778 client certificate was expired. A simple patch fixes this, though I don't
3779 understand the full logic of why the verify callback is called multiple
3782 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3783 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3784 which causes problems with some clients (such as the Certicom SSL Plus
3785 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3786 disables the coutermeasure allowing Eudora to connect."
3788 38. Exim was not checking that a write() to a log file succeeded. This could
3789 lead to Bad Things if a log got too big, in particular if it hit a file
3790 size limit. Exim now panics and dies if it cannot write to a log file, just
3791 as it does if it cannot open a log file.
3793 39. Modified OS/Makefile-Linux so that it now contains
3795 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3797 The two -D definitions ensure that Exim is compiled with large file
3798 support, which makes it possible to handle log files that are bigger than
3801 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3802 instance) a domain was checked against a named list that involved a lookup,
3803 causing $domain_data to be set, then another domain was checked against the
3804 same list, then the first domain was re-checked, the value of $domain_data
3805 after the final check could be wrong. In particular, if the second check
3806 failed, it could be set empty. This bug probably also applied to
3809 41. The strip_trailing_dot option was not being applied to the address given
3810 with the -f command-line option.
3812 42. The code for reading a message's header from the spool was incrementing
3813 $received_count, but never initializing it. This meant that the value was
3814 incorrect (doubled) while delivering a message in the same process in which
3815 it was received. In the most common configuration of Exim, this never
3816 happens - a fresh exec is done - but it can happen when
3817 deliver_drop_privilege is set.
3819 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3820 it now includes up to 150 characters of the unexpected data in the log
3823 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3824 and building data strings. The size of both of these buffers was 10 000
3825 bytes - far larger than anybody would *ever* want, thought I. Needless to
3826 say, somebody hit the limit. I have increased the maximum line length to
3827 20 000 and the maximum data length of concatenated lines to 100 000. I have
3828 also fixed two bugs, because there was no checking on these buffers. Tsk,
3829 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3830 buffer is too small.
3832 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3833 lsearch lookups. Now it does.
3835 46. When parsing a route_list item in a manualroute router, a fixed-length
3836 buffer was used for the list of hosts. I made this 1024 bytes long,
3837 thinking that nobody would ever have a list of hosts that long. Wrong.
3838 Somebody had a whole pile of complicated expansion conditions, and the
3839 string was silently truncated, leading to an expansion error. It turns out
3840 that it is easier to change to an unlimited length (owing to other changes
3841 that have happened since this code was originally written) than to build
3842 structure for giving a limitation error. The length of the item that
3843 expands into the list of hosts is now unlimited.
3845 47. The lsearch lookup could not handle data where the length of text line was
3846 more than 4095 characters. Such lines were truncated, leading to shortened
3847 data being returned. It should now handle lines of any length.
3849 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3850 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3853 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3854 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3855 no longer happens in comment lines. A list of replacements is now placed
3856 at the head of all of the source files, except those whose only change is
3857 to replace PERL_COMMAND in the very first #! line.
3859 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3860 messages on the queue, with a bottom-up merge sort, using code contributed
3861 by Michael Haardt. This should make operations like -bp somewhat faster on
3862 large queues. It won't affect queue runners, except when queue_run_in_order
3865 51. Installed eximstats 1.31 in the distribution.
3867 52. Added support for SRV lookups to the dnslookup router.
3869 53. If an ACL referred to $message_body or $message_body_end, the value was not
3870 reset for any messages that followed in the same SMTP session.
3872 54. The store-handling optimization for building very long strings was not
3873 differentiating between the different store pools. I don't think this
3874 actually made any difference in practice, but I've tidied it.
3876 55. While running the routers to verify a sender address, $sender_address
3877 was still set to the sender address. This is wrong, because when routing to
3878 send a bounce to the sender, it would be empty. Therefore, I have changed
3879 it so that, while verifying a sender address, $sender_address is set to <>.
3880 (There is no change to what happens when verifying a recipient address.)
3882 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3883 target A or AAAA records (if not already returned) without resetting the
3884 qualify_single or search_parents options of the DNS resolver. These are
3885 inappropriate in this case because the targets of MX and SRV records must
3886 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3887 target that, when qualified, matched something in the local domain. These
3888 two options are now turned off when doing these lookups.
3890 57. It seems that at least some releases of Reiserfs (which does not have the
3891 concept of a fixed number of inodes) returns zero and not -1 for the
3892 number of available inodes. This interacted badly with check_spool_inodes,
3893 which assumed that -1 was the "no such thing" setting. What I have done is
3894 to check that the total number of inodes is greater than zero before doing
3895 the test of how many are available.
3897 58. When a "warn" ACL statement has a log_message modifier, the message is
3898 remembered, and not repeated. This is to avoid a lot of repetition when a
3899 message has many recipients that cause the same warning to be written.
3900 Howewer, Exim was preserving the list of already written lines for an
3901 entire SMTP session, which doesn't seem right. The memory is now reset if a
3902 new message is started.
3904 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3905 debugging output unless log_rewrite was also set.
3907 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3908 of (int)(handle) when we know that handle contains (void *)(-1).
