1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.554 2008/09/29 11:41:07 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
78 TK/01 Add preliminary DKIM support. Currently requires a forked version of
79 ALT-N's libdkim that I have put here:
80 http://duncanthrax.net/exim-experimental/
82 Note to Michael Haardt: I had to rename some vars in sieve.c. They
83 were called 'true' and it seems that C99 defines that as a reserved
84 keyword to be used with 'bool' variable types. That means you could
85 not include C99-style headers which use bools without triggering
86 build errors in sieve.c.
88 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
89 as mailq or other aliases. Changed the --help handling significantly
90 to do whats expected. exim_usage() emits usage/help information.
92 SC/01 Added the -bylocaldomain option to eximstats.
94 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
96 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
98 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
104 PH/01 Another patch from the Sieve maintainer.
106 PH/02 When an IPv6 address is converted to a string for single-key lookup
107 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
108 dots are used instead of colons so that keys in lsearch files need not
109 contain colons. This was done some time before quoting was made available
110 in lsearch files. However, iplsearch files do require colons in IPv6 keys
111 (notated using the quote facility) so as to distinguish them from IPv4
112 keys. This meant that lookups for IP addresses in host lists did not work
113 for iplsearch lookups.
115 This has been fixed by arranging for IPv6 addresses to be expressed with
116 colons if the lookup type is iplsearch. This is not incompatible, because
117 previously such lookups could never work.
119 The situation is now rather anomolous, since one *can* have colons in
120 ordinary lsearch keys. However, making the change in all cases is
121 incompatible and would probably break a number of configurations.
123 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
126 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
127 conversion specification without a maximum field width, thereby enabling
128 a rogue spamd server to cause a buffer overflow. While nobody in their
129 right mind would setup Exim to query an untrusted spamd server, an
130 attacker that gains access to a server running spamd could potentially
131 exploit this vulnerability to run arbitrary code as the Exim user.
133 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
134 $primary_hostname instead of what libspf2 thinks the hosts name is.
136 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
137 a directory entry by the name of the lookup key. Previously, if a
138 symlink pointed to a non-existing file or a file in a directory that
139 Exim lacked permissions to read, a lookup for a key matching that
140 symlink would fail. Now it is enough that a matching directory entry
141 exists, symlink or not. (Bugzilla 503.)
143 PH/03 The body_linecount and body_zerocount variables are now exported in the
146 PH/04 Added the $dnslist_matched variable.
148 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
149 This means they are set thereafter only if the connection becomes
152 PH/06 Added the client_condition to authenticators so that some can be skipped
153 by clients under certain conditions.
155 PH/07 The error message for a badly-placed control=no_multiline_responses left
156 "_responses" off the end of the name.
158 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
160 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
161 (without spaces) instead of just copying the configuration text.
163 PH/10 Added the /noupdate option to the ratelimit ACL condition.
165 PH/11 Added $max_received_linelength.
167 PH/12 Added +ignore_defer and +include_defer to host lists.
169 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
170 way in which PCRE > 7.0 is built.
172 PH/14 Implemented queue_only_load_latch.
174 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
175 MAIL command. The effect was to mangle the value on 64-bit systems.
177 PH/16 Another patch from the Sieve maintainer.
179 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
181 PH/18 If a system quota error occurred while trying to create the file for
182 a maildir delivery, the message "Mailbox is full" was not appended to the
183 bounce if the delivery eventually timed out. Change 4.67/27 below applied
184 only to a quota excession during the actual writing of the file.
186 PH/19 It seems that peer DN values may contain newlines (and other non-printing
187 characters?) which causes problems in log lines. The DN values are now
188 passed through string_printing() before being added to log lines.
190 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
191 and InterBase are left for another time.)
193 PH/21 Added message_body_newlines option.
195 PH/22 Guard against possible overflow in moan_check_errorcopy().
197 PH/23 POSIX allows open() to be a macro; guard against that.
199 PH/24 If the recipient of an error message contained an @ in the local part
200 (suitably quoted, of course), incorrect values were put in $domain and
201 $local_part during the evaluation of errors_copy.
207 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
208 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
211 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
212 issue a MAIL command.
214 PH/02 In an ACL statement such as
216 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
218 if a client was not listed at all, or was listed with a value other than
219 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
220 the condition was not true (as it should be), so access was not denied.
221 The bug was that the ! inversion was incorrectly passed on to the second
222 item. This has been fixed.
224 PH/03 Added additional dnslists conditions == and =& which are different from
225 = and & when the dns lookup returns more than one IP address.
227 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
228 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
230 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
231 FSYNC, which compiles an option called disable_fsync that allows for
232 bypassing fsync(). The documentation is heavily laced with warnings.
234 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
236 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
237 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
238 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
239 including adding "make clean"; (3) Added -fPIC when compiling the test
240 dynamically loaded module, to get rid of a warning.
242 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
243 message fails, move_frozen_messages = true and ignore_bounce_errors_after
244 = 0s. The bug is otherwise harmless.
246 PH/07 There was a bug in the dovecot authenticator such that the value of
247 $auth1 could be overwritten, and so not correctly preserved, after a
248 successful authentication. This usually meant that the value preserved by
249 the server_setid option was incorrect.
251 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
253 PH/09 Installed PCRE release 7.0.
255 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
256 run for batched SMTP input. It is now run at the start of every message
257 in the batch. While fixing this I discovered that the process information
258 (output by running exiwhat) was not always getting set for -bs and -bS
259 input. This is fixed, and it now also says "batched" for BSMTP.
261 PH/11 Added control=no_pipelining.
263 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
264 patch, slightly modified), and move the expansion of helo_data till after
265 the connection is made in the smtp transport (so it can use these
268 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
270 PH/14 Added log_selector = +pid.
272 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
274 PH/16 Add ${if forany and ${if forall.
276 PH/17 Added dsn_from option to vary the From: line in DSNs.
278 PH/18 Flush SMTP output before performing a callout, unless control =
279 no_callout_flush is set.
281 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
282 was true (the default) a successful delivery failed to delete the retry
283 item, thus causing premature timeout of the address. The bug is now
286 PH/20 Added hosts_avoid_pipelining to the smtp transport.
288 PH/21 Long custom messages for fakedefer and fakereject are now split up
289 into multiline reponses in the same way that messages for "deny" and
290 other ACL rejections are.
292 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
293 with slight modification.
295 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
296 draft, changing the syntax and factoring some duplicate code".
298 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
299 for deliveries of the second and subsequent messages over the same SMTP
302 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
303 ${reduce, with only minor "tidies".
305 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
307 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
308 expansion side effects.
310 PH/27 When a message times out after an over-quota error from an Exim-imposed
311 quota, the bounce message says "mailbox is full". This message was not
312 being given when it was a system quota that was exceeded. It now should
315 MH/03 Made $recipients available in local_scan(). local_scan() already has
316 better access to the recipient list through recipients_list[], but
317 $recipients can be useful in postmaster-provided expansion strings.
319 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
320 in the case of a MAIL command with additional options following the
321 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
322 were accidentally chopped off.
324 PH/29 SMTP synchronization checks are implemented when a command is read -
325 there is a check that no more input is waiting when there shouldn't be
326 any. However, for some commands, a delay in an ACL can mean that it is
327 some time before the response is written. In this time, more input might
328 arrive, invalidly. So now there are extra checks after an ACL has run for
329 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
330 pipelining has not been advertised.
332 PH/30 MH's patch to allow iscntrl() characters to be list separators.
334 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
335 returned in the SMTP response when smtp_return_error_details was false.
338 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
339 instead of the C I/O that was originally supplied, because problems were
342 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
343 Exim which did not show up earlier: it was assuming that a call to
344 SSL_CTX_set_info_callback() might give an error value. In fact, there is
345 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
346 was a macro that became an assignment, so it seemed to work. This has
347 changed to a proper function call with a void return, hence the compile
348 error. Exim's code has been fixed.
350 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
353 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
355 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
357 PH/37 If a message is not accepted after it has had an id assigned (e.g.
358 because it turns out to be too big or there is a timeout) there is no
359 "Completed" line in the log. When some messages of this type were
360 selected by exigrep, they were listed as "not completed". Others were
361 picked up by some special patterns. I have improved the selection
362 criteria to be more general.
364 PH/38 The host_find_failed option in the manualroute router can now be set
365 to "ignore", to completely ignore a host whose IP address cannot be
366 found. If all hosts are ignored, the behaviour is controlled by the new
367 host_all_ignored option.
369 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
370 homing or because of multiple MX records with /mx) generated more than
371 one IP address, and the following item turned out to be the local host,
372 all the secondary addresses of the first item were incorrectly removed
373 from the list, along with the local host and any following hosts (which
374 is what is supposed to happen).
376 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
377 whoever called Exim into the -H file. In the case of the daemon it was
378 behaving confusingly. When first started, it used values for whoever
379 started the daemon, but after a SIGHUP it used the Exim user (because it
380 calls itself on a restart). I have changed the code so that it now always
383 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
384 message are rejected with the same error (e.g. no authentication or bad
385 sender address), and a DATA command is nevertheless sent (as can happen
386 with PIPELINING or a stupid MUA), the error message that was given to the
387 RCPT commands is included in the rejection of the DATA command. This is
388 intended to be helpful for MUAs that show only the final error to their
391 PH/42 Another patch from the Sieve maintainer.
393 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
394 Eximstats - Fixed some broken HTML links and added missing column headers
396 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
397 columns for Rejects, Temp Rejects, Ham, and Spam rows.
399 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
401 PH/43 Yet another patch from the Sieve maintainer.
403 PH/44 I found a way to check for a TCP/IP connection going away before sending
404 the response to the final '.' that terminates a message, but only in the
405 case where the client has not sent further data following the '.'
406 (unfortunately, this is allowed). However, in many cases there won't be
407 any further data because there won't be any more messages to send. A call
408 to select() can be used: if it shows that the input is "ready", there is
409 either input waiting, or the socket has been closed. An attempt to read
410 the next input character can distinguish the two cases. Previously, Exim
411 would have sent an OK response which the client would never have see.
412 This could lead to message repetition. This fix should cure that, at
413 least in a lot of common cases.
415 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
416 advertised in response to EHLO.
422 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
423 fixed by 4.65/MH/01 (is this a record?) are fixed:
425 (i) An empty string was always treated as zero by the numeric comparison
426 operators. This behaviour has been restored.
428 (ii) It is documented that the numeric comparison operators always treat
429 their arguments as decimal numbers. This was broken in that numbers
430 starting with 0 were being interpreted as octal.
432 While fixing these problems I realized that there was another issue that
433 hadn't been noticed. Values of message_size_limit (both the global option
434 and the transport option) were treated as octal if they started with 0.
435 The documentation was vague. These values are now always treated as
436 decimal, and I will make that clear in the documentation.
442 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
443 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
446 MH/01 Don't check that the operands of numeric comparison operators are
447 integers when their expansion is in "skipping" mode (fixes bug
448 introduced by 4.64-PH/07).
450 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
451 child addresses, Exim now panics and dies. Previously, because the count
452 is held in a short int, deliveries were likely to be lost. As such a
453 large number of recipients for a single message is ridiculous
454 (performance will be very, very poor), I have chosen to impose a limit
455 rather than extend the field.
461 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
462 leftover -K file (the existence of which was triggered by #402).
463 While we were at it, introduced process PID as part of the -K
464 filename. This should rule out race conditions when creating
467 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
468 processing considerably. Previous code took too long for large mails,
469 triggering a timeout which in turn triggers #401.
471 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
472 in the DK code in transports.c. sendfile() is not really portable,
473 hence the _LINUX specificness.
475 TF/01 In the add_headers option to the mail command in an Exim filter,
476 there was a bug that Exim would claim a syntax error in any
477 header after the first one which had an odd number of characters
480 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
481 callout verification, Exim cached a "reject" for the entire domain. This
482 is correct for most verifications, but it is not correct for a recipient
483 verification with use_sender or use_postmaster set, because in that case
484 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
485 case of MAIL FROM:<> rejection from other early rejections (e.g.
486 rejection of HELO). When verifying a recipient using a non-null MAIL
487 address, the cache is ignored if it shows MAIL FROM:<> rejection.
488 Whatever the result of the callout, the value of the domain cache is
489 left unchanged (for any other kind of callout, getting as far as trying
490 RCPT means that the domain itself is ok).
492 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
495 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
496 manifest itself as EPIPE rather than ECONNECT. When tidying away a
497 session, the daemon ignores ECONNECT errors and logs others; it now
498 ignores EPIPE as well.
500 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
501 (quoted-printable decoding).
503 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
504 later the small subsequent patch to fix an introduced bug.
506 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
508 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
510 PH/08 An error is now given if message_size_limit is specified negative.
512 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
513 to be given (somewhat) arbitrary names.
515 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
518 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
519 miscellaneous code fixes
521 PH/10 Added the log_reject_target ACL modifier to specify where to log
524 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
525 hostname. This is wrong, because it relates to the incoming message (and
526 probably the interface on which it is arriving) and not to the outgoing
527 callout (which could be using a different interface). This has been
528 changed to use the value of the helo_data option from the smtp transport
529 instead - this is what is used when a message is actually being sent. If
530 there is no remote transport (possible with a router that sets up host
531 addresses), $smtp_active_hostname is used.
533 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
534 tweaks were necessary in order to get it to work (see also 21 below):
535 (a) The code assumed that strncpy() returns a negative number on buffer
536 overflow, which isn't the case. Replaced with Exim's string_format()
538 (b) There were several signed/unsigned issues. I just did the minimum
539 hacking in of casts. There is scope for a larger refactoring.
540 (c) The code used strcasecmp() which is not a standard C function.
541 Replaced with Exim's strcmpic() function.
542 (d) The code set only $1; it now sets $auth1 as well.
543 (e) A simple test gave the error "authentication client didn't specify
544 service in request". It would seem that Dovecot has changed its
545 interface. Fortunately there's a specification; I followed it and
546 changed what the client sends and it appears to be working now.
548 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
551 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
552 address A is aliased to B and C, where B exists and C does not. Without
553 -v the output is "A verified" because verification stops after a
554 successful redirection if more than one address is generated. However,
555 with -v the child addresses are also verified. Exim was outputting "A
556 failed to verify" and then showing the successful verification for C,
557 with its parentage. It now outputs "B failed to verify", showing B's
558 parentage before showing the successful verification of C.
560 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
561 look up a TXT record in a specific list after matching in a combined
564 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
565 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
566 they consult the DNS. I had assumed they would set it the way they
567 wanted; and indeed my experiments on Linux seem to show that in some
568 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
569 To be on the safe side, however, I have now made the interface to
570 host_find_byname() similar to host_find_bydns(), with an argument
571 containing the DNS resolver options. The host_find_byname() function now
572 sets these options at its start, just as host_find_bydns() does. The smtp
573 transport options dns_qualify_single and dns_search_parents are passed to
574 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
575 of host_find_byname() use the default settings of RES_DEFNAMES
576 (qualify_single) but not RES_DNSRCH (search_parents).
578 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
579 spool_read_header() do less string testing, by means of a preliminary
580 switch on the second character of optional "-foo" lines. (This is
581 overdue, caused by the large number of possibilities that now exist.
582 Originally there were few.) While I was there, I also converted the
583 str(n)cmp tests so they don't re-test the leading "-" and the first
584 character, in the hope this might squeeze out yet more improvement.
586 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
587 flag allowing group syntax was set by the header_syntax check but not
588 turned off, possible causing trouble later; (2) The flag was not being
589 set at all for the header_verify test, causing "group"-style headers to
590 be rejected. I have now set it in this case, and also caused header_
591 verify to ignore an empty address taken from a group. While doing this, I
592 came across some other cases where the code for allowing group syntax
593 while scanning a header line wasn't quite right (mostly, not resetting
594 the flag correctly in the right place). These bugs could have caused
595 trouble for malformed header lines. I hope it is now all correct.
597 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
598 with the "reply" argument non-NULL. The code, however (which originally
599 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
600 but it didn't always do it. This confused somebody who was copying the
601 code for some other use. I have removed all the tests.
603 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
604 feature that was used to support insecure browsers during the U.S. crypto
605 embargo. It requires special client support, and Exim is probably the
606 only MTA that supported it -- and would never use it because real RSA is
607 always available. This code has been removed, because it had the bad
608 effect of slowing Exim down by computing (never used) parameters for the
609 RSA_EXPORT functionality.
611 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
612 authenticator to fail if there's a tab character in the incoming data
613 (there should never be unless someone is messing about, as it's supposed
614 to be base64-encoded). Also added, on Timo's advice, the "secured" option
615 if the connection is using TLS or if the remote IP is the same as the
616 local IP, and the "valid-client-cert option" if a client certificate has
619 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
620 authenticators. This can be used for authorization after authentication
621 succeeds. (In the case of plaintext, it servers for both authentication
624 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
625 if any retry times were supplied.
627 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
628 connection, where there is no client IP address to check. In this
629 situation, the verify now always succeeds.
631 PH/25 Applied John Jetmore's -Mset patch.
633 PH/26 Added -bem to be like -Mset, but loading a message from a file.
635 PH/27 In a string expansion for a processed (not raw) header when multiple
636 headers of the same name were present, leading whitespace was being
637 removed from all of them, but trailing whitespace was being removed only
638 from the last one. Now trailing whitespace is removed from each header
639 before concatenation. Completely empty headers in a concatenation (as
642 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
643 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
645 PH/29 [Removed. This was a change that I later backed out, and forgot to
646 correct the ChangeLog entry (that I had efficiently created) before
647 committing the later change.]
649 PH/30 Exim was sometimes attempting to deliver messages that had suffered
650 address errors (4xx response to RCPT) over the same connection as other
651 messages routed to the same hosts. Such deliveries are always "forced",
652 so retry times are not inspected. This resulted in far too many retries
653 for the affected addresses. The effect occurred only when there were more
654 hosts than the hosts_max_try setting in the smtp transport when it had
655 the 4xx errors. Those hosts that it had tried were not added to the list
656 of hosts for which the message was waiting, so if all were tried, there
657 was no problem. Two fixes have been applied:
659 (i) If there are any address or message errors in an SMTP delivery, none
660 of the hosts (tried or untried) are now added to the list of hosts
661 for which the message is waiting, so the message should not be a
662 candidate for sending over the same connection that was used for a
663 successful delivery of some other message. This seems entirely
664 reasonable: after all the message is NOT "waiting for some host".
665 This is so "obvious" that I'm not sure why it wasn't done
666 previously. Hope I haven't missed anything, but it can't do any
667 harm, as the worst effect is to miss an optimization.
669 (ii) If, despite (i), such a delivery is accidentally attempted, the
670 routing retry time is respected, so at least it doesn't keep
671 hammering the server.
673 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
674 in ${readsocket because some servers need this prod.
676 PH/32 Added some extra debug output when updating a wait-xxx database.
678 PH/33 The hint "could be header name not terminated by colon", which has been
679 given for certain expansion errors for a long time, was not being given
680 for the ${if def:h_colon_omitted{... case.
682 PH/34 The spec says: "With one important exception, whenever a domain list is
683 being scanned, $domain contains the subject domain." There was at least
684 one case where this was not true.
686 PH/35 The error "getsockname() failed: connection reset by peer" was being
687 written to the panic log as well as the main log, but it isn't really
688 panic-worthy as it just means the connection died rather early on. I have
689 removed the panic log writing for the ECONNRESET error when getsockname()
692 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
693 runs only) independently of the message's sender address. This meant
694 that, if the 4xx error was in fact related to the sender, a different
695 message to the same recipient with a different sender could confuse
696 things. In particualar, this can happen when sending to a greylisting
697 server, but other circumstances could also provoke similar problems.
698 I have changed the default so that the retry time for these errors is now
699 based a combination of the sender and recipient addresses. This change
700 can be overridden by setting address_retry_include_sender=false in the
703 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
704 remote server are returned as part of bounce messages. This was not
705 happening for LMTP over a pipe (the lmtp transport), but now it is the
706 same for both kinds of LMTP.
708 PH/38 Despite being documented as not happening, Exim was rewriting addresses
709 in header lines that were in fact CNAMEs. This is no longer the case.
711 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
712 and queue runs started by the daemon processed all messages. This has
713 been fixed so that -R and -S can now usefully be given with -q<time>.
715 PH/40 Import PCRE release 6.7 (fixes some bugs).
717 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
719 PH/42 Give an error if -q is specified more than once.
721 PH/43 Renamed the variables $interface_address and $interface_port as
722 $received_ip_address and $received_port, to make it clear that these
723 values apply to message reception, and not to the outgoing interface when
724 a message is delivered. (The old names remain recognized, of course.)
726 PH/44 There was no timeout on the connect() call when using a Unix domain
727 socket in the ${readsocket expansion. There now is.
729 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
730 be meaningful with "accept".
732 SC/01 Eximstats V1.43
733 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
735 SC/02 Eximstats V1.44
736 Use a glob alias rather than an array ref in the generated
737 parser. This improves both readability and performance.
739 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
740 Collect SpamAssassin and rejection statistics.
741 Don't display local sender or destination tables unless
742 there is data to show.
743 Added average volumes into the top table text output.
745 SC/04 Eximstats V1.46
746 Collect data on the number of addresses (recipients)
747 as well as the number of messages.
749 SC/05 Eximstats V1.47
750 Added 'Message too big' to the list of mail rejection
751 reasons (thanks to Marco Gaiarin).
753 SC/06 Eximstats V1.48
754 Mainlog lines which have GMT offsets and are too short to
755 have a flag are now skipped.
757 SC/07 Eximstats V1.49 (Alain Williams)
758 Added the -emptyok flag.
760 SC/08 Eximstats V1.50
761 Fixes for obtaining the IP address from reject messages.
763 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
764 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
765 whitesspace changes from 4.64-PH/27
767 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
770 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
771 are found, allow negative numbers in numeric criteria)
773 JJ/06 exipick.20061117.2, added new $message_body_missing variable
775 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
776 to match changes made in 4.64-PH/43
778 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
780 PH/47 Put in an explicit test for a DNS lookup of an address record where the
781 "domain" is actually an IP address, and force a failure. This locks out
782 those revolvers/nameservers that support "A-for-A" lookups, in
783 contravention of the specifications.
785 PH/48 When a host name was looked up from an IP address, and the subsequent
786 forward lookup of the name timed out, the host name was left in
787 $sender_host_name, contrary to the specification.
789 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
790 restricted to single-key lookups, Exim was not diagnosing an error if
791 * or *@ was used with a query-style lookup.
793 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
795 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
796 long ago, but noone interested enough thought of it. Let's just say that
797 the "1.1" means that there are some new functions that weren't there at
798 some point in the past.
800 PH/51 Error processing for expansion failure of helo_data from an smtp
801 transport during callout processing was broken.
803 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
804 tested/used via the -bh/-bhc/-bs options.
806 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
807 bug, fixed in subsequent PCRE releases).
809 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
810 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
812 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
818 SC/01 Use a glob alias rather than an array ref in eximstats generated
819 parser. This improves both readability and performance.
821 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
822 Don't display local sender or destination tables in eximstats unless
823 there is data to show.
824 Added average volumes into the eximstats top table text output.
826 SC/03 Collect data on the number of addresses (recipients) as well
827 as the number of messages in eximstats.
829 TF/01 Correct an error in the documentation for the redirect router. Exim
830 does (usually) call initgroups() when daemonizing.
832 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
833 with consistent privilege compared to when running as a daemon.
835 TF/03 Note in the spec that $authenticated_id is not set for local
836 submissions from trusted users.
838 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
839 Thanks to Dean Brooks <dean@iglou.com> for the patch.
841 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
842 by adding some example configuration directives to the default
843 configuration file. A little bit of work is required to uncomment the
844 directives and define how usernames and passwords are checked, but
845 there is now a framework to start from.
847 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
848 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
849 without this. I don't know how relevant this is to other LDAP libraries.
851 PH/02 Add the verb name to the "unknown ACL verb" error.
853 PH/03 Magnus Holmgren's patch for filter_prepend_home.
855 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
857 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
858 directory not expanded when it should be if an expanded home directory
859 was set for the address (which is overridden by the transport).
861 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
864 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
865 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
866 because it is too late at that time, and has no effect.
868 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
869 security issue with \' (bugzilla #107). I could not use the
870 PQescapeStringConn() function, because it needs a PGconn value as one of
873 PH/08 When testing addresses using -bt, indicate those final addresses that
874 are duplicates that would not cause an additional delivery. At least one
875 person was confused, thinking that -bt output corresponded to deliveries.
876 (Suppressing duplicates isn't a good idea as you lose the information
877 about possibly different redirections that led to the duplicates.)
879 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
880 systems where poll() doesn't work, in particular OS X.
882 PH/10 Added more information to debugging output for retry time not reached.
884 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
885 operations in malware.c.
887 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
890 PH/13 If write_rejectlog was set false when logging was sent to syslog with
891 syslog_duplication set false, log lines that would normally be written
892 both the the main log and to the reject log were not written to syslog at
895 PH/14 In the default configuration, change the use of "message" in ACL warn
896 statements to "add_header".
898 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
899 not followed by a command (e.g. "seen endif").
901 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
902 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
905 PH/17 Added extra conditions to the default value of delay_warning_condition
909 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
910 { match{$h_precedence:}{(?i)bulk|list|junk} } \
911 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
914 The Auto-Submitted: and various List- headers are standardised, whereas I
915 don't think Precedence: ever was.
917 PH/18 Refactored debugging code in route_finduser() to show more information,
918 in particular, the error code if getpwnam() issues one.
920 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
921 This is apparently needed in addition to the PH/07 change above to avoid
922 any possible encoding problems.
924 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
925 but not after initializing Perl.
927 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
928 output them only if debugging. By default they are written stderr,
929 apparently, which is not desirable.
931 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
934 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
937 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
939 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
940 authenticated or an ident call has been made. Suppress the default
941 values for $authenticated_id and $authenticated_sender (but permit -oMai
942 and -oMas) when testing with -bh.
944 PH/24 Re-jigged the order of the tests in the default configuration so that the
945 tests for valid domains and recipients precede the DNS black list and CSA
946 tests, on the grounds that those ones are more expensive.
948 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
949 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
950 This bug exists in every version of Exim that I still have, right back to
953 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
954 However, an attempt to turn on case-sensitivity in a regex key by
955 including (?-i) didn't work because the subject string was already
956 lowercased, and the effects were non-intuitive. It turns out that a
957 one-line patch can be used to allow (?-i) to work as expected.
963 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
964 other effects) broke the use of negated acl sub-conditions.
966 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
969 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
970 "Deny" causes Exim to reject the incoming connection with a 554 error.
971 Unfortunately, if there is a major crisis, such as a disk failure,
972 tcp-wrappers gives "deny", whereas what one would like would be some
973 kind of temporary error. A kludge has been added to help with this.
974 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
975 554 error is used if errno is still zero or contains ENOENT (which occurs
976 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
979 PH/03 Add -lutil to the default FreeBSD LIBS setting.
981 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
982 errors. Otherwise a message that provokes a temporary error (when other
983 messages do not) can cause a whole host to time out.
985 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
986 addresses were routed directly to files or pipes from a redirect router.
987 File deliveries just didn't batch; pipe deliveries might have suffered
990 PH/06 A failure to get a lock for a hints database would erroneously always say
991 "Failed to get write lock", even when it was really a read lock.
993 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
994 of 0600. This has been changed to use the value of the lockfile_mode
995 option (which defaults to 0600).
997 PH/08 Applied small patch from the Sieve maintainer.
999 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1000 folder from quota calculations, a direct delivery into this folder messed
1001 up the contents of the maildirsize file. This was because the regex was
1002 used only to exclude .Trash (or whatever) when the size of the mailbox
1003 was calculated. There was no check that a delivery was happening into an
1004 excluded directory. This bug has been fixed by ignoring all quota
1005 processing for deliveries into excluded directories.
1007 PH/10 Added the maildirfolder_create_regex option to appendfile.
1013 PH/01 The code for finding all the local interface addresses on a FreeBSD
1014 system running IPv6 was broken. This may well have applied to all BSD
1015 systems, as well as to others that have similar system calls. The broken
1016 code found IPv4 interfaces correctly, but gave incorrect values for the
1017 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1018 that it would not match correctly against @[] and not recognize the IPv6
1021 PH/02 The ipliteral router was not recognizing addresses of the form user@
1022 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1024 PH/03 Added disable_ipv6.
1026 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1027 decoded form, because it is most often used to construct To: headers
1028 lines in autoreplies, and the decoded form may well be syntactically
1029 invalid. However, $reply_address has leading white space removed, and all
1030 newlines turned into spaces so that the autoreply transport does not
1033 PH/05 If group was specified without a user on a router, and no group or user
1034 was specified on a transport, the group from the router was ignored.
1036 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1037 for visible compile-time settings that can be used to change these
1038 numbers, for those that want even more. Backwards compatibility with old
1039 spool files has been maintained. However, going back to a previous Exim
1040 release will lost any variables that are in spool files.
1042 PH/07 Two small changes when running in the test harness: increase delay when
1043 passing a TCP/IP connection to a new process, in case the original
1044 process has to generate a bounce, and remove special handling of
1045 127.0.0.2 (sic), which is no longer necessary.
1047 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1048 be the same on different OS.