3910 61. The Exim daemon panic-logs an error return when it closes the incoming
3911 connection. However "connection reset by peer" seems to be common, and
3912 isn't really an error worthy of noting specially, so that particular error
3915 62. When Exim is trying to find all the local interfaces, it used to panic and
3916 die if the ioctl to get the interface flags failed. However, it seems that
3917 on at least one OS (Solaris 9) it is possible to have an interface that is
3918 included in the list of interfaces, but for which you get a failure error
3919 for this call. This happens when the interface is not "plumbed" into a
3920 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3921 failure of the "get flags" call assumes that the interface is down.
3923 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3924 makes life easier for people who are doing arithmetic on fields extracted
3925 from dates, where you often get leading zeros that should not be
3926 interpreted as octal.
3928 64. Added qualify_domain to the redirect router, to override the global
3931 65. If a pathologically long header line contained very many addresses (the
3932 report of this problem mentioned 10 000) and each of them was rewritten,
3933 Exim could use up a very large amount of memory. (It kept on making new
3934 copies of the header line as it rewrote, and never released the old ones.)
3935 At the expense of a bit more processing, the header rewriting function has
3936 been changed so that it no longer eats memory in this way.
3938 66. The generation of the Received: header has been moved from the time that a
3939 message starts to be received, to the time that it finishes. The timestamp
3940 in the Received: header should now be very close to that of the <= log
3941 line. There are two side-effects of this change:
3943 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3944 logged header lines no longer include the local Received: line, because
3945 it has not yet been created. The same applies to a copy of the message
3946 that is returned to a non-SMTP sender when a message is rejected.
3948 (b) When a filter file is tested using -bf, no additional Received: header
3949 is added to the test message. After some thought, I decided that this
3952 This change does not affect the value of $received_for. It is still set
3953 after address rewriting, but before local_scan() is called.
3955 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3957 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3958 gave an unhelpful panic error message, and a defer error. I have managed to
3959 change this behaviour so that it now rejects any supplied certificate,
3960 which seems right, as the list of acceptable certificates is empty.
3962 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3963 gave an unhelpful defer error. I have not managed to make this reject any
3964 supplied certificates, but the error message it gives is "no certificate
3965 supplied", which is not helpful.
3967 70. exigrep's output now also includes lines that are not associated with any
3968 message, but which match the given pattern. Implemented by a patch from
3969 Martin Sluka, which also tidied up the Perl a bit.
3971 71. Recipient callout verification, like sender verification, was using <> in
3972 the MAIL FROM command. This isn't really the right thing, since the actual
3973 sender may affect whether the remote host accepts the recipient or not. I
3974 have changed it to use the actual sender in the callout; this means that
3975 the cache record is now keyed on a recipient/sender pair, not just the
3976 recipient address. There doesn't seem to be a real danger of callout loops,
3977 since a callout by the remote host to check the sender would use <>.
3978 [SEE ABOVE: changed after hitting problems.]
3980 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3981 temporary errors. However, in the case of such a code being given after
3982 the end of a data transmission (i.e. after ".") Exim was failing to write
3983 a retry record for the message. (Yes, there was some broken host that was
3984 actually sending 8xx at this point.)
3986 73. An unknown lookup type in a host list could cause Exim to panic-die when
3987 the list was checked. (An example that provoked this was putting <; in the
3988 middle of a list instead of at the start.) If this happened during a DATA
3989 ACL check, a -D file could be left lying around. This kind of configuration
3990 error no longer causes Exim to die; instead it causes a defer errror. The
3991 incident is still logged to the main and panic logs.
3993 74. Buglet left over from Exim 3 conversion. The message "too many messages
3994 in one connection" was written to the rejectlog but not the mainlog, except
3995 when address rewriting (yes!) was being logged.
3997 75. Added write_rejectlog option.
3999 76. When a system filter was run not as root (that is, when system_filter_user
4000 was set), the values of the $n variables were not being returned to the
4001 main process; thus, they were not subsequently available in the $sn
4004 77. Added +return_path_on_delivery log selector.
4006 78. A connection timeout was being treated differently from recipients deferred
4007 when testing hosts_max_try with a message that was older than the host's
4008 retry timeout. (The host should not be counted, thus allowing all hosts to
4009 be tried at least once before bouncing.) This may have been the cause of an
4010 occasionally reported bug whereby a message would remain on the queue
4011 longer than the retry timeout, but would be bounced if a delivery was
4012 forced. I say "may" because I never totally pinned down the problem;
4013 setting up timeout/retry tests is difficult. See also the next item.
4015 79. The ultimate address timeout was not being applied to errors that involved
4016 a combination of host plus message (for example, a timeout on a MAIL
4017 command). When an address resolved to a number of possible hosts, and they
4018 were not all tried for each delivery (e.g. because of hosts_max_try), a
4019 message could remain on the queue longer than the retry timeout.
4021 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4024 81. Fixed an obscure SMTP outgoing bug which required at least the following
4025 conditions: (a) there was another message waiting for the same server;
4026 (b) the server returned 5xx to all RCPT commands in the first message so
4027 that the message was not completed; (c) the server dropped the connection
4028 or gave a negative response to the RSET that Exim sends to abort the
4029 transaction. The observed case was a dropped connection after DATA that had
4030 been sent in pipelining mode. That is, the server had advertised PIPELINING
4031 but was not implementing it correctly. The effect of the bug was incorrect
4032 behaviour, such as trying another host, and this could lead to a crash.
4038 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4039 and daemon.c were passed as pointers to ints; they should have been
4040 pointers to socklen_t variables (which are typically unsigned ints).