1050 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1053 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1054 whether --show-vars was specified or not
1056 JJ/02 exipick: Added support for new ACL variable spool format introduced
1059 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1060 syntactically invalid From: or Reply-to: line, and a filter used this to
1061 generate an autoreply, and therefore failed to obtain an address for the
1062 autoreply, Exim could try to deliver to a non-existent relative file
1063 name, causing unrelated and misleading errors. What now happens is that
1064 it logs this as a hard delivery error, but does not attempt to create a
1067 PH/11 The exinext utility has a -C option for testing purposes, but although
1068 the given file was scanned by exinext itself; it wasn't being passed on
1069 when Exim was called.
1071 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1072 an end-of-file indication when reading a command response.
1074 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1075 compiled. In many other places in Exim, IPv6 addresses are always
1076 recognized, so I have changed this. It also means that IPv4 domain
1077 literals of the form [IPV4:n.n.n.n] are now always recognized.
1079 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1080 used if the router is not running as root, for example, when verifying at
1081 ACL time, or when using -bh. The debugging output from this situation was
1082 non-existent - all you got was a failure to exec. I have made two
1085 (a) Failures to set uid/gid, the current directory, or a process leader
1086 in a subprocess such as that created by queryprogram now generate
1087 suitable debugging ouput when -d is set.
1089 (b) The queryprogram router detects when it is not running as root,
1090 outputs suitable debugging information if -d is set, and then runs
1091 the subprocess without attempting to change uid/gid.
1093 PH/15 Minor change to Makefile for building test_host (undocumented testing
1096 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1097 additional section of a DNS packet that returns MX or SRV records.
1098 Instead, it always explicitly searches for A/AAAA records. This avoids
1099 major problems that occur when a DNS server includes only records of one
1100 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1101 fixed another bug: if SRV records were looked up and the corresponding
1102 address records were *not* found in the additional section, the port
1103 values from the SRV records were lost.
1105 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1106 using the correct key (the original address) when searching the retry
1107 rules in order to find which one to use for generating the retry hint.
1109 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1110 adding the default one. Similarly, if it contains a Reply-To: header, the
1111 errors_reply_to option, if set, is not used.
1113 PH/19 When calculating a retry time, Exim used to measure the "time since
1114 failure" by looking at the "first failed" field in the retry record. Now
1115 it does not use this if it is later than than the arrival time of the
1116 message. Instead it uses the arrival time. This makes for better
1117 behaviour in cases where some deliveries succeed, thus re-setting the
1118 "first failed" field. An example is a quota failure for a huge message
1119 when small messages continue to be delivered. Without this change, the
1120 "time since failure" will always be short, possible causing more frequent
1121 delivery attempts for the huge message than are intended.
1122 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1124 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1125 $1, $2, $3) because the numerical variables can be reset during some
1126 expansion items (e.g. "match"), thereby losing the authentication data.
1128 PH/21 Make -bV show the size of off_t variables so that the test suite can
1129 decide whether to run tests for quotas > 2G.
1131 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1132 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1133 filecount value is greater than 2G or if a quota value is greater than 2G
1134 on a system where the size of off_t is not greater than 4, a panic error
1137 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1138 never match. The debug and -bh output now contains an explicit error
1139 message indicating a malformed IPv4 address or mask.
1141 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1142 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1143 PH/23 above applies.
1145 PH/25 Do not write to syslog when running in the test harness. The only
1146 occasion when this arises is a failure to open the main or panic logs
1147 (for which there is an explicit test).
1149 PH/26 Added the /no_tell option to "control=freeze".
1151 PH/27 If a host name lookup failed very early in a connection, for example, if
1152 the IP address matched host_lookup and the reverse lookup yielded a name
1153 that did not have a forward lookup, an error message of the form "no IP
1154 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1155 could be logged. Now it outputs the IP address instead of "NULL".
1157 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1158 allows the sender and the authenticated sender to be set when
1159 submitting a message from within Exim. Since child_open_exim() is
1160 documented for local_scan(), the new function should be too.
1162 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1163 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1164 results in an empty string is now treated as unset.
1166 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1168 PH/31 Added sender_verify_fail logging option.
1170 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1171 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1172 tidied the source and removed it altogether.
1174 PH/33 When a queue run was abandoned because the load average was too high, a
1175 log line was always written; now it is written only if the queue_run log
1176 selector is set. In addition, the log line for abandonment now contains
1177 information about the queue run such as the pid. This is always present
1178 in "start" and "stop" lines but was omitted from the "abandon" line.
1180 PH/34 Omit spaces between a header name and the colon in the error message that
1181 is given when verify = headers_syntax fails (if there are lots of them,
1182 the message gets confusing).
1184 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1185 names, as there are now some PTR records that contain slashes. This check
1186 is only to protect against broken name servers that fall over on strange
1187 characters, so the fact that it applies to all lookups doesn't matter.
1189 PH/36 Now that the new test suite is complete, we can remove some of the
1190 special code in Exim that was needed for the old test suite. For example,
1191 sorting DNS records because real resolvers return them in an arbitrary
1192 order. The new test suite's fake resolver always returns records in the
1195 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1196 bounces) except when queue_only is set, to avoid logging races between
1197 the different processes.
1199 PH/38 Panic-die if .include specifies a non-absolute path.
1201 PH/39 A tweak to the "H" retry rule from its user.
1203 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1204 a label. They prevented compilation on older perls.
1206 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1207 a warning to be raised on newish perls.
1209 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1210 on queue. Changes to match documented behaviour of showing count of
1211 messages matching specified criteria.
1213 PH/40 Changed the default ident timeout from 30s to 5s.
1215 PH/41 Added support for the use of login_cap features, on those BSD systems
1216 that have them, for controlling the resources used by pipe deliveries.
1218 PH/42 The content-scanning code uses fopen() to create files in which to put
1219 message data. Previously it was not paying any attention to the mode of
1220 the files. Exim runs with umask(0) because the rest of the code creates
1221 files with open(), and sets the required mode explicitly. Thus, these
1222 files were ending up world-writeable. This was not a big issue, because,
1223 being within the spool directory, they were not world-accessible. I have
1224 created a function called modefopen, which takes an additional mode
1225 argument. It sets umask(777), creates the file, chmods it to the required
1226 mode, then resets the umask. All the relevant calls to fopen() in the
1227 content scanning code have been changed to use this function.
1229 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1230 to 24 hours. This avoids potential overflow problems when processing G
1231 and H retry rules. I suspect nobody ever tinkers with this value.
1233 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1235 PH/45 When the plaintext authenticator is running as a client, the server's
1236 challenges are checked to ensure they are valid base64 strings. By
1237 default, the authentication attempt is cancelled if an invalid string is
1238 received. Setting client_ignore_invalid_base64 true ignores these errors.
1239 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1240 they are received. Thus, the responses can be made to depend on the
1241 challenges. If an invalid string is ignored, an empty string is placed in
1244 PH/46 Messages that are created by the autoreply transport now contains a
1245 References: header, in accordance with RFCs 2822 and 3834.
1247 PH/47 Added authenticated_sender_force to the smtp transport.
1249 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1251 PH/49 Installed latest patch from the Sieve maintainer.
1253 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1254 was also set, the appendfile transport was unnecessarily scanning a
1255 directory of message files (e.g. for maildir delivery) to find the count
1256 of files (along with the size), even though it did not need this
1257 information. It now does the scan only if it needs to find either the
1258 size of the count of files.
1260 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1262 PH/52 Two bugs concerned with error handling when the smtp transport is
1265 (i) Exim was not creating retry information for temporary errors given
1266 for individual recipients after the DATA command when the smtp transport
1267 was used in LMTP mode. This meant that they could be retried too
1268 frequently, and not timed out correctly.
1270 (ii) Exim was setting the flag that allows error details to be returned
1271 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1272 recipients that were returned after the DATA command.
1274 PH/53 This is related to PH/52, but is more general: for any failing address,
1275 when detailed error information was permitted to be returned to the
1276 sender, but the error was temporary, then after the final timeout, only
1277 "retry timeout exceeded" was returned. Now it returns the full error as
1278 well as "retry timeout exceeded".
1280 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1281 do this, and (what is worse) MTAs that accept it.
1283 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1284 will now be deprecated.
1286 PH/56 New os.c-cygwin from the Cygwin maintainer.
1288 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1289 formats (previously only available in exim formats via -bpr, -bpru,
1290 and -bpra. Now also available in native and exiqgrep formats)
1292 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1293 with very large, slow to parse queues
1295 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1297 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1299 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1300 responses to authentication challenges, though it was showing the
1301 challenges; (ii) I've removed the CR characters from the debug output for
1304 PH/58 Allow for the insertion of a newline as well as a space when a string
1305 is turned into more than one encoded-word during RFC 2047 encoding. The
1306 Sieve code now uses this.
1308 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1309 data_4xx, lost_connection, tls_required.
1311 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1312 message was being sent as an SMTP response.
1314 PH/61 Add -l and -k options to exicyclog.
1316 PH/62 When verifying, if an address was redirected to one new address, so that
1317 verification continued, and the new address failed or deferred after
1318 having set something in $address_data, the value of $address_data was not
1319 passed back to the ACL. This was different to the case when no
1320 redirection occurred. The value is now passed back in both cases.
1322 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1323 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1324 use login_cap.h, so on its own it isn't the distinguishing feature. The
1325 new name refers directly to the setclassresources() function.
1327 PH/65 Added configuration files for NetBSD3.
1329 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1331 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1332 is preferred over IPv4.
1334 PH/68 The bounce_return_message and bounce_return_body options were not being
1335 honoured for bounces generated during the reception of non-SMTP messages.
1336 In particular, this applied to messages rejected by the ACL. This bug has
1337 been fixed. However, if bounce_return_message is true and bounce_return_
1338 body is false, the headers that are returned for a non-SMTP message
1339 include only those that have been read before the error was detected.
1340 (In the case of an ACL rejection, they have all been read.)
1342 PH/69 The HTML version of the specification is now built in a directory called
1343 spec_html instead of spec.html, because the latter looks like a path with
1344 a MIME-type, and this confuses some software.
1346 PH/70 Catch two compiler warnings in sieve.c.
1348 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1349 function verify_get_ident() calls ip_connect() to connect a socket, but
1350 if the "connect()" function timed out, ip_connect() used to close the
1351 socket. However, verify_get_ident() also closes the socket later, and in
1352 between Exim writes to the log, which may get opened at this point. When
1353 the socket was closed in ip_connect(), the log could get the same file
1354 descriptor number as the socket. This naturally causes chaos. The fix is
1355 not to close the socket in ip_connect(); the socket should be closed by
1356 the function that creates it. There was only one place in the code where
1357 this was missing, in the iplookup router, which I don't think anybody now
1358 uses, but I've fixed it anyway.
1360 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1361 well as to direct DNS lookups. Otherwise the handling of names in host
1362 lists is inconsistent and therefore confusing.
1368 PH/01 Two changes to the default runtime configuration:
1370 (1) Move the checks for relay_from_hosts and authenticated clients from
1371 after to before the (commented out) DNS black list checks.
1373 (2) Add control=submission to the relay_from_hosts and authenticated
1374 clients checks, on the grounds that messages accepted by these
1375 statements are most likely to be submissions.
1377 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1379 (1) Generate an error if the third argument for the ${prvs expansion is
1382 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1385 (3) Reset the variables that are obtained from the first argument of
1386 ${prvscheck and used in the second argument before leaving the code,
1387 because their memory is reclaimed, so using them afterwards may do
1390 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1391 one (it's much easier than Tom thought :-).
1393 (5) Because of (4), we can now allow for the use of $prvscheck_result
1394 inside the third argument.
1396 PH/03 For some reason, the default setting of PATH when running a command from
1397 a pipe transport was just "/usr/bin". I have changed it to
1400 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1401 anything to be listed in the output from -bV.
1403 PH/05 When a filter generated an autoreply, the entire To: header line was
1404 quoted in the delivery log line, like this:
1406 => >A.N.Other <ano@some.domain> <original@ddress> ...
1408 This has been changed so that it extracts the operative address. There
1409 may be more than one such address. If so, they are comma-separated, like
1412 => >ano@some.domain,ona@other.domain <original@ddress> ...
1414 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1415 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1416 not being looked up in the rDNS to get a host name, Exim was showing the
1417 IP address twice in Received: lines, even though the IP addresses were
1418 identical. For example:
1420 Received: from [1.2.3.4] (helo=[1.2.3.4])
1422 However, if the real host name was known, it was omitting the HELO data
1423 if it matched the actual IP address. This has been tidied up so that it
1424 doesn't show the same IP address twice.
1426 PH/07 When both +timestamp and +memory debugging was on, the value given by
1427 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1428 called by the debug printing, thereby overwriting the timestamp buffer.
1429 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1431 PH/08 When the original message was included in an autoreply transport, it
1432 always said "this is a copy of the message, including all the headers",
1433 even if body_only or headers_only was set. It now gives an appropriate
1436 PH/09 Applied a patch from the Sieve maintainer which:
1438 o fixes some comments
1439 o adds the (disabled) notify extension core
1440 o adds some debug output for the result of if/elsif tests
1441 o points to the current vacation draft in the documentation
1442 and documents the missing references header update
1446 o fixes a bug in processing the envelope test (when testing
1447 multiple envelope elements, the last element determinted the
1450 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1451 Electronic Mail") by including:
1453 Auto-submitted: auto-generated
1455 in the messages that it generates (bounce messages and others, such as
1456 warnings). In the case of bounce messages for non-SMTP mesages, there was
1457 also a typo: it was using "Auto_submitted" (underscore instead of
1458 hyphen). Since every message generated by Exim is necessarily in response
1459 to another message, thes have all been changed to:
1461 Auto-Submitted: auto-replied
1463 in accordance with these statements in the RFC:
1465 The auto-replied keyword:
1467 - SHOULD be used on messages sent in direct response to another
1468 message by an automatic process,
1470 - MUST NOT be used on manually-generated messages,
1472 - MAY be used on Delivery Status Notifications (DSNs) and Message
1473 Disposition Notifications (MDNs),
1475 - MUST NOT be used on messages generated by automatic or periodic
1476 processes, except for messages which are automatic responses to
1479 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1480 to the default Received: header definition.
1482 PH/12 Added log selector acl_warn_skipped (default on).
1484 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1485 variables because (a) they are in the wrong storage pool and (b) even if
1486 they were copied, it wouldn't work properly because of the caching.
1488 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1489 checking when decoding. Apparently there are clients that generate
1490 overlong encoded strings. Why am I not surprised?
1492 PH/15 If the first argument of "${if match_address" was not empty, but did not
1493 contain an "@" character, Exim crashed. Now it writes a panic log message
1494 and treats the condition as false.
1496 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1498 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1499 extension "envelope-auth". The code is finished and in agreement with
1500 other implementations, but there is no documentation so far and in fact,
1501 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1502 not changing the active code.
1504 Print executed "if" and "elsif" statements when debugging is used. This
1505 helps a great deal to understand what a filter does.
1507 Document more things not specified clearly in RFC3028. I had all this
1508 sorted out, when out of a sudden new issues came to my mind. Oops."
1510 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1513 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1514 canonical form (as documented). However, after a host name lookup from
1515 the IP address, check_host() was doing a simple string comparison with
1516 addresses acquired from the DNS when checking that the found name did
1517 have the original IP as one of its addresses. Since any found IPv6
1518 addresses are likely to be in abbreviated form, the comparison could
1519 fail. Luckily, there already exists a function for doing the comparison
1520 by converting both addresses to binary, so now that is used instead of
1521 the text comparison.
1523 PH/20 There was another similar case to PH/19, when a complete host name was
1524 given in a host list; looking up its IP address could give an abbreviated
1525 form, whereas the current host's name might or might not be abbreviated.
1526 The same fix has been applied.
1532 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1533 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1536 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1537 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1539 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1541 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1542 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1543 Also, for "Auto-Submitted", treat anything other than "no" as
1544 non-personal, in accordance with RFC 3834. (Previously it treated
1545 anything starting "auto-" as non-personal.)
1547 TF/01 The control=submission/name=... option had a problem with syntax
1548 errors if the name included a slash character. The /name= option
1549 now slurps the rest of the string, so it can include any characters
1550 but it must come last in the list of options (after /sender_retain
1553 PH/05 Some modifications to the interface to the fake nameserver for the new
1561 TK/01 Added the "success_on_redirect" address verification option. See
1562 NewStuff for rationale and an example.
1564 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1566 PH/02 Patch to exigrep to allow it to work on syslog lines.
1568 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1569 fread() to skip over the body file's header line, because in Cygwin the
1570 header line is locked and is inaccessible.
1572 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1573 co-exist for some time) to make it clear that it is the Exim ID that is
1574 referenced, not the Message-ID: header line.
1576 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1577 string_format() function, because snprintf() does not exist on all
1580 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1581 ${stat: expansion item.
1583 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1584 protocol synchronization error", to keep the pedants happy.
1586 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1587 well as for IRIX systems, when gcc is being used. See the host.c source
1590 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1592 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1595 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1596 transport and to the smtp transport in LMTP mode.
1598 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1600 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1601 run a filter in a subprocess. This could lead to confusion in subsequent
1602 lookups in the parent process. There should also be a search_tidyup() at
1603 the end of the subprocess.
1605 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1606 only if the host matched helo_try_verify_hosts, which caused the
1607 verification to occur when the EHLO/HELO command was issued. The ACL just
1608 tested the remembered result. Now, if a previous verification attempt has
1609 not happened, "verify = helo" does it there and then.
1611 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1613 TK/03 Fix log output including CR from clamd.
1615 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1616 exist provoked a memory error which could cause a segfault.
1618 PH/15 Installed PCRE 6.2
1620 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1622 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1623 of the problem. Specifically, suggested +O2 rather than +O1 for the
1626 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1628 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1629 record, the port was not correctly logged when the outgoing_port log
1630 selector was set (it logged the transort's default port).
1632 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1633 fallback_hosts, and "hosts" in the smtp transport.
1635 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1636 host errors such as "Connection refused".
1638 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1639 authentication with radiusclient 0.4.9:
1641 - Error returned from rc_read_config was caught wrongly
1642 - Username/password not passed on to radius server due to wrong length.
1644 The presumption is that some radiusclient API changes for 4.51/PH/17
1645 were not taken care of correctly. The code is still untested by me (my
1646 Linux distribution still has 0.3.2 of radiusclient), but it was
1647 contributed by a Radius user.
1649 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1650 expanding the "port" option of the smtp transport.
1652 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1653 while reading a MIME header. Thanks to Tom Hughes for a patch.
1655 PH/24 Include config.h inside local_scan.h so that configuration settings are
1658 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1659 that in an ACL for RCPT (for example), you can examine exactly what was
1662 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1663 commands, but it was not correctly comparing the address with the actual
1664 client host address. Thus, it would show the EHLO address in Received:
1665 header lines when this was not necessary.
1667 PH/27 Added the % operator to ${eval:}.
1669 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1670 it should be ignoring failures (because with -C, for example, it has lost
1671 privilege). It wasn't ignoring creation failures other than "already
1674 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1677 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1678 because an input line was too long, either on its own, or by virtue of
1679 too many continuations, the temporary file was not being removed, and the
1680 return code was incorrect.
1682 PH/31 Missing "BOOL" in function definition in filtertest.c.
1684 PH/32 Applied Sieve patches from the maintainer.
1686 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1688 PH/33 Added "verify = not_blind".
1690 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1691 Local/Makefile (with some defaults set). These are used in built scripts
1692 such as exicyclog, but they have never been used in the exim_install
1693 script (though there are many overriding facilities there). I have
1694 arranged that the exim_install script now takes note of these two
1697 PH/35 Installed configuration files for Dragonfly.
1699 PH/36 When a locally submitted message by a trusted user did not contain a
1700 From: header, and the sender address was obtained from -f or from an SMTP
1701 MAIL command, and the trusted user did not use -F to supply a sender
1702 name, $originator_name was incorrectly used when constructing a From:
1703 header. Furthermore, $originator_name was used for submission mode
1704 messages from external hosts without From: headers in a similar way,
1705 which is clearly wrong.
1707 PH/37 Added control=suppress_local_fixups.
1709 PH/38 When log_selector = +received_sender was set, and the addition of the
1710 sender made the log line's construction buffer exactly full, or one byte
1711 less than full, an overflow happened when the terminating "\n" was
1714 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1715 when the result of a list match is failure because a DNS lookup failed.
1717 PH/40 RM_COMMAND is now used in the building process.
1719 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1720 the "build-* directories that it finds.
1722 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1723 address in a domain literal was a prefix of an interface address.
1725 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1726 when verifying a sender address, unless rewrite_headers is false.
1728 PH/44 Wrote a long comment about why errors_to addresses are verified as
1729 recipients, not senders.
1731 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1732 the ratelimit ACL was added.
1734 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1736 PH/46 Added extra information about PostgreSQL errors to the error string.
1738 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1739 avoiding the need to install special zones in a real server. This is
1740 backwards compatible; if it can't find the fake resolver, it drops back.
1741 Thus, both old and new test suites can be run.
1743 TF/02 Added util/ratelimit.pl
1745 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1746 clock is set back in time.
1748 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1749 Candler <B.Candler@pobox.com>.
1751 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1752 OK for addresses that are the result of redirections.
1754 PH/48 A number of further additions for the benefit of the new test suite,
1755 including a fake gethostbyname() that interfaces to the fake DNS resolver
1758 TF/06 The fix for widen_domains has also been applied to qualify_single and
1759 search_parents which are the other dnslookup options that can cause
1762 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1765 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1767 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1768 service" if it encountered an MX record with an empty target hostname.
1769 The message is now "an MX or SRV record indicated no SMTP service".
1771 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1772 if the DNS of the sending site is misconfigured. This is quite a
1773 common situation. This change restores the behaviour of treating a
1774 helo verification defer as a failure.
1776 PH/51 If self=fail was set on a router, the bounce message did not include the
1777 actual error message.
1783 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1785 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1786 command itself ended in error, the underlying message about the transport
1787 filter timeout was being overwritten with the pipe command error. Now the
1788 underlying error message should be appended to the second error message.
1790 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1792 PH/02 Reduce the amount of output that "make" produces by default. Full output
1793 can still be requested.
1795 PH/03 The warning log line about a condition test deferring for a "warn" verb
1796 was being output only once per connection, rather than after each
1797 occurrence (because it was using the same function as for successful
1798 "warn" verbs). This seems wrong, so I have changed it.
1800 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1801 it should not have, which might have caused a crash in the right
1802 circumstances, but probably never did.
1804 PH/04 Installed a modified version of Tony Finch's patch to make submission
1805 mode fix the return path as well as the Sender: header line, and to
1806 add a /name= option so that you can make the user's friendly name appear
1809 TF/03 Added the control = fakedefer ACL modifier.
1811 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1812 Mark Lowes for thorough testing.
1814 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1816 TK/03 Merged latest SRS patch from Miles Wilton.
1818 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1819 with the definition in sysexits.h (which is #included earlier).
1820 Fortunately, Exim does not actually use EX_OK. The code used to try to
1821 preserve the sysexits.h value, by assumimg that macro definitions were
1822 scanned for macro replacements. I have been disabused of this notion,
1823 so now the code just undefines EX_OK before #including unistd.h.
1825 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1826 in the smtp transport. When a block could not be written in a single
1827 write() function, the timeout was being re-applied to each part-write.
1828 This seems wrong - if the receiver was accepting one byte at a time it
1829 would take for ever. The timeout is now adjusted when this happens. It
1830 doesn't have to be particularly precise.
1832 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1833 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1835 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1836 without a domain if the check to <postmaster@domain> fails.
1838 SC/01 Eximstats: added -xls and the ability to specify output files
1839 (patch written by Frank Heydlauf).
1841 SC/02 Eximstats: use FileHandles for outputing results.
1843 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1845 SC/04 Eximstats: fixed display of large numbers with -nvr option
1847 SC/05 Eximstats: fixed merging of reports with empty tables.
1849 SC/06 Eximstats: added the -include_original_destination flag
1851 SC/07 Eximstats: removed tabs and trailing whitespace.
1853 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1855 TK/06 MBOX spool code: Add real "From " MBOX separator line
1856 so the .eml file is really in mbox format (even though
1857 most programs do not really care). Patch from Alex Miller.
1859 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1860 The latter is generated from $received_to and is only set if the
1861 message has one envelope recipient. SA can use these headers,
1862 obviously out-of-the-box. Patch from Alex Miller.
1864 PH/08 The ${def test on a variable was returning false if the variable's
1865 value was "0", contrary to what the specification has always said!
1866 The result should be true unless the variable is empty.
1868 PH/09 The syntax error of a character other than { following "${if
1869 def:variable_name" (after optional whitespace) was not being diagnosed.
1870 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1871 accidental colon was present, for example, could give incorrect results.
1873 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1874 result is used (not including appendfile, where other changes are about
1877 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1878 This involved changing a lot of size variables from int to off_t. It
1879 should work with maildirs and everything.
1881 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1882 spamd dying while we are connected to it.
1884 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1887 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1888 function for BDB 4.3.
1890 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1892 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1893 the sources for PCRE have been re-arranged and more files are now
1896 PH/15 The code I had for printing potentially long long variables in PH/11
1897 above was not the best (it lost precision). The length of off_t variables
1898 is now inspected at build time, and an appropriate printing format (%ld
1899 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1900 to be "long long int" or "long int". This is needed for the internal
1901 formatting function string_vformat().
1903 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1904 the configuration file to be ":syslog", then the script "guesses" where
1905 the logs files are, rather than using the compiled in default. In our
1906 case the guess is not the same as the compiled default, so the script
1907 suddenly stopped working when I started to use syslog. The patch checks
1908 to see if log_file_path is "". If so, it attempts to read it from exim
1909 with no configuration file to get the compiled in version, before it
1910 falls back to the previous guessing code."
1912 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1913 implementing BATV in an Exim configuration. See NewStuff for the gory
1916 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1917 Makefile that are specific to HP-UX.
1919 PH/18 If the "use_postmaster" option was set for a recipient callout together
1920 with the "random" option, the postmaster address was used as the MAIL
1921 FROM address for the random test, but not for the subsequent recipient
1922 test. It is now used for both.
1924 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1925 patch removes a few documentation additions to RFC 3028, because the
1926 latest draft now contains them. It adds the new en;ascii-case comparator
1927 and a new error check for 8bit text in MIME parts. Comparator and
1928 require names are now matched exactly. I enabled the subaddress
1929 extension, but it is not well tested yet (read: it works for me)."
1931 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1932 rework some of the code of TK/09 above to avoid the hardwired use of
1933 "%lld" and "long long". Replaced the call to snprintf() with a call to
1936 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1937 records point to non-existent hosts", "retry timeout exceeded", and
1938 "retry time not reached for any host after a long failure period".
1940 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1941 experimental DomainKeys support:
1943 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1944 (2) On an error such as an illegally used "control", the wrong name for
1945 the control was given.
1947 These problems did NOT occur unless DomainKeys support was compiled.
1949 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1951 PH/24 Added ${if match_ip condition.
1953 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1954 databases so that it will be absolutely obvious if a crash occurs in the
1955 DB library. This is a regular occurrence (often caused by mis-matched
1958 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1959 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1960 picked up on a user's system that detects such things. There doesn't seem
1961 to be a gcc warning option for this - only an attribute that has to be
1962 put on the function's prototype. It seems that in Fedora Core 4 they have
1963 set this on a number of new functions. No doubt there will be more in due
1966 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1967 specify a transport. However, if an address that was verified by such a
1968 router was the subject of a callout, Exim crashed because it tried to
1969 read the rcpt_include_affixes from the non-existent transport. Now it
1970 just assumes that the setting of that option is false. This bug was
1971 introduced by 4.51/PH/31.
1973 PH/28 Changed -d+all to exclude +memory, because that information is very
1974 rarely of interest, but it makes the output a lot bigger. People tend to
1975 do -d+all out of habit.
1977 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1978 code in os-type was giving problems when libc.so lives in lib64, like on
1981 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1982 aren't the modern standard, and it seems that some systems' include files
1983 don't always have them. Exim was already checking for some of the newer
1984 ones like T_AAAA, and defining it itself. I've added checks for all the
1985 record types that Exim uses.
1987 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1988 not automatically generating a new one, as it is supposed to. This
1989 prevented TLS from working. If the file did exist, but contained invalid
1990 data, a new version was generated, as expected. It was only the case of a
1991 non-existent file that was broken.
1993 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1994 with a change in libdomainkeys > 0.64.
1996 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1997 from DNS. If the selector record carries the flag, it now has
1998 precedence over the domain-wide flag.
2000 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2002 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2003 the use of a port name that isn't defined in /etc/services) occurred, the
2004 message was deferred as in a normal delivery, and thus remained on the
2005 spool, instead of being failed because of the mua_wrapper setting. This
2006 is now fixed, and I tidied up some of the mua_wrapper messages at the
2009 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2010 the messages in a hash of arrays rather than using individual hashes.
2011 This is a bit cleaner and results in dramatic memory savings, albeit
2012 at a slight CPU cost.
2014 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2015 as requested by Marc Sherman.
2017 SC/10 Eximstats: added histograms for user specified patterns as requested
2020 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2022 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2023 fopen() in the content-scanning modules that did not already have it.
2029 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2030 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2032 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2034 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2036 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2037 to test Sieve filters that use "vacation".
2039 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2040 that changes the way the GnuTLS parameters are stored in the cache file.