4042 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4045 3. Fixed a really odd bug that affected only the testing scheme; patching a
4046 certain fixed string in the binary changed the value of another string that
4047 happened to be identical to the end of the original first string.
4049 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4050 name", it returns that address as the IP address. On some operating
4051 systems (e.g. Solaris), it also passes back the IP address string as the
4052 "host name". However, on others (e.g. Linux), it passes back an empty
4053 string. Exim wasn't checking for this, and was changing the host name to an
4054 empty string, assuming it had been canonicized.
4056 5. Although rare, it is permitted to have more than one PTR record for a given
4057 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4058 all the names associated with an address, because they do in Solaris.
4059 However, it seems that they do not in Linux for data that comes from the
4060 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4061 I found this out when I moved to a new Linux workstation and tried to run
4062 the Exim test suite.
4064 To get round this problem I have changed the code so that it now does its
4065 own call to the DNS to look up PTR records when searching for a host name.
4066 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4067 addresses that are only in /etc/hosts are still found.
4069 This behaviour is, however, controlled by an option called host_lookup_
4070 order, which defaults to "bydns:byaddr". If people want to use the other
4071 order, or indeed, just use one or the other means of lookup, they can
4072 specify it in this variable.
4074 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4075 some operating systems, this comes back from gethostbyaddr() as an empty
4076 string, and this is what Exim used to test for. However, it seems that in
4077 other systems, "." is yielded. Exim now tests for this case too.
4079 7. The values of check_spool_space and check_log_space are now held internally
4080 as a number of kilobytes instead of an absolute number of bytes. If a
4081 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4082 kilobyte. This means that much larger values can be stored.
4084 8. Exim monitor: an attempt to get the action menu when not actually pointing
4085 at a message produces an empty menu entitled "No message selected". This
4086 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4087 no entries in it ("Shell widget menu has zero width and/or height"). So I
4088 have added a single, blank menu entry in this case.
4090 9. Added ${quote_local_part.
4092 10. MIME decoding is now applied to the contents of Subject: header lines when
4095 11. Now that a reference to $sender_host_address automatically causes a reverse
4096 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4097 host lookup before query-style lookups in lists that might use this
4098 variable. This has therefore been abolished, and the "net-" prefix is no
4099 longer necessary for query-style lookups.
4101 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4102 have been a typo for LFLAGS, so it has been changed.
4104 13. The install script calls Exim with "-C /dev/null" in order to find the
4105 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4106 to be output. Howeve, since Exim outputs its version number before the
4107 error, it didn't break the script. It just looked ugly. I fixed this by
4108 always allowing "-C /dev/null" if the caller is root.
4110 14. Ignore overlarge ACL variable number when reading spool file - insurance
4111 against a later release with more variables having written the file.
4113 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4114 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4115 this, as well as the form without the "IPv6" on the front.
4117 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4118 OS/Makefile-Darwin file.
4120 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4121 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4123 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4124 variable. However, some OS use ints for inodes. I've added cast to long int
4125 to get rid of the compiler warning.
4127 19. I had forgotten to lock out "/../" in configuration file names when
4128 ALT_CONFIG_PREFIX was set.
4130 20. Routers used for verification do not need to specify transports. However,
4131 if such a router generated a host list, and callout was configured, Exim
4132 crashed, because it could not find a port number from the (non-existent)
4133 transport. It now assumes port 25 in this circumstance.
4135 21. Added the -t option to exigrep.
4137 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4138 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4139 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4140 them is set, LOOKUP_LSEARCH is forced.
4142 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4144 24. Added sender and host information to the "rejected by local_scan()" log
4145 line; previously there was no indication of these.
4147 25. Added .include_if_exists.
4149 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4150 Linux. It turns out that not all file systems support this. Apparently some
4151 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4152 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4153 not supported on the file descriptor, is now ignored when Exim is trying to
4154 sync a directory. This applies only to Linux.
4156 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4158 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4159 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4162 29. Added support for maildirsize files from supplied patch (modified a bit).
4164 30. The use of :fail: followed by an empty string could lead Exim to respond to
4165 sender verification failures with (e.g.):
4167 550 Verification failed for <xxx>
4168 550 Sender verify failed
4170 where the first response line was missing the '-' that indicates it is not
4171 the final line of the response.
4173 31. The loop for finding the name of the user that called Exim had a hardwired
4174 limit of 10; it now uses the value of finduser_retries, which is used for
4175 all other user lookups.
4177 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4180 33. Exim was neglecting to zero errno before one call of strtol() when
4181 expanding a string and expecting an integer value. On some systems this
4182 resulted in spurious "integer overflow" errors. Also, it was casting the
4183 result into an int without checking.
4185 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4186 did not work. The code looks as if it has *never* worked, though it appears
4187 to have been documented since at least releast 1.62. I have made it work.
4189 35. The "timeout_DNS" error in retry rules, also documented since at least
4190 1.62, also never worked. As it isn't clear exactly what this means, and
4191 clearly it isn't a major issue, I have abolished the feature by treating it
4192 as "timeout", and writing a warning to the main and panic logs.
4194 36. The display of retry rules for -brt wasn't always showing the error code
4197 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4198 timeout_connect_A, timeout_connect_MX.
4200 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4201 to the empty sender.