2041 The new format can be generated externally. For backward compatibility,
2042 if the data in the cache doesn't make sense, Exim assumes it has read an
2043 old-format file, and it generates new data and writes a new file. This
2044 means that you can't go back to an older release without removing the
2047 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2048 work if there are any delivery delays because "one_time" forces the
2049 parent to be marked "delivered", so its unseen clone is never tried
2050 again. For this reason, Exim now forbids the simultaneous setting of
2053 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2054 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2055 entry if you want to know the details. The fix, however, neglected to
2056 consider the case when local delivery batching is involved. The test for
2057 "previously delivered" was not happening when checking to see if an
2058 address could be batched with a previous (undelivered) one; under
2059 certain circumstances this could lead to multiple deliveries to the same
2062 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2063 in its include files, and this causes problems building Exim.
2065 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2066 header_syntax) but Exim was just ignoring anything given after a slash.
2067 In particular, this caused confusion with an attempt to use "verify =
2068 reverse_host_lookup/defer_ok". An error is now given when options are
2069 supplied for verify items that do not have them. (Maybe reverse_host_
2070 lookup should have a defer_ok option, but that's a different point.)
2072 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2073 defined by RFC 821) to 2048, because there were problems with some AUTH
2074 commands, and RFC 1869 says the size should be increased for extended
2075 SMTP commands that take arguments.
2077 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2080 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2081 "unknown" error; now it says that the functionality isn't in the binary.
2083 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2084 an address' error message when a string expansion fails (syntax or
2085 whatever). Otherwise the password may appear in the log. Following change
2086 PH/42 below, there is no longer a chance of it appearing in a bounce
2089 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2091 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2092 of its addresses was ever tried. (Bugzilla bug #2.)
2094 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2095 the result incorrectly in the debug output. (It correctly added a newline
2096 to what was transported.)
2098 TF/01 Added $received_time.
2100 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2101 commented out examples of how to interface to a virus scanner and to
2102 SpamAssassin. Also added commented examples of av_scanner and
2103 spamd_address settings.
2105 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2106 and controls are allowed in which ACLs. There were a couple of minor
2107 errors. Some of the entries in the conditions table (which is a table of
2108 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2109 negation of where the condition IS allowed.
2111 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2113 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2114 header file does not have a version number, so I've had to invent a new
2115 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2116 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2117 radiusclient), but it was contributed by a Radius user.
2119 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2120 files or directories, for OpenSSL.
2122 PH/19 When an Exim process that is running as root has to create an Exim log
2123 file, it does so in a subprocess that runs as exim:exim so as to get the
2124 ownership right at creation (otherwise, other Exim processes might see
2125 the file with the wrong ownership). There was no test for failure of this
2126 fork() call, which would lead to the process getting stuck as it waited
2127 for a non-existent subprocess. Forks do occasionally fail when resources
2128 run out. I reviewed all the other calls to fork(); they all seem to check
2131 PH/20 When checking for unexpected SMTP input at connect time (before writing
2132 the banner), Exim was not dealing correctly with a non-positive return
2133 from the read() function. If the client had disconnected by this time,
2134 the result was a log entry for a synchronization error with an empty
2135 string after "input=" when read() returned zero. If read() returned -1
2136 (an event I could not check), uninitialized data bytes were printed.
2137 There were reports of junk text (parts of files, etc) appearing after
2140 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2142 PH/22 Added support for macro redefinition, and (re)definition in between
2143 driver and ACL definitions.
2145 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2146 forgetting to use the resulting value; it was using the unexpanded value.
2148 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2149 hadn't been configured. The fix is from Juergen Kreileder, who
2150 understands it better than I do:
2152 "Here's what I see happening with three configured cyrus_sasl
2153 authenticators configured (plain, login, cram-md5):
2155 On startup auth_cyrus_sasl_init() gets called for each of these.
2156 This means three calls to sasl_listmech() without a specified mech_list.
2157 => SASL tests which mechs of all available mechs actually work
2158 => three warnings about OTP not working
2159 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2161 With the patch, sasl_listmech() also gets called three times. But now
2162 SASL's mech_list option is set to the server_mech specified in the the
2163 authenticator. Or in other words, the answer from sasl_listmech()
2164 gets limited to just the mech you're testing for (which is different
2166 => the return list contains just 'plain' or 'login', 'cram-md5' or
2167 nothing depending on the value of ob->server_mech.
2169 I've just tested the patch: Authentication still works fine,
2170 unavailable mechs specified in the exim configuration are still
2171 caught, and the auth.log warnings about OTP are gone."
2173 PH/25 When debugging is enabled, the contents of the command line are added
2174 to the debugging output, even when log_selector=+arguments is not
2177 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2178 answer is "GNU", and only if the return is "GNU/something" is the answer
2181 PH/27 $acl_verify_message is now set immediately after the failure of a
2182 verification in an ACL, and so is available in subsequent modifiers. In
2183 particular, the message can be preserved by coding like this:
2185 warn !verify = sender
2186 set acl_m0 = $acl_verify_message
2188 Previously, $acl_verify_message was set only while expanding "message"
2189 and "log_message" when a very denied access.
2191 PH/28 Modified OS/os.c-Linux with
2193 -#ifndef OS_LOAD_AVERAGE
2194 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2196 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2197 nomenclature these days.)
2199 PH/29 Installed patch from the Sieve maintainer that adds the options
2200 sieve_useraddress and sieve_subaddress to the redirect router.
2202 PH/30 In these circumstances:
2203 . Two addresses routed to the same list of hosts;
2204 . First host does not offer TLS;
2205 . First host accepts first address;
2206 . First host gives temporary error to second address;
2207 . Second host offers TLS and a TLS session is established;
2208 . Second host accepts second address.
2209 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2210 and peerdn, if requested) that were in fact used only for the second
2213 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2214 attention to any local part prefix or suffix that was matched by the
2215 router that accepted the address. It now behaves in the same way as it
2216 does for delivery: the affixes are removed from the local part unless
2217 rcpt_include_affixes is set on the transport.
2219 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2220 timeout during the DATA phase of an incoming message.
2222 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2223 applied a patch sanctioned by the Sieve maintainer.
2225 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2226 the uid or gid is negative. A case of a negative gid caused this to be
2227 noticed. The fix allows for either to be negative.
2229 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2230 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2233 PH/36 The change PH/12 above was broken. Fixed it.
2235 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2236 the grounds that routing the same address twice would always produce the
2237 same answer. This might have been true once, but it is certainly no
2238 longer true now. Routing a child address may depend on the previous
2239 routing that produced that child. Some complicated redirection strategies
2240 went wrong when messages had multiple recipients, and made Exim's
2241 behaviour dependent on the order in which the addresses were given.
2243 I have moved the duplicate checking until after the routing is complete.
2244 Exim scans the addresses that are assigned to local and remote
2245 transports, and removes any duplicates. This means that more work will be
2246 done, as duplicates will always all be routed, but duplicates are
2247 presumably rare, so I don't expect this is of any significance.
2249 For deliveries to pipes, files, and autoreplies, the duplicate checking
2250 still happens during the routing process, since they are not going to be
2253 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2254 It corrects a timeout issue with spamd. This is Ian's comment: "The
2255 background is that sometimes spamd either never reads data from a
2256 connection it has accepted, or it never writes response data. The exiscan
2257 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2258 blindly assumes that writes won't block so it may never time out."
2260 PH/39 Allow G after quota size as well as K and M.
2262 PH/40 The value set for $authenticated_id in an authenticator may not contain
2263 binary zeroes or newlines because the value is written to log lines and
2264 to spool files. There was no check on this. Now the value is run through
2265 the string_printing() function so that such characters are converted to
2266 printable escape sequences.
2268 PH/41 $message_linecount is a new variable that contains the total number of
2269 lines in the message. Compare $body_linecount, which is the count for the
2272 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2273 bounce and delay warning messages, except in certain special cases, which
2276 (a) An SMTP error message from a remote host;
2277 (b) A message specified in a :fail: redirection;
2278 (c) A message specified in a "fail" command in a system filter;
2279 (d) A message specified in a FAIL return from the queryprogram router;
2280 (e) A message specified by the cannot_route_message router option.
2282 In these cases only, Exim does include the error details in bounce and
2283 warning messages. There are also a few cases where bland messages such
2284 as "unrouteable address" or "local delivery error" are given.
2286 PH/43 $value is now also set for the "else" part of a ${run expansion.
2288 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2289 being worked on, but at least Exim now implements the latest version to
2292 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2293 process to complete was treated as a delivery failure, a timeout while
2294 writing the message to the pipe was logged, but erroneously treated as a
2295 successful delivery. Such timeouts include transport filter timeouts. For
2296 consistency with the overall process timeout, these timeouts are now
2297 treated as errors, giving rise to delivery failures by default. However,
2298 there is now a new Boolean option for the pipe transport called
2299 timeout_defer, which, if set TRUE, converts the failures into defers for
2300 both kinds of timeout. A transport filter timeout is now identified in
2303 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2304 systems where "make" and "gmake" are different, calling "gmake" at top
2305 level broke things. I've arranged for the value of $(MAKE) to be passed
2306 from the Makefile to this script so that it can call the same version of
2310 A note about Exim versions 4.44 and 4.50
2311 ----------------------------------------
2313 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2314 changes of various kinds. As a consequence, a big documentation update was
2315 needed. This delayed the release for rather longer than seemed good, especially
2316 in the light of a couple of (minor) security issues. Therefore, the changes
2317 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2318 release. So 4.44 and 4.50 are in effect two different branches that both start
2321 I have left the 4.50 change log unchanged; it contains all the changes since
2322 4.43. The change log for 4.44 is below; many of its items are identical to
2323 those for 4.50. This seems to be the most sensible way to preserve the
2324 historical information.
2330 1. Minor wording change to the doc/README.SIEVE file.
2332 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2333 computation of the current number of files was incorrect.
2335 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2336 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2339 4. Give more explanation in the error message when the command for a transport
2340 filter fails to execute.
2342 5. There are several places where Exim runs a non-Exim command in a
2343 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2344 was being done only for the command run by the queryprogram router. It is
2345 now done for all such subprocesses. The other cases are: ${run, transport
2346 filters, and the commands run by the lmtp and pipe transports.
2348 6. Added CONFIGURE_GROUP build-time option.
2350 7. Some older OS have a limit of 256 on the maximum number of file
2351 descriptors. Exim was using setrlimit() to set 1000 as a large value
2352 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2353 systems. I've change it so that if it can't get 1000, it tries for 256.
2355 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2356 was an oversight, and furthermore, ever since the addition of extra
2357 controls (e.g. 4.43/32), the checks on when to allow different forms of
2358 "control" were broken. There should now be diagnostics for all cases when a
2359 control that does not make sense is encountered.
2361 9. Added the /retain_sender option to "control=submission".
2363 10. $recipients is now available in the predata ACL (oversight).
2365 11. Tidy the search cache before the fork to do a delivery from a message
2366 received from the command line. Otherwise the child will trigger a lookup
2367 failure and thereby defer the delivery if it tries to use (for example) a
2368 cached ldap connection that the parent has called unbind on.
2370 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2371 of $address_data from the recipient verification was clobbered by the
2372 sender verification.
2374 13. The value of address_data from a sender verification is now available in
2375 $sender_address_data in subsequent conditions in the ACL statement.
2377 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2379 15. Added a new option "connect=<time>" to callout options, to set a different
2382 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2383 was its contents. (It was OK if the option was not defined at all.)
2385 17. A "Completed" log line is now written for messages that are removed from
2386 the spool by the -Mrm option.
2388 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2389 information about exactly what failed.
2391 19. Added -dd to debug only the daemon process.
2393 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2394 handles timeouts, both on the server side and network timeouts. Renamed the
2395 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2397 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2398 It is now set to "smtps".
2400 22. $host_address is now set to the target address during the checking of
2401 ignore_target_hosts.
2403 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2404 being passed; this would have caused $sender_host_name to have been used if
2405 matching the list had actually called for a host name (not very likely,
2406 since this list is usually IP addresses). A host name is now passed as
2409 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2410 code that specifies a non-restarting handler (typically sigaction() in
2411 modern systems) in an attempt to fix a rare and obscure crash bug.
2413 25. Narrowed the window for a race in the daemon that could cause it to ignore
2414 SIGCHLD signals. This is not a major problem, because they are used only to
2415 wake it up if nothing else does.
2417 26. A malformed maildirsize file could cause Exim to calculate negative values
2418 for the mailbox size or file count. Odd effects could occur as a result.
2419 The maildirsize information is now recalculated if the size or filecount
2422 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2423 support for a long time. Removed HAVE_SYS_VFS_H.
2425 28. Installed the latest version of exipick from John Jetmore.
2427 29. In an address list, if the pattern was not a regular expression, an empty
2428 subject address (from a bounce message) matched only if the pattern was an
2429 empty string. Non-empty patterns were not even tested. This was the wrong
2430 because it is perfectly reasonable to use an empty address as part of a
2431 database query. An empty address is now tested by patterns that are
2432 lookups. However, all the other forms of pattern expect the subject to
2433 contain a local part and a domain, and therefore, for them, an empty
2434 address still always fails if the pattern is not itself empty.
2436 30. Exim went into a mad DNS loop when attempting to do a callout where the
2437 host was specified on an smtp transport, and looking it up yielded more
2438 than one IP address.
2440 31. Re-factored the code for checking spool and log partition space into a
2441 function that finds that data and another that does the check. The former
2442 is then used to implement four new variables: $spool_space, $log_space,
2443 $spool_inodes, and $log_inodes.
2445 32. The RFC2047 encoding function was originally intended for short strings
2446 such as real names; it was not keeping to the 75-character limit for
2447 encoded words that the RFC imposes. It now respects the limit, and
2448 generates multiple encoded words if necessary. To be on the safe side, I
2449 have increased the buffer size for the ${rfc2047: expansion operator from
2452 33. It is now permitted to omit both strings after an "if" condition; if the
2453 condition is true, the result is "true". As before, when the second string
2454 is omitted, a false condition yields an empty string. This makes it less
2455 cumbersome to write custom ACL and router conditions.
2457 34. Failure to deliver a bounce message always caused it to be frozen, even if
2458 there was an errors_to setting on the router. The errors_to setting is now
2461 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2462 canonical form (fully expanded) before being placed in
2463 $sender_host_address.
2465 36. The table in the code that translates DNS record types into text (T_A to
2466 "A" for instance) was missing entries for NS and CNAME. It is just possible
2467 that this could have caused confusion if both these types were looked up
2468 for the same domain, because the text type is used as part of Exim's
2469 per-process caching. But the chance of anyone hitting this buglet seems
2472 37. The dnsdb lookup has been extended in a number of ways.
2474 (1) There is a new type, "zns", which walks up the domain tree until it
2475 finds some nameserver records. It should be used with care.
2477 (2) There is a new type, "mxh", which is like "mx" except that it returns
2478 just the host names, not the priorities.