4203 39. The daemon was not analysing the content of -oX till after it had closed
4204 stderr and disconnected from the controlling terminal. This meant that any
4205 syntax errors were only noted on the panic log, and the return code from
4206 the command was 0. By re-arranging the code a little, I've made the
4207 decoding happen first, so such errors now appear on stderr, and the return
4208 code is 1. However, the actual setting up of the sockets still happens in
4209 the disconnected process, so errors there are still only recorded on the
4212 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4213 connections (as happens on some IP stacks) was logged at start up time as
4214 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4215 it from "IPv6 and IPv4", which means that two separate sockets are being
4218 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4219 says whether AF_INET or AF_INET6 was passed as an argument.
4221 42. Exiwhat output was messed up when time zones were included in log
4224 43. Exiwhat now gives more information about the daemon's listening ports,
4225 and whether -tls-on-connect was used.
4227 44. The "port" option of the smtp transport is now expanded.
4229 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4230 silently ignored. Now an error message is written to the main and panic
4233 46. There's a new ACL modifier called "logwrite" which writes to a log file
4234 as soon as it is encountered.
4236 47. Added $local_user_uid and $local_user_gid at routing time.
4238 48. Exim crashed when trying to verify a sender address that was being
4241 49. Exim was recognizing only a space character after ".include". It now also
4242 recognizes a tab character.
4244 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4245 extracting the relevant information from the specification. The man page no
4246 longer contains scrambled data for the -d option, and I've added a section
4247 at the front about calling Exim under different names.
4249 51. Added "extra_headers" argument to the "mail" command in filter files.
4251 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4254 53. Installed eximstats 1.29.
4256 54. Added transport_filter_timeout as a generic transport option.
4258 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4259 Cc: header lines. This was required by RFC 822, but it not required by RFC
4262 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4263 incoming messages that did not have them. Now it does so only if the
4264 message originates locally, that is, if there is no associated remote host
4265 address. When Resent- header lines are present, this applies to the Resent-
4266 lines rather than the non-Resent- lines.
4268 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4269 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4271 58. Messages for configuration errors now include the name of the main
4272 configuration file - useful now that there may be more than one file in a
4273 list (.included file names were always shown).
4275 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4276 for those rare installations that do not start the daemon as root or run it
4277 setuid root. I've cut out the call to initgroups() if the daemon is not
4280 60. The Exim user and group can now be bound into the binary as text strings
4281 that are looked up at the start of Exim's processing.
4283 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4285 62. Added $mailstore_basename variable.
4287 63. Installed patch to sieve.c from Michael Haardt.
4289 64. When Exim failed to open the panic log after failing to open the main log,
4290 the original message it was trying to log was written to stderr and debug
4291 output, but if they were not available (the usual case in production), it
4292 was lost. Now it is written to syslog before the two lines that record the
4293 failures to open the logs.
4295 65. Users' Exim filters run in subprocesses under the user's uid. It is
4296 possible for a "deliver" command or an alias in a "personal" command to
4297 provoke an address rewrite. If logging of address rewriting is configured,
4298 this fails because the process is not running as root or exim. There may be
4299 a better way of dealing with this, but for the moment (because 4.30 needs
4300 to be released), I have disabled address rewrite logging when running a
4301 filter in a non-root, non-exim process.
4307 1. The buildconfig auxiliary program wasn't quoting the value set for
4308 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4309 not defined. This bug was masked in 4.22 by the effect that was fixed in
4312 2. Some messages that were rejected after a message id was allocated were
4313 shown as "incomplete" by exigrep. It no longer does this for messages that
4314 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4316 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4317 have allow_domain_literals set, the ID did not get logged in the <= line.
4318 Domain literals are now always recognized in Message-ID: header lines.
4320 4. The first argument for a ${extract expansion item is the key name or field
4321 number. Leading and trailing spaces in this item were not being ignored,
4322 causing some misleading effects.
4324 5. When deliver_drop_privilege was set, single queue runner processes started
4325 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4326 same command in the process it spins off) were not dropping privilege.
4328 6. When the daemon running as "exim" started a queue runner, it always
4329 re-executed Exim in the spun-off process. This is a waste of effort when
4330 deliver_drop_privilege is set. The new process now just calls the
4331 queue-runner function directly.
4337 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4340 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4341 ignored. Though the use of -oP was forcing the writing of a pid file, it
4342 was always written to the default place.
4344 3. If the message "no IP address found for host xxxx" is generated during
4345 incoming verification, it is now followed by identification of the incoming
4346 connection (so you can more easily find what provoked it).
4348 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4350 5. Added some features to "harden" Exim a bit more against certain attacks:
4352 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4353 be put in Local/Makefile. This is like the never_users runtime option,
4354 but it cannot be overridden. The default setting is "root".
4356 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4357 prefix string with which any file named in a -C command line option
4360 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4361 is retained for -C and -D only if the caller of Exim is root. Without
4362 it, the exim user may also use -C and -D and retain privilege.
4364 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4365 command line option is disabled.
4367 6. Macro names set by the -D option must start with an upper case letter, just
4368 like macro names defined in the configuration file.
4370 7. Added "dereference=" facility to LDAP.
4372 8. Two instances of the typo "uknown" in the source files are fixed.
4374 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4375 the Configure-Makefile script screwed up while processing it.
4377 10. Incorporated PCRE 4.4.
4379 11. The SMTP synchronization check was not operating right at the start of an
4380 SMTP session. For example, it could not catch a HELO sent before the client
4381 waited for the greeting. There is now a check for outstanding input at the
4382 point when the greeting is written. Because of the duplex, asynchronous
4383 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4384 way, but not yet received, when the check is performed.