2480 (3) It is now possible to give a list of domains (or IP addresses) to be
2481 looked up. The behaviour when one of the lookups defers can be
2482 controlled by a keyword.
2484 (4) It is now possible to specify the separator character for use when
2485 multiple records are returned.
2487 38. The dnslists ACL condition has been extended: it is now possible to supply
2488 a list of IP addresses and/or domains to be looked up in a particular DNS
2491 39. Added log_selector=+queue_time_overall.
2493 40. When running the queue in the test harness, wait just a tad after forking a
2494 delivery process, to get repeatability of debugging output.
2496 41. Include certificate and key file names in error message when GnuTLS fails
2497 to set them up, because the GnuTLS error message doesn't include the name
2498 of the failing file when there is a problem reading it.
2500 42. Allow both -bf and -bF in the same test run.
2502 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2504 44. The "Exiscan patch" is now merged into the mainline Exim source.
2506 45. Sometimes the final signoff response after QUIT could fail to get
2507 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2508 < 0 before doing a fflush(). This bug looks as though it goes back to the
2509 introduction of TLS in release 3.20, but "sometimes" must have been rare
2510 because the tests only now provoked it.
2512 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2513 (this can affect the format of dates).
2515 47. exim_tidydb, when checking for the continued existence of a message for
2516 which it has found a message-specific retry record, was not finding
2517 messages that were in split spool directories. Consequently, it was
2518 deleting retry records that should have stayed in existence.
2520 48. Steve fixed some bugs in eximstats.
2522 49. The SPA authentication driver was not abandoning authentication and moving
2523 on to the next authenticator when an expansion was forced to fail,
2524 contradicting the general specification for all authenticators. Instead it
2525 was generating a temporary error. It now behaves as specified.
2527 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2528 (the order specifies the preference for clients). The order is now AES256,
2529 AES128, 3DES, ARCFOUR128.
2531 51. Small patch to Sieve code - explicitly set From: when generating an
2534 52. Exim crashed if a remote delivery caused a very long error message to be
2535 recorded - for instance if somebody sent an entire SpamAssassin report back
2536 as a large number of 550 error lines. This bug was coincidentally fixed by
2537 increasing the size of one of Exim's internal buffers (big_buffer) that
2538 happened as part of the Exiscan merge. However, to be on the safe side, I
2539 have made the code more robust (and fixed the comments that describe what
2542 53. Now that there can be additional text after "Completed" in log lines (if
2543 the queue_time_overall log selector is set), a one-byte patch to exigrep
2544 was needed to allow it to recognize "Completed" as not the last thing in
2547 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2548 patch that reportedly fixes this has been added. I am not expert enough to
2549 create a test for it. This is what the patch creator wrote:
2551 "I found a little strange behaviour of ldap code when working with
2552 Windows 2003 AD Domain, where users was placed in more than one
2553 Organization Units. When I tried to give exim partial DN, the exit code
2554 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2555 But simultaneously result of request was absolutely normal ldap result,
2556 so I produce this patch..."
2558 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2559 REFERENCE, so I have modified the code to exclude the patch when that macro
2562 55. Some experimental protocols are using DNS PTR records for new purposes. The
2563 keys for these records are domain names, not reversed IP addresses. The
2564 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2565 leaves it alone. Component reversal etc. now happens only for IP addresses.
2568 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2570 57. Double the size of the debug message buffer (to 2048) so that more of very
2571 long debug lines gets shown.
2573 58. The exicyclog utility now does better if the number of log files to keep
2574 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2576 59. Two changes related to the smtp_active_hostname option:
2578 (1) $smtp_active_hostname is now available as a variable.
2579 (2) The default for smtp_banner uses $smtp_active_hostname instead
2580 of $primary_hostname.
2582 60. The host_aton() function is supposed to be passed a string that is known
2583 to be a valid IP address. However, in the case of IPv6 addresses, it was
2584 not checking this. This is a hostage to fortune. Exim now panics and dies
2585 if the condition is not met. A case was found where this could be provoked
2586 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2587 components; fortuitously, this particular loophole had already been fixed
2588 by change 4.50/55 above.
2590 If there are any other similar loopholes, the new check in host_aton()
2591 itself should stop them being exploited. The report I received stated that
2592 data on the command line could provoke the exploit when Exim was running as
2593 exim, but did not say which command line option was involved. All I could
2594 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2595 running as the user.
2598 61. There was a buffer overflow vulnerability in the SPA authentication code
2599 (which came originally from the Samba project). I have added a test to the
2600 spa_base64_to_bits() function which I hope fixes it.
2603 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2604 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2606 63. The daemon start-up calls getloadavg() while still root for those OS that
2607 need the first call to be done as root, but it missed one case: when
2608 deliver_queue_load_max is set with deliver_drop_privilege. This is
2609 necessary for the benefit of the queue runner, because there is no re-exec
2610 when deliver_drop_privilege is set.
2612 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2613 This has been fixed.
2615 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2616 was in use, was not putting the data itself into the right store pool;
2617 consequently, it could be overwritten for a subsequent message in the same
2618 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2621 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2623 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2624 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2625 the function were treating the return as a boolean value, which happened to
2626 work because 0=false and not-0=true, but is not correct code.
2628 68. The host_aton() function was not handling scoped IPv6 addresses (those
2629 with, for example, "%eth0" on the end) correctly.
2631 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2632 negated items (that is, ~something) in unsigned ints. Some compilers
2633 apparently mutter when there is no cast.
2635 70. If an address verification called from an ACL failed, and did not produce a
2636 user-specific message (i.e. there was only a "system" message), nothing was
2637 put in $acl_verify_message. In this situation, it now puts the system
2640 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2641 session; change 4.31/43 added the unwanted input to the log line - except
2642 that it did not do this in the start of session case. It now does.
2644 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2645 This is wrong and can cause the other end to generate a synchronization
2646 error if it is another Exim or anything else that does the synchronization
2647 check. A QUIT command is no longer sent after a timeout.
2649 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2650 during host lookups.
2652 74. The defer_ok option of callout verification was not working if it was used
2653 when verifying addresses in header lines, that is, for this case:
2655 verify = header_sender/callout=defer_ok
2657 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2658 those file descriptors could be used for SMTP connections. If anything
2659 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2660 could be sent to the SMTP client, causing chaos. The daemon now opens
2661 stdin, stdout, and stderr to /dev/null when it puts itself into the
2664 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2665 log by default. The user can override this with suitable Perl magic.
2667 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2668 the log message when discard triggers, was not working for the DATA ACL or
2669 for the non-SMTP ACL.
2671 78. Error message wording change in sieve.c.
2673 79. If smtp_accept_max_per_host was set, the number of connections could be
2674 restricted to fewer than expected, because the daemon was trying to set up
2675 a new connection before checking whether the processes handling previous
2676 connections had finished. The check for completed processes is now done
2677 earlier. On busy systems, this bug wouldn't be noticed because something
2678 else would have woken the daemon, and it would have reaped the completed
2681 80. If a message was submitted locally by a user whose login name contained one
2682 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2683 It caused a spool format error. I have fixed the spool reading code. A
2684 related problem was that the "from" clause in the Received: line became
2685 illegal because of the space(s). It is now covered by ${quote_local_part.
2687 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2690 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2692 83. Patch from Sieve maintainer to fix unterminated string problem in
2693 "vacation" handling.
2695 84. Some minor changes to the Linux configuration files to help with other
2696 OS variants using glibc.
2698 85. One more patch for Sieve to update vacation handling to latest spec.
2701 ----------------------------------------------------
2702 See the note above about the 4.44 and 4.50 releases.
2703 ----------------------------------------------------
2709 1. Change 4.43/35 introduced a bug that caused file counts to be
2710 incorrectly computed when quota_filecount was set in an appendfile
2713 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2714 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2717 3. Give more explanation in the error message when the command for a transport
2718 filter fails to execute.
2720 4. There are several places where Exim runs a non-Exim command in a
2721 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2722 was being done only for the command run by the queryprogram router. It is
2723 now done for all such subprocesses. The other cases are: ${run, transport
2724 filters, and the commands run by the lmtp and pipe transports.
2726 5. Some older OS have a limit of 256 on the maximum number of file
2727 descriptors. Exim was using setrlimit() to set 1000 as a large value
2728 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2729 systems. I've change it so that if it can't get 1000, it tries for 256.
2731 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2732 was an oversight, and furthermore, ever since the addition of extra
2733 controls (e.g. 4.43/32), the checks on when to allow different forms of
2734 "control" were broken. There should now be diagnostics for all cases when a
2735 control that does not make sense is encountered.
2737 7. $recipients is now available in the predata ACL (oversight).
2739 8. Tidy the search cache before the fork to do a delivery from a message
2740 received from the command line. Otherwise the child will trigger a lookup
2741 failure and thereby defer the delivery if it tries to use (for example) a
2742 cached ldap connection that the parent has called unbind on.
2744 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2745 of $address_data from the recipient verification was clobbered by the
2746 sender verification.
2748 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2749 was its contents. (It was OK if the option was not defined at all.)
2751 11. A "Completed" log line is now written for messages that are removed from
2752 the spool by the -Mrm option.
2754 12. $host_address is now set to the target address during the checking of
2755 ignore_target_hosts.
2757 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2758 being passed; this would have caused $sender_host_name to have been used if
2759 matching the list had actually called for a host name (not very likely,
2760 since this list is usually IP addresses). A host name is now passed as
2763 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2764 code that specifies a non-restarting handler (typically sigaction() in
2765 modern systems) in an attempt to fix a rare and obscure crash bug.
2767 15. Narrowed the window for a race in the daemon that could cause it to ignore
2768 SIGCHLD signals. This is not a major problem, because they are used only to
2769 wake it up if nothing else does.
2771 16. A malformed maildirsize file could cause Exim to calculate negative values
2772 for the mailbox size or file count. Odd effects could occur as a result.
2773 The maildirsize information is now recalculated if the size or filecount
2776 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2777 support for a long time. Removed HAVE_SYS_VFS_H.
2779 18. Updated exipick to current release from John Jetmore.
2781 19. Allow an empty sender to be matched against a lookup in an address list.
2782 Previously the only cases considered were a regular expression, or an
2785 20. Exim went into a mad DNS lookup loop when doing a callout where the
2786 host was specified on the transport, if the DNS lookup yielded more than
2789 21. The RFC2047 encoding function was originally intended for short strings
2790 such as real names; it was not keeping to the 75-character limit for
2791 encoded words that the RFC imposes. It now respects the limit, and
2792 generates multiple encoded words if necessary. To be on the safe side, I
2793 have increased the buffer size for the ${rfc2047: expansion operator from
2796 22. Failure to deliver a bounce message always caused it to be frozen, even if
2797 there was an errors_to setting on the router. The errors_to setting is now
2800 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2801 canonical form (fully expanded) before being placed in
2802 $sender_host_address.
2804 24. Updated eximstats to version 1.33
2806 25. Include certificate and key file names in error message when GnuTLS fails
2807 to set them up, because the GnuTLS error message doesn't include the name
2808 of the failing file when there is a problem reading it.
2810 26. Expand error message when OpenSSL has problems setting up cert/key files.
2813 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2814 (this can affect the format of dates).
2816 28. exim_tidydb, when checking for the continued existence of a message for
2817 which it has found a message-specific retry record, was not finding
2818 messages that were in split spool directories. Consequently, it was
2819 deleting retry records that should have stayed in existence.
2821 29. eximstats updated to version 1.35
2822 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2823 1.35 - bugfix such that pie charts by volume are generated correctly
2825 30. The SPA authentication driver was not abandoning authentication and moving
2826 on to the next authenticator when an expansion was forced to fail,
2827 contradicting the general specification for all authenticators. Instead it
2828 was generating a temporary error. It now behaves as specified.
2830 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2831 (the order specifies the preference for clients). The order is now AES256,
2832 AES128, 3DES, ARCFOUR128.
2834 31. Small patch to Sieve code - explicitly set From: when generating an
2837 32. Exim crashed if a remote delivery caused a very long error message to be
2838 recorded - for instance if somebody sent an entire SpamAssassin report back
2839 as a large number of 550 error lines. This bug was coincidentally fixed by
2840 increasing the size of one of Exim's internal buffers (big_buffer) that
2841 happened as part of the Exiscan merge. However, to be on the safe side, I
2842 have made the code more robust (and fixed the comments that describe what
2845 33. Some experimental protocols are using DNS PTR records for new purposes. The
2846 keys for these records are domain names, not reversed IP addresses. The
2847 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2848 leaves it alone. Component reversal etc. now happens only for IP addresses.
2851 34. The host_aton() function is supposed to be passed a string that is known
2852 to be a valid IP address. However, in the case of IPv6 addresses, it was
2853 not checking this. This is a hostage to fortune. Exim now panics and dies
2854 if the condition is not met. A case was found where this could be provoked
2855 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2856 components; fortuitously, this particular loophole had already been fixed
2857 by change 4.50/55 or 4.44/33 above.
2859 If there are any other similar loopholes, the new check in host_aton()
2860 itself should stop them being exploited. The report I received stated that
2861 data on the command line could provoke the exploit when Exim was running as
2862 exim, but did not say which command line option was involved. All I could
2863 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2864 running as the user.
2867 35. There was a buffer overflow vulnerability in the SPA authentication code
2868 (which came originally from the Samba project). I have added a test to the
2869 spa_base64_to_bits() function which I hope fixes it.
2872 36. The daemon start-up calls getloadavg() while still root for those OS that
2873 need the first call to be done as root, but it missed one case: when
2874 deliver_queue_load_max is set with deliver_drop_privilege. This is
2875 necessary for the benefit of the queue runner, because there is no re-exec
2876 when deliver_drop_privilege is set.
2878 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2879 was in use, was not putting the data itself into the right store pool;
2880 consequently, it could be overwritten for a subsequent message in the same
2881 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2884 38. Sometimes the final signoff response after QUIT could fail to get
2885 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2886 < 0 before doing a fflush(). This bug looks as though it goes back to the
2887 introduction of TLS in release 3.20, but "sometimes" must have been rare
2888 because the tests only now provoked it.
2894 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2895 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2896 Now it returns an integer. A number of calls were still expecting a T/F
2897 return. Fortuitously, in all cases, the tests worked in OK situations,
2898 which is the norm. However, things would have gone wrong on any write
2899 failures on the smtp file descriptor. This function is used when sending
2900 messages over SMTP and also when doing verify callouts.
2902 2. When Exim is called to do synchronous delivery of a locally submitted
2903 message (the -odf or -odi options), it no longer closes stderr before doing
2906 3. Implemented the mua_wrapper option.
2908 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2910 5. Implemented the functions header_remove(), header_testname(),
2911 header_add_at_position(), and receive_remove_recipient(), and exported them
2914 6. If an ACL "warn" statement specified the addition of headers, Exim already
2915 inserted X-ACL-Warn: at the start if there was no header name. However, it
2916 was not making this test for the second and subsequent header lines if
2917 there were newlines in the string. This meant that an invalid header could
2918 be inserted if Exim was badly configured.