4386 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4387 on TCP/IP sockets, because this apparently causes some broken clients to
4390 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4391 unchanged) from the Cygwin maintainer.
4393 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4394 was supported instead of testing for mbx. Effectively a typo.
4396 15. The spa authenticator server code was not checking that the input it
4397 received was valid base64.
4399 16. The debug output line for the "set" modifier in ACLs was not showing the
4400 name of the variable that was being set.
4402 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4404 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4405 lookup on its own. Something else was needed to trigger the lookup. For
4406 example, a match in host_lookup or the need for a host name in a host list.
4407 Now, if $sender_host_name is referenced and the host name has not yet been
4408 looked up, a lookup is performed. If the lookup fails, the variable remains
4409 empty, and $host_lookup_failed is set to "1".
4411 19. Added "eqi" as a case-independent comparison operator.
4413 20. The saslauthd authentication condition could segfault if neither service
4414 nor realm was specified.
4416 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4417 error message that was logged was misleading, and incoming SMTP
4418 connections were dropped. The message is now more accurate, and temporary
4419 errors are given to SMTP connections.
4421 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4422 (except RSET) with a 421 error, until QUIT is received. However, it was
4423 failing to send a response to QUIT.
4425 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4426 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4427 ACL is now run after the helo_try_verify_hosts code.
4429 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4430 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4431 case-independent, so other case variants are also recognized). Apparently
4432 some systems use these upper case variants.
4434 25. If more than two messages were waiting for the same host, and a transport
4435 filter was specified for the transport, Exim sent two messages over the
4436 same TCP/IP connection, and then failed with "socket operation on non-
4437 socket" when it tried to send the third.
4439 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4441 27. The extern definition of crypt16() in expand.c was not being excluded when
4442 the OS had its own crypt16() function.
4444 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4445 as a preferred synonym for return_size_limit, both as an option and as an
4448 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4450 30. Changed the default configuration ACL to relax the local part checking rule
4451 for addresses that are not in any local domains. For these addresses,
4452 slashes and pipe symbols are allowed within local parts, but the sequence
4453 /../ is explicitly forbidden.
4455 31. SPA server authentication was not clearing the challenge buffer before
4458 32. log_message in a "warn" ACL statement was writing to the reject log as
4459 well as to the main log, which contradicts the documentation and doesn't
4460 seem right (because no rejection is happening). So I have stopped it.
4462 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4463 However, I am unable to do any testing of this.
4465 34. Fixed an infelicity in the appendfile transport. When checking directories
4466 for a mailbox, to see if any needed to be created, it was accidentally
4467 using path names with one or more superfluous leading slashes; tracing
4468 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4470 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4471 contents are added to the log line that is written for every discarded
4472 recipient. (Previously a log_message setting was ignored.)
4474 36. The ${quote: operator now quotes the string if it is empty.
4476 37. The install script runs exim in order to find its version number. If for
4477 some reason other than non-existence or emptiness, which it checks, it
4478 could not run './exim', it was installing it with an empty version number,
4479 i.e. as "exim-". This error state is now caught, and the installation is
4482 38. An argument was missing from the function that creates an error message
4483 when Exim fails to connect to the socket for saslauthd authentication.
4484 This could cause Exim to crash, or give a corrupted message.
4486 39. Added isip, isip4, and isip6 to ${if conditions.
4488 40. The ACL variables $acl_xx are now saved with the message, and can be
4489 accessed later in routers, transports, and filters.
4491 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4492 strings in the file are not string-expanded.
4494 42. If a MAIL command specified a SIZE value that was too large to fit into an
4495 int variable, the check against message_size_limit failed. Such values are
4496 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4497 day this will have to be increased, but I don't think I want to be around
4498 when emails are that large.
4505 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4506 iconv() is not standard in FreeBSD.
4508 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4509 IPv6 enabled. The observed symptom was a segmentation fault on return from
4510 the function os_common_find_running_interfaces() in src/os.c.
4512 3. In the check_special_case() function in daemon.c I had used "errno" as an
4513 argument name, which causes warnings on some systems. This was basically a
4514 typo, since it was named "eno" in the comments!
4516 4. The code that waits for the clock to tick (at a resolution of some fraction
4517 of a second) so as to ensure message-id uniqueness was always waiting for
4518 at least one whole tick, when it could have waited for less. [This is
4519 almost certainly not relevant at current processor speeds, where it is
4520 unlikely to ever wait at all. But we try to future-proof.]
4522 5. The function that sleeps for a time interval that includes fractions of a
4523 second contained a race. It did not block SIGALRM between setting the
4524 timer, and suspending (a couple of lines later). If the interval was short
4525 and the sigsuspend() was delayed until after it had expired, the suspension
4526 never ended. On busy systems this could lead to processes getting stuck for
4529 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4530 process, before it forks any delivery processes. The open lookup caching
4531 mechanism meant that the open file or database connection was passed into
4532 the delivery process. The problem was that delivery processes always tidy
4533 up cached lookup data. This could cause a problem for the next delivery
4534 process started by the queue runner, because the external queue runner
4535 process does not know about the closure. So the next delivery process
4536 still has data in the lookup cache. In the case of a file lookup, there was
4537 no problem because closing a file descriptor in a subprocess doesn't affect
4538 the parent. However, if the lookup was caching a connection to a database,
4539 the connection was closed, and the second delivery process was likely to
4540 see errors such as "PGSQL: query failed: server closed the connection
4541 unexpectedly". The problem has been fixed by closing all cached lookups
4542 in a queue runner before running a delivery process.