2920 7. Allow an ACL "warn" statement to add header lines at the start or after all
2921 the Received: headers, as well as at the end.
2923 8. Added the rcpt_4xx retry error code.
2925 9. Added postmaster_mailfrom=xxx to callout verification option.
2927 10. Added mailfrom=xxxx to the callout verification option, for verify=
2930 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2931 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2933 12. Inserted some casts to stop certain compilers warning when using pointer
2934 differences as field lengths or precisions in printf-type calls (mostly
2935 affecting debugging statements).
2937 13. Added optional readline() support for -be (dynamically loaded).
2939 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2940 same clock tick as a message's arrival, so that its received time was the
2941 same as the "first fail" time on the retry record, and that message
2942 remained on the queue past the ultimate address timeout, every queue runner
2943 would try a delivery (because it was past the ultimate address timeout) but
2944 after another failure, the ultimate address timeout, which should have then
2945 bounced the address, did not kick in. This was a "< instead of <=" error;
2946 in most cases the first failure would have been in the next clock tick
2947 after the received time, and all would be well.
2949 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2950 being recognized when the domain list was tested by the match_domain
2951 condition in an expansion string.
2953 16. Added the ${str2b64: operator.
2955 17. Exim was always calling setrlimit() to set a large limit for the number of
2956 processes, without checking whether the existing limit was already
2957 adequate. (It did check for the limit on file descriptors.) Furthermore,
2958 errors from getrlimit() and setrlimit() were being ignored. Now they are
2959 logged to the main and panic logs, but Exim does carry on, to try to do its
2960 job under whatever limits there are.
2962 18. Imported PCRE 5.0.
2964 19. Trivial typo in log message " temporarily refused connection" (the leading
2967 20. If the log selector return_path_on_delivery was set and an address was
2968 redirected to /dev/null, the delivery process crashed because it assumed
2969 that a return path would always be set for a "successful" delivery. In this
2970 case, the whole delivery is bypassed as an optimization, and therefore no
2973 21. Internal re-arrangement: the function for sending a challenge and reading
2974 a response while authentication was assuming a zero-terminated challenge
2975 string. It's now changed to take a pointer and a length, to allow for
2976 binary data in such strings.
2978 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2980 23. Exim was not respecting finduser_retries when seeking the login of the
2981 uid under which it was called; it was always trying 10 times. (The default
2982 setting of finduser_retries is zero.) Also, it was sleeping after the final
2983 failure, which is pointless.
2985 24. Implemented tls_on_connect_ports.
2987 25. Implemented acl_smtp_predata.
2989 26. If the domain in control=submission is set empty, Exim assumes that the
2990 authenticated id is a complete email address when it generates From: or
2991 Sender: header lines.
2993 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2994 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2995 chown and chgrp in /bin and hostname in /usr/bin.
2997 28. Exim was keeping the "process log" file open after each use, just as it
2998 does for the main log. This opens the possibility of it remaining open for
2999 long periods when the USR1 signal hits a daemon. Occasional processlog
3000 errors were reported, that could have been caused by this. Anyway, it seems
3001 much more sensible not to leave this file open at all, so that is what now
3004 29. The long-running daemon process does not normally write to the log once it
3005 has entered its main loop, and it closes the log before doing so. This is
3006 so that log files can straightforwardly be renamed and moved. However,
3007 there are a couple of unusual error situations where the daemon does write
3008 log entries, and I had neglected to close the log afterwards.
3010 30. The text of an SMTP error response that was received during a remote
3011 delivery was being truncated at 512 bytes. This is too short for some of
3012 the long messages that one sometimes sees. I've increased the limit to
3015 31. It is now possible to make retry rules that apply only when a message has a
3016 specific sender, in particular, an empty sender.
3018 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3019 it possible to be selective about when SMTP synchronization is enforced.
3021 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3023 32. Implemented hosts_connection_nolog.
3025 33. Added an ACL for QUIT.
3027 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3030 35. Added mailbox_size and mailbox_filecount to appendfile.
3032 36. Added control = no_multiline_responses to ACLs.
3034 37. There was a bug in the logic of the code that waits for the clock to tick
3035 in the case where the clock went backwards by a substantial amount such
3036 that the microsecond fraction of "now" was more than the microsecond
3037 fraction of "then" (but the whole seconds number was less).
3039 38. Added support for the libradius Radius client library this is found on
3040 FreeBSD (previously only the radiusclient library was supported).
3046 1. When certain lookups returned multiple values in the form name=value, the
3047 quoting of the values was not always being done properly. Specifically:
3048 (a) If the value started with a double quote, but contained no whitespace,
3050 (b) If the value contained whitespace other than a space character (i.e.
3051 tabs or newlines or carriage returns) it was not quoted.
3052 This fix has been applied to the mysql and pgsql lookups by writing a
3053 separate quoting function and calling it from the lookup code. The fix
3054 should probably also be applied to nisplus, ibase and oracle lookups, but
3055 since I cannot test any of those, I have not disturbed their existing code.
3057 2. A hit in the callout cache for a specific address caused a log line with no
3058 reason for rejecting RCPT. Now it says "Previous (cached) callout
3059 verification failure".
3061 3. There was an off-by-one bug in the queryprogram router. An over-long
3062 return line was truncated at 256 instead of 255 characters, thereby
3063 overflowing its buffer with the terminating zero. As well as fixing this, I
3064 have increased the buffer size to 1024 (and made a note to document this).
3066 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3067 when Exim is waiting for an SMTP response from a remote server, Exim
3068 restarts its select() call on the socket, thereby resetting its timeout.
3069 This is not a problem when such interrupts are rare. Somebody set up a cron
3070 job to run exiwhat every 2 minutes, which is less than the normal select()
3071 timeout (5 or 10 minutes). This meant that the select() timeout never
3072 kicked in because it was always reset. I have fixed this by comparing the
3073 time when an interrupt arrives with the time at the start of the first call
3074 to select(). If more time than the timeout has elapsed, the interrupt is
3075 treated as a timeout.
3077 5. Some internal re-factoring in preparation for the addition of Sieve
3078 extensions (by MH). In particular, the "personal" test is moved to a
3079 separate function, and given an option for scanning Cc: and Bcc: (which is
3080 not set for Exim filters).
3082 6. When Exim created an email address using the login of the caller as the
3083 local part (e.g. when creating a From: or Sender: header line), it was not
3084 quoting the local part when it contained special characters such as @.
3086 7. Installed new OpenBSD configuration files.
3088 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3089 try to make them clearer.
3091 9. Callout options, other than the timeout value, were being ignored when
3092 verifying sender addresses in header lines. For example, when using
3094 verify = header_sender/callout=no_cache
3096 the cache was (incorrectly) being used.
3098 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3099 only the Cygwin environment.
3101 11. When return_path_on_delivery was set as a log selector, if different remote
3102 addresses in the same message used different return paths and parallel
3103 remote delivery occurred, the wrong values would sometimes be logged.
3104 (Whenever a remote delivery process finished, the return path value from
3105 the most recently started remote delivery process was logged.)
3107 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3108 lines when AUTH and/or TLS are in use. This is the "received protocol"
3109 field. Exim used to use "asmtp" for authenticated SMTP, without any
3110 indication (in the protocol name) for TLS use. Now it follows the RFC and
3111 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3112 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3113 names appear in log lines as well as in Received: header lines.
3115 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3116 extensions, and comparison tests, and to fix some bugs.
3118 14. Changes to the "personal" filter test:
3120 (1) The test was buggy in that it was just doing the equivalent of
3121 "contains" tests on header lines. For example, if a user's address was
3122 anne@some.where, the "personal" test would incorrectly be true for
3124 To: susanne@some.where
3126 This test is now done by extracting each address from the header in turn,
3127 and checking the entire address. Other tests that are part of "personal"
3128 are now done using regular expressions (for example, to check local parts
3129 of addresses in From: header lines).
3131 (2) The list of non-personal local parts in From: addresses has been
3132 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3133 taken from the Sieve specification recommendations.
3135 (3) If the message contains any header line starting with "List-" it is
3136 treated as non-personal.
3138 (4) The test for "circular" in the Subject: header line has been removed
3139 because it now seems ill-conceived.
3141 15. Minor typos in src/EDITME comments corrected.
3143 16. Installed latest exipick from John Jetmore.
3145 17. If headers_add on a router specified a text string that was too long for
3146 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3147 of string_sprintf() is now avoided.
3149 18. $message_body_size was not set (it was always zero) when running the DATA
3150 ACL and the local_scan() function.
3152 19. For the "mail" command in an Exim filter, no default was being set for
3153 the once_repeat time, causing a random time value to be used if "once" was
3154 specified. (If the value happened to be <= 0, no repeat happened.) The
3155 default is now 0s, meaning "never repeat". The "vacation" command was OK
3156 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3157 (I found it when inspecting the code).
3159 20. There is now an overall timeout for performing a callout verification. It
3160 defaults to 4 times the callout timeout, which applies to individual SMTP
3161 commands during the callout. The overall timeout applies when there is more
3162 than one host that can be tried. The timeout is checked before trying the
3163 next host. This prevents very long delays if there are a large number of
3164 hosts and all are timing out (e.g. when the network connections are timing
3165 out). The value of the overall timeout can be changed by specifying an
3166 additional sub-option for "callout", called "maxwait". For example:
3168 verify = sender/callout=5s,maxwait=20s
3170 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3171 to the end before writing, but this should make it even safer).
3173 22. Exim was forgetting that it had advertised PIPELINING for the second and
3174 subsequent messages on an SMTP connection. It was also not resetting its
3175 memory on STARTTLS and an internal HELO.
3177 23. When Exim logs an SMTP synchronization error within a session, it now
3178 records whether PIPELINING has been advertised or not.
3180 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3181 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3182 rather than long int.
3184 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3186 26. Added the never_mail option to autoreply.
3192 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3193 crash if the getsockname() call failed; this can happen if a connection is
3194 closed very soon after it is established. The problem was simply in the
3195 order in which certain operations were done, causing Exim to try to write
3196 to the SMTP stream before it had set up the file descriptor. The bug has
3197 been fixed by making things happen in the correct order.
3203 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3204 before the connection was closed, thus losing the rejection response.
3206 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3207 some early Solaris releases, but causes trouble in current releases where
3208 socklen_t is defined.
3210 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3213 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3216 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3217 that releasing the top of it at the end releases what was used for sub-
3218 expansions (unless the block got too big). However, discard this block if
3219 the first thing is a variable or header, so that we can use its block when
3220 it is dynamic (useful for very large $message_headers, for example).
3222 6. Lookups now cache *every* query, not just the most recent. A new, separate
3223 store pool is used for this. It can be recovered when all lookup caches are
3224 flushed. Lookups now release memory at the end of their result strings.
3225 This has involved some general refactoring of the lookup sources.
3227 7. Some code has been added to the store_xxx() functions to reduce the amount
3228 of flapping under certain conditions.
3230 8. log_incoming_interface used to affect only the <= reception log lines. Now
3231 it causes the local interface and port to be added to several more SMTP log
3232 lines, for example "SMTP connection from", and rejection lines.
3234 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3236 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3238 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3239 could be overwritten at the end of the current message (or the start of a
3240 new message if it was set in a HELO ACL). The value is now preserved for
3241 the duration of the SMTP connection.
3243 12. If a transport had a headers_rewrite setting, and a matching header line
3244 contained an unqualified address, that address was qualified, even if it
3245 did not match any rewriting rules. The underlying bug was that the values
3246 of the flags that permit the existence of unqualified sender and recipient
3247 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3248 non-local messages, and by -bnq for local messages) were not being
3249 preserved with the message after it was received.
3251 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3252 "next input=" as part of the text comprising the host identity instead of
3253 the correct text. The code was using the same buffer for two different
3254 strings. However, depending on which order the printing function evaluated
3255 its arguments, the bug did not always show up. Under Linux, for example, my
3256 test suite worked just fine.
3258 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3259 doesn't work with some older versions of Perl. It has been changed to "my",
3260 which in any case is probably the better facility to use.
3262 15. A really picky compiler found some instances of statements for creating
3263 error messages that either had too many or two few arguments for the format
3266 16. The size of the buffer for calls to the DNS resolver has been increased
3267 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3268 for addresses that have a lot of PTR records. This alleviates a problem; it
3269 does not fully solve it.
3271 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3272 buffer now truncates the list and logs the incident, which is the same
3273 action as happens when Exim is looking up a host name and its aliases.
3274 Previously in this situation something unpredictable would happen;
3275 sometimes it was "internal error: store_reset failed".
3277 18. If a server dropped the connection unexpectedly when an Exim client was
3278 using GnuTLS and trying to read a response, the client delivery process
3279 crashed while trying to generate an error log message.
3281 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3282 string, for example:
3284 warn message = H1: something\nH2: something
3286 the text was added as a single header line from Exim's point of view
3287 though it ended up OK in the delivered message. However, searching for the
3288 second and subsequent header lines using $h_h2: did not work. This has been
3289 fixed. Similarly, if a system filter added multiple headers in this way,
3290 the routers could not see them.
3292 20. Expanded the error message when iplsearch is called with an invalid key to
3293 suggest using net-iplsearch in a host list.
3295 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3296 ACLs are no longer actually imposed (and a message to that effect is
3299 22. If a "gecos" field in a passwd entry contained escaped characters, in
3300 particular, if it contained a \" sequence, Exim got it wrong when building
3301 a From: or a Sender: header from that name. A second bug also caused
3302 incorrect handling when an unquoted " was present following a character
3303 that needed quoting.
3305 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3306 was not being matched caselessly.
3308 24. Arranged for all hyphens in the exim.8 source to be escaped with
3311 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3312 properly. Recipient callout cache records were still being keyed to include
3313 the sender, even when use_sender was set false. This led to far more
3314 callouts that were necessary. The sender is no longer included in the key
3315 when use_sender is false.
3317 26. Added "control = submission" modifier to ACLs.
3319 27. Added the ${base62d: operator to decode base 62 numbers.
3321 28. dnsdb lookups can now access SRV records.
3323 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3324 the configuration file.
3326 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3327 (-v) mode. This makes the output for a verbose queue run more intelligible.
3329 31. Added a use_postmaster feature to recipient callouts.
3331 32. Added the $body_zerocount variable, containing the number of binary zero
3332 bytes in the message body.
3334 33. The time of last modification of the "new" subdirectory is now used as the
3335 "mailbox time last read" when there is a quota error for a maildir
3338 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3340 35. Added +ignore_unknown as a special item in host lists.
3342 36. Code for decoding IPv6 addresses in host lists is now included, even if
3343 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3344 address was recognized as an IP address, but was then not correctly decoded
3345 into binary, causing unexpected and incorrect effects when compared with
3352 1. Very minor rewording of debugging text in manualroute to say "list of
3353 hosts" instead of "hostlist".