4544 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4545 seem to have the "const" qualifier which it has on other OS. I've
4548 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4549 *of the same type* (client or server) with the same public name that an
4550 error should be diagnosed.
4552 9. When Exim looked up a host name for an IP address, but failed to find the
4553 original IP address when looking up the host name (a safety check), it
4554 output the message "<ip address> does not match any IP for NULL", which was
4555 confusing, to say the least. The bug was that the host name should have
4556 appeared instead of "NULL".
4558 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4559 user that is built into the binary, and the -C or -D options is used, root
4560 privilege is dropped before the configuration file is read. In addition,
4561 logging is switched to stderr instead of the normal log files. If the
4562 configuration then re-defines the Exim user, the unprivileged environment
4563 is probably not what is expected, so Exim logs a panic warning message (but
4566 However, if deliver_drop_privilege is set, the unprivileged state may well
4567 be exactly what is intended, so the warning has been cut out in that case,
4568 and Exim is allowed to try to write to its normal log files.
4574 1. smtp_return_error_details was not giving details for temporary sender
4575 or receiver verification errors.
4577 2. Diagnose a configuration error if two authenticators have the same public
4580 3. Exim used not to create the message log file for a message until the first
4581 delivery attempt. This could be confusing when incoming messages were held
4582 for policy or load reasons. The message log file is now created at the time
4583 the message is received, and an initial "Received" line is written to it.
4585 4. The automatically generated man page for command line options had a minor
4586 bug that caused no ill effects; however, a more serious problem was that
4587 the procedure for building the man page automatically didn't always
4588 operate. Consequently, release 4.20 contains an out-of-date version. This
4589 shouldn't happen again.
4591 5. When building Exim with embedded Perl support, the script that builds the
4592 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4593 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4595 6. The freeze_tell option was not being used for messages that were frozen on
4596 arrival, either by an ACL or by local_scan().
4598 7. Added the smtp_incomplete_transaction log selector.
4600 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4601 was accepting AUTH without a new EHLO.
4603 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4604 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4607 10. Logging of TCP/IP connections (when configured) now happens in the main
4608 daemon process instead of the child process, so that the TCP/IP connection
4609 count is more accurate (but it can never be perfect).
4611 11. The use of "drop" in a nested ACL was not being handled correctly in the
4612 outer ACL. Now, if condition failure induced by the nested "drop" causes
4613 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4614 or "require"), the connection is dropped.
4616 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4617 that yield "discard" can now be used with an "accept" or a "discard" verb,
4618 but an error is generated for any others (because I can't see a useful way
4619 to define what should happen).
4621 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4622 are now processed in the same way as lines in the Exim configuration file.
4623 In particular, continuation lines are supported.
4625 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4627 15. Added -ti meaning -t -i.
4629 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4630 domains, and warn by logging if others are found.
4632 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4633 returned from GIFCONF when Exim is trying to find the list of interfaces on
4634 a host. The code in os.c has been modified to copy each ifreq to an aligned
4635 structure in all cases.
4637 Also, in some cases, the returned ifreq's were being copied to a 'struct
4638 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4639 means the last couple of bytes of an IPv6 address could be chopped if the
4640 ifreq contained only a normal sockaddr (14 bytes storage).
4642 18. Named domain lists were not supported in the hosts_treat_as_local option.
4643 An entry such as +xxxx was not recognized, and was treated as a literal
4646 19. Ensure that header lines added by a DATA ACL are included in the reject log
4647 if the ACL subsequently rejects the message.
4649 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4650 MD5 (which is deprecated).
4652 21. When testing a filter file using -bf, Exim was writing a message when it
4653 took the sender from a "From " line in the message, but it was not doing so
4654 when it took $return_path from a Return-Path: header line. It now does.
4656 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4657 with a valid header line field name (a series of printing characters
4658 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4660 23. Changed "disc" in the source to "disk" to conform to the documentation and
4661 the book and for uniformity.
4663 24. Ignore Sendmail's -Ooption=value command line item.
4665 25. When execve() failed while trying to run a command in a pipe transport,
4666 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4667 could be confused with a return value of 69 from the command itself. This
4668 has been changed to 127, the value the shell returns if it is asked to run
4669 a non-existent command. The wording for the related log line suggests a
4670 non-existent command as the problem.
4672 26. If received_header_text expands to an empty string, do not add a Received:
4673 header line to the message. (Well, it adds a token one on the spool, but
4674 marks it "old" so that it doesn't get used or transmitted.)
4676 27. Installed eximstats 1.28 (addition of -nt option).
4678 28. There was no check for failure on the call to getsockname() in the daemon
4679 code. This can fail if there is a shortage of resources on the system, with
4680 ENOMEM, for example. A temporary error is now given on failure.
4682 29. Contrary to the C standard, it seems that in some environments, the
4683 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4684 program. Exim now does this explicitly; it affects the formatting of
4685 timestamps using strftime().
4687 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4688 complaints. I've now initialized all the variables, to avoid this.
4690 32. Header lines added by a system filter were not being "seen" during
4691 transport-time rewrites.