3355 2. If verify=header_syntax was set, and a header line with an unqualified
3356 address (no domain) and a large number of spaces between the end of the
3357 name and the colon was received, the reception process suffered a buffer
3358 overflow, and (when I tested it) crashed. This was caused by some obsolete
3359 code that should have been removed. The fix is to remove it!
3361 3. When running in the test harness, delay a bit after writing a bounce
3362 message to get a bit more predictability in the log output.
3364 4. Added a call to search_tidyup() just before forking a reception process. In
3365 theory, someone could use a lookup in the expansion of smtp_accept_max_
3366 per_host which, without the tidyup, could leave open a database connection.
3368 5. Added the variables $recipient_data and $sender_data which get set from a
3369 lookup success in an ACL "recipients" or "senders" condition, or a router
3370 "senders" option, similar to $domain_data and $local_part_data.
3372 6. Moved the writing of debug_print from before to after the "senders" test
3375 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3376 problems for message scanning, either using a data ACL, or using
3377 local_scan() because the Received: header was not generated till after they
3378 were called (in order to set the time as the time of reception completion).
3379 I have revised the way this works. The header is now generated after the
3380 body is received, but before the ACL or local_scan() are called. After they
3381 are run, the timestamp in the header is updated.
3387 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3388 before starting a queue runner without re-exec. This happened only when
3389 deliver_drop_privilege was set or when the Exim user was set to root. The
3390 effect of the bug was that timeouts during subsequent deliveries caused
3391 crashes instead of being properly handled. The handler is now left at its
3392 default (and expected) setting.
3394 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3395 message, again when deliver_drop_privilege is set or Exim is run as root.
3396 The bug described in (1) was not present in this case, but the tidying up
3397 of the other signals was missing. I have made the two cases consistent.
3399 3. The ignore_target_hosts setting on a manualroute router was being ignored
3400 for hosts that were looked up using the /MX notation.
3402 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3405 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3406 operated on the sender address. After changing the $sender_address to <>
3407 for the sender address verify, Exim was re-instated it as the original
3408 (before rewriting) address, but remembering that it had rewritten it, so it
3409 wasn't rewriting it again. This bug also had the effect of breaking the
3410 sender address verification caching when the sender address was rewritten.
3412 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3413 This has been changed so that if the ip literal address matches
3414 ignore_target_hosts, the router declines.
3416 7. Added expansion conditions match_domain, match_address, and match_local_
3417 part (NOT match_host).
3419 8. The placeholder for the Received: header didn't have a length field set.
3421 9. Added code to Exim itself and to exim_lock to test for a specific race
3422 condition that could lead to file corruption when using MBX delivery. The
3423 issue is with the lockfile that is created in /tmp. If this file is removed
3424 after a process has opened it but before that process has acquired a lock,
3425 there is the potential for a second process to recreate the file and also
3426 acquire a lock. This could lead to two Exim processes writing to the file
3427 at the same time. The added code performs the same test as UW imapd; it
3428 checks after acquiring the lock that its file descriptor still refers to
3429 the same named file.
3431 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3432 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3435 11. Added the smtp_active_hostname option. If used, this will typically be made
3436 to depend on the incoming interface address. Because $interface_address is
3437 not set up until the daemon has forked a reception process, error responses
3438 that can happen earlier (such as "too many connections") no longer contain
3441 12. If an expansion in a condition on a "warn" statement fails because a lookup
3442 defers, the "warn" statement is abandoned, and the next ACL statement is
3443 processed. Previously this caused the whole ACL to be aborted.
3445 13. Added the iplsearch lookup type.
3447 14. Added ident_timeout as a log selector.
3449 15. Added tls_certificate_verified as a log selector.
3451 16. Added a global option tls_require_ciphers (compare the smtp transport
3452 option of the same name). This controls incoming TLS connections.
3454 17. I finally figured out how to make tls_require_ciphers do a similar thing
3455 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3456 before starting the TLS session.
3458 18. Tabs are now shown as \t in -bP output.
3460 19. If the log selector return_path_on_delivery was set, Exim crashed when
3461 bouncing a message because it had too many Received: header lines.
3463 20. If two routers both had headers_remove settings, and the first one included
3464 a superfluous trailing colon, the final name in the first list and the
3465 first name in the second list were incorrectly joined into one item (with a
3466 colon in the middle).
3472 1. Added -C and -D options to the exinext utility, mainly to make it easier
3473 to include in the automated testing, but these could be helpful when
3474 multiple configurations are in use.
3476 2. The exinext utility was not formatting the output nicely when there was
3477 an alternate port involved in the retry record key, nor when there was a
3478 message id as well (for retries that were specific to a specific message
3479 and a specific host). It was also confused by IPv6 addresses, because of
3480 the additional colons they contain. I have fixed the IPv4 problem, and
3481 patched it up to do a reasonable job for IPv6.
3483 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3484 delivery, the log line now contains "pipelined" if PIPELINING was used.
3486 4. An SMTP transport process used to panic and die if the bind() call to set
3487 an explicit outgoing interface failed. This has been changed; it is now
3488 treated in the same way as a connect() failure.
3490 5. A reference to $sender_host_name in the part of a conditional expansion
3491 that was being skipped was still causing a DNS lookup. This no longer
3494 6. The def: expansion condition was not recognizing references to header lines
3495 that used bh_ and bheader_.
3497 7. Added the _cache feature to named lists.
3499 8. The code for checking quota_filecount in the appendfile transport was
3500 allowing one more file than it should have been.
3502 9. For compatibility with Sendmail, the command line option
3510 and sets the incoming protocol and host name (for trusted callers). The
3511 host name and its colon can be omitted when only the protocol is to be set.
3512 Note the Exim already has two private options, -pd and -ps, that refer to
3513 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3514 "s", but I don't think that's a major issue.
3516 10. A number of refactoring changes to the code, none of which should affect
3519 (a) The number of logging options was getting close to filling up the
3520 32-bit word that was used as a bit map. I have split them into two classes:
3521 those that are passed in the argument to log_write(), and those that are
3522 only ever tested independently outside of that function. These are now in
3523 separate 32-bit words, so there is plenty of room for expansion again.
3524 There is no change in the user interface or the logging behaviour.
3526 (b) When building, for example, log lines, the code previously used a
3527 macro that called string_cat() twice, in order to add two strings. This is
3528 not really sufficiently general. Furthermore, there was one instance where
3529 it was actually wrong because one of the argument was used twice, and in
3530 one call a function was used. (As it happened, calling the function twice
3531 did not affect the overall behaviour.) The macro has been replaced by a
3532 function that can join an arbitrary number of extra strings onto a growing
3535 (c) The code for expansion conditions now uses a table and a binary chop
3536 instead of a serial search (which was left over from when there were very
3537 few conditions). Also, it now recognizes conditions like "pam" even when
3538 the relevant support is not compiled in: a suitably worded error message is
3539 given if an attempt is made to use such a condition.
3541 11. Added ${time_interval:xxxxx}.
3543 12. A bug was causing one of the ddress fields not to be passed back correctly
3544 from remote delivery subprocesses. The field in question was not being
3545 subsequently used, so this caused to problems in practice.
3547 13. Added new log selectors queue_time and deliver_time.
3549 14. Might have fixed a bug in maildirsizefile handling that threw up
3550 "unexpected character" debug warnings, and recalculated the data
3551 unnecessarily. In any case, I expanded the warning message to give more
3554 15. Added the message "Restricted characters in address" to the statements in
3555 the default ACL that block characters like @ and % in local parts.
3557 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3558 Three changes have been made:
3560 (a) There was a serious bug; a negative response to MAIL caused the whole
3561 recipient domain to be cached as invalid, thereby blocking all messages
3562 to all local parts at the same domain, from all senders. This bug has
3563 been fixed. The domain is no longer cached after a negative response to
3564 MAIL if the sender used is not empty.
3566 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3569 (c) A new callout option, "use_sender" has been added for people who want
3570 the modified behaviour.
3576 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3579 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3580 indeed breaks things for older releases.
3582 3. Added additional logging to the case where there is a problem reading data
3583 from a filter that is running in a subprocess using a pipe, in order to
3584 try to track down a specific problem.
3586 4. Testing facility fudge: when running in the test harness and attempting
3587 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3588 getting "No route to host". Convert this to a timeout.
3590 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3593 6. Some OS don't have socklen_t but use size_t instead. This affects the
3594 fifth argument of getsockopt() amongst other things. This is now
3595 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3596 can be set for individual OS. I have set it for SunOS5, OSF1, and
3597 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3598 some earlier ones do not.
3600 7. Change 4.30/15 was not doing the test caselessly.
3602 8. The standard form for an IPv6 address literal was being rejected by address
3603 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3604 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3605 this, as well as the form without the "IPv6" on the front (but only when
3606 address literals are enabled, of course).
3608 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3610 10. Exim crashed if a message with an empty sender address specified by -f
3611 encountered a router with an errors_to setting. This could be provoked only
3612 by a command such as
3616 where an empty string was supplied; "<>" did not hit this bug.
3618 11. Installed PCRE release 4.5.
3620 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3621 remained set. It is now erased.
3623 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3624 times from message ids (which are base 36 rather than the normal 62).
3626 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3627 were being counted as actual protocol errors, and logged if the log
3628 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3629 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3630 and DATA following a set of rejected RCPTs do not count as protocol errors.
3631 In other words, Exim assumes they were pipelined, though this may not
3632 actually be the case. Of course, in all cases the client gets an
3633 appropriate error code.
3635 15. If a lookup fails in an ACL condition, a message about the failure may
3636 be available; it is used if testing the ACL cannot continue, because most
3637 such messages specify what the cause of the deferral is. However, some
3638 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3639 that caused an old message to be retained and used if a later statement
3640 caused a defer, replacing the real cause of the deferral.
3642 16. If an IP address had so many PTR records that the DNS lookup buffer
3643 was not large enough to hold them, Exim could crash while trying to process
3644 the truncated data. It now detects and logs this case.
3646 17. Further to 4.21/58, another change has been made: if (and only if) the
3647 first line of a message (the first header line) ends with CRLF, a bare LF
3648 in a subsequent header line has a space inserted after it, so as not to
3649 terminate the header.
3651 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3652 unnecessarily, used atoi() instead of strtol(), and didn't check the
3653 termination when getting file sizes from file names by regex.
3655 19. Completely re-implemented the support for maildirsize files, in the light
3656 of a number of problems with the previous contributed implementation
3657 (4.30/29). In particular:
3659 . If the quota is zero, the maildirsize file is maintained, but no quota is
3662 . If the maildir directory does not exist, it is created before any attempt
3663 to write a maildirsize file.
3665 . The quota value in the file is just a cache; if the quota is changed in
3666 the transport, the new value overrides.
3668 . A regular expression is available for excluding directories from the
3671 20. The autoreply transport checks the characters in options that define the
3672 message's headers; it allows continued headers, but it was checking with
3673 isspace() after an embedded newline instead of explicitly looking for a
3676 21. If all the "regular" hosts to which an address was routed had passed their
3677 expiry times, and had not reached their retry times, the address was
3678 bounced, even if fallback hosts were defined. Now Exim should go on to try
3681 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3682 equivalent code in the SMTP transport. Some hosts send humungous responses
3683 to HELO/EHLO, more than 1024 it seems.
3685 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3686 gives compiler warnings in some environments. I've now done it "properly",
3689 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3690 (because of problems with the built-in one) was declared to return uschar *
3691 instead of char *, causing compiler failure.
3693 25. Fixed a file descriptor leak when processing alias/forward files.
3695 26. Fixed a minor format string issue in dbfn.c.
3697 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3699 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3700 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3703 29. When a sender address is verified, it is cached, to save repeating the test
3704 when there is more than one recipient in a message. However, when the
3705 verification involves a callout, it is possible for different callout
3706 options to be set for different recipients. It is too complicated to keep
3707 track of this in the cache, so now Exim always runs a verification when a
3708 callout is required, relying on the callout cache for the optimization.
3709 The overhead is duplication of the address routing, but this should not be
3712 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3713 to be verified with callout=postmaster, and the main callout worked but the
3714 postmaster check failed, the verification correctly failed. However, if a
3715 subsequent RCPT command asked for sender verification *without* the
3716 postmaster check, incorrect caching caused this verification also to fail,
3719 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3720 it was not caching the DNS options (qualify_single, search_parents) that
3721 were used when the lookup failed. A subsequent lookup with different
3722 options therefore always gave the same answer, though there were cases
3723 where it should not have. (Example: a "domains = !$mx_any" option on a
3724 dnslookup router: the "domains" option is always processed without any
3725 widening, but the router might have qualify_single set.) Now Exim uses the
3726 cached value only when the same options are set.
3728 32. Added John Jetmore's "exipick" utility to the distribution.
3730 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3731 than a timeout (e.g. a certificate is required, and is not provided), an
3732 Exim server now closes the connection immediately. Previously it waited for
3733 the client to close - but if the client is SSL, it seems that they each
3734 wait for each other, leading to a delay before one of them times out.
3736 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3737 maintained 0.8.x compatibility because I don't think many are using it, and
3738 it is clearly obsolete.
3740 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3743 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3744 client certificate was expired. A simple patch fixes this, though I don't
3745 understand the full logic of why the verify callback is called multiple
3748 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3749 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3750 which causes problems with some clients (such as the Certicom SSL Plus
3751 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3752 disables the coutermeasure allowing Eudora to connect."
3754 38. Exim was not checking that a write() to a log file succeeded. This could
3755 lead to Bad Things if a log got too big, in particular if it hit a file
3756 size limit. Exim now panics and dies if it cannot write to a log file, just
3757 as it does if it cannot open a log file.
3759 39. Modified OS/Makefile-Linux so that it now contains
3761 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3763 The two -D definitions ensure that Exim is compiled with large file
3764 support, which makes it possible to handle log files that are bigger than
3767 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3768 instance) a domain was checked against a named list that involved a lookup,
3769 causing $domain_data to be set, then another domain was checked against the
3770 same list, then the first domain was re-checked, the value of $domain_data
3771 after the final check could be wrong. In particular, if the second check
3772 failed, it could be set empty. This bug probably also applied to
3775 41. The strip_trailing_dot option was not being applied to the address given
3776 with the -f command-line option.
3778 42. The code for reading a message's header from the spool was incrementing
3779 $received_count, but never initializing it. This meant that the value was
3780 incorrect (doubled) while delivering a message in the same process in which
3781 it was received. In the most common configuration of Exim, this never
3782 happens - a fresh exec is done - but it can happen when
3783 deliver_drop_privilege is set.
3785 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3786 it now includes up to 150 characters of the unexpected data in the log
3789 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3790 and building data strings. The size of both of these buffers was 10 000
3791 bytes - far larger than anybody