4693 33. The info_callback() function passed to OpenSSL is set up with type void
4694 (*)(SSL *, int, int), as described somewhere. However, when calling the
4695 function (actually a macro) that sets it up, the type void(*)() is
4696 expected. I've put in a cast to prevent warnings from picky compilers.
4698 34. If a DNS black list lookup found a CNAME record, but there were no A
4699 records associated with the domain it pointed at, Exim crashed.
4701 35. If a DNS black list lookup returned more than one A record, Exim ignored
4702 all but the first. It now scans all returned addresses if a particular IP
4703 value is being sought. In this situation, the contents of the
4704 $dnslist_value variable are a list of all the addresses, separated by a
4707 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4708 to accept a host name and all its aliases if the forward lookup for any of
4709 them yielded the IP address of the incoming connection. Now it accepts only
4710 those names whose forward lookup yields the correct IP address. Any other
4711 names are discarded. This closes a loophole whereby a rogue DNS
4712 administrator could create reverse DNS records to break through a
4713 wildcarded host restriction in an ACL.
4715 37. If a user filter or a system filter that ran in a subprocess used any of
4716 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4717 the wrong values were passed to the pipe command ($thisaddress had the
4718 value of $0, $0 had the value of $1, etc). This bug was introduced by
4719 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4721 38. Improved the line breaking for long SMTP error messages from ACLs.
4722 Previously, if there was no break point between 40 and 75 characters, Exim
4723 left the rest of the message alone. Two changes have been made: (a) I've
4724 reduced the minimum length to 35 characters; (b) if it can't find a break
4725 point between 35 and 75 characters, it looks ahead and uses the first one
4726 that it finds. This may give the occasional overlong line, but at least the
4727 remaining text gets split now.
4729 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4730 file descriptors might be low, and that setting 1000 would always raise it.
4731 It turns out that in some environments, the limit is already over 1000 and
4732 that lowering it causes trouble. So now Exim takes care not to decrease it.
4734 40. When delivering a message, the value of $return_path is set to $sender_
4735 address at the start of routing (routers may change the value). By an
4736 oversight, this default was not being set up when an address was tested by
4737 -bt or -bv, which affected the outcome if any router or filter referred to
4740 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4741 log only when "message" or "log_message" are set. However, if one of the
4742 conditions was an address verification, or a call to a nested ACL, the
4743 messages generated by the underlying test were being passed through. This
4744 no longer happens. The underlying message is available in $acl_verify_
4745 message for both "message" and "log_message" expansions, so it can be
4746 passed through if needed.
4748 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4749 new expansion $bh_ to give the encoded byte string without charset
4750 translation. Translation happens only if iconv() is available; HAVE_ICONV
4751 indicates this at build time. HEADERS_CHARSET gives the charset to
4752 translate to; headers_charset can change it in the configuration, and
4753 "headers charset" can change it in an individual filter file.
4755 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4756 that creates RFC 2047 encoded "words" labels them as that charset instead
4757 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4758 expansion operator; (ii) when Exim creates a From: line for a local
4759 message; (iii) when a header line is rewritten to include a "phrase" part.
4761 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4762 buggy, causing it to skip the first lines of messages whose message ID
4763 ended in 'D'. This would not have bitten before Exim release 4.14, because
4764 message IDs were unlikely to end in 'D' before then. The effect was to have
4765 incorrect size information for certain domains.
4767 45. #include "config.h" was missing at the start of the crypt16.c module. This
4768 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4771 46. If there was a timeout during a "random" callout check, Exim treated it as
4772 a failure of the random address, and carried on sending RSET and the real
4773 address. If the delay was just some slowness somewhere, the response to the
4774 original RCPT would be taken as a response to RSET and so on, causing
4775 mayhem of various kinds.
4777 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4778 when I implemented it. It didn't allow for the fact that some option values
4779 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4780 the right test for positive values.
4782 48. Domain names in DNS records are case-independent. Exim always looks them up
4783 in lower case. Some resolvers return domain names in exactly the case they
4784 appear in the zone file, that is, they may contain uppercase letters. Not
4785 all resolvers do this - some return always lower case. Exim was treating a
4786 change of case by a resolver as a change of domain, similar to a widening
4787 of a domain abbreviation. This triggered its re-routing code and so it was
4788 trying to route what was effectively the same domain again. This normally
4789 caused routing to fail (because the router wouldn't handle the domain
4790 twice). Now Exim checks for this case specially, and just changes the
4791 casing of the domain that it ultimately uses when it transmits the message
4794 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4797 50. If a filter generated a file delivery with a non-absolute name (possible if
4798 no home directory exists for the router), the forbid_file option was not
4801 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4802 the existing equality matching.
4804 52. Exim was using ints instead of ino_t variables in some places where it was
4805 dealing with inode numbers.
4807 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4808 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4809 called TMPDIR, and if it finds it is different, it changes its value.
4811 54. The smtp_printf() function is now made available to local_scan() so
4812 additional output lines can be written before returning. There is also an
4813 smtp_fflush() function to enable the detection of a dropped connection.
4814 The variables smtp_input and smtp_batched_input are exported to
4817 55. Changed the default runtime configuration: the message "Unknown user"
4818 has been removed from the ACL, and instead placed on the localuser router,
4819 using the cannot_route_message feature. This means that any verification
4820 failures that generate their own messages won't get overridden. Similarly,
4821 the "Unrouteable address" message that was in the ACL for unverifiable
4822 relay addresses has also been removed.
4824 56. Added hosts_avoid_esmtp to the smtp transport.
4826 57. The exicyclog script was not checking for the esoteric option
4827 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4828 will work only if exicyclog is run under the appropriate euid.
4830 58. Following a discussion on the list, the rules by which Exim recognises line
4831 endings on incoming messages have been changed. The -dropcr and drop_cr
4832 options are now no-ops, retained only for backwards compatibility. The
4833 following line terminators are recognized: LF CRLF CR. However, special
4834 processing applies to CR:
4836 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4837 nor a local message in the state where . is a terminator.
4839 (ii) If a bare CR is encountered in a header line, an extra space is added
4840 after the line terminator so as not to end the header. The reasoning
4841 behind this is that bare CRs in header lines are most likely either
4842 to be mistakes, or people trying to play silly games.
4844 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4845 was being incorrectly given as 18 bytes larger than it should have been.
4846 This is a VOB (very old bug).
4848 60. This may never have affected anything current, but just in case it has:
4849 When the local host is found other than at the start of a list of hosts,
4850 the local host, those with the same MX, and any that follow, are discarded.
4851 When the list in question was part of a longer list of hosts, the following
4852 hosts (not currently being processed) were also being discarded. This no
4853 longer happens. I'm not sure if this situation could ever has previously
4856 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4859 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4860 header. This is something that is recommended in a new Internet Draft, and
4861 is something that is documented as being done by Sendmail. There are two
4862 possible values. For messages generated by the autoreply transport, Exim
4865 Auto-Submitted: auto-replied
4867 whereas for all other generated messages (e.g. bounces) it adds
4869 Auto-Submitted: auto-generated
4871 63. The "personal" condition in filters now includes a test for the
4872 Auto-Submitted: header. If it contains the string "auto-" the message it
4873 not considered personal.
4875 64. Added rcpt_include_affixes as a generic transport option.
4877 65. Added queue_only_override (default true).
4879 66. Added the syslog_duplication option.
4881 67. If what should have been the first header line of a message consisted of
4882 a space followed by a colon, Exim was mis-interpreting it as a header line.
4883 It isn't of course - it is syntactically invalid and should therefore be
4884 treated as the start of the message body. The misbehaviour could have
4885 caused a number of strange effects, including loss of data in subsequent
4886 header lines, and spool format errors.
4888 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4889 client host had authenticated. This control can now be exercised by an ACL
4890 for more flexibility.
4892 69. By default, callouts do not happen when testing with -bh. There is now a
4893 variant, -bhc, which does actually run the callout code, including
4894 consulting and updating the callout cache.
4896 70. Added support for saslauthd authentication, courtesy of Alexander
4899 71. If statvfs() failed on the spool or log directories while checking their
4900 size for availability, Exim confusingly gave the error "space shortage".
4901 Furthermore, in debugging mode it crashed with a floating point exception.
4902 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4903 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4904 serious problem, Exim now writes to the main and panic logs when this
4905 happens, with details of the failure. It then refuses to accept the
4906 incoming message, giving the message "spool directory problem" or "log
4907 directory problem" with a 421 code for SMTP messages.
4909 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4910 0, 1, and 2 exist, because some OS complain for execs without them (see
4911 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4912 descriptors. However, the code omitted to check that the open succeeded,
4913 causing mysterious errors if for some reason the permissions on /dev/null
4914 got screwed. Now Exim writes a message to the main and panic logs, and
4915 bombs out if it can't open /dev/null.
4917 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4918 interact so that it is all more flexible. It is supposed to remain
4919 backwards compatible. Also added extra_local_interfaces.
4921 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4922 to bomb out with an assertion failure - to the client this appears as a
4923 connection drop. This problem occurs in the part of the code that was taken
4924 from the Samba project. Fortunately, the assertion is in a very simple
4925 function, so I have fixed this by reproducing the function inline in the
4926 one place where it is called, and arranging for authentication to fail
4927 instead of killing the process with assert().
4929 75. The SPA client code was not working when the server requested OEM rather
4930 than Unicode encoding.
4932 76. Added code to make require_files with a specific uid setting more usable in
4933 the case where statting the file as root fails - usually a non-root-mounted
4934 NFS file system. When this happens and the failure is EACCES, Exim now
4935 forks a subprocess and does the per-uid checking as the relevant uid.
4937 77. Added process_log_path.
4939 78. If log_file_path was not explicitly set, a setting of check_log_space or
4940 check_log_inodes was ignored.
4942 79. If a space check for the spool or log partitions fails, the incident is now
4943 logged. Of course, in the latter case the data may get lost...
4945 80. Added the %p formatting code to string_format() so that it can be used to
4946 print addresses in debug_print(). Adjusted all the address printing in the
4947 debugging in store.c to use %p rather than %d.
4949 81. There was a concern that a line of code in smtp_in.c could overflow a
4950 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4951 initially expressed, the concern was not well-founded, because trailing
4952 spaces are removed early. However, if the trailing spaces were followed by
4953 a NULL, they did not get removed, so the overflow was possible. Two fixes
4956 (a) I re-wrote the offending code in a cleaner fashion.
4957 (b) If an incoming SMTP command contains a NULL character, it is rejected
4960 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4961 runs initgroups(), so that if the Exim user is in any additional groups,
4962 they will be used during message reception.
4968 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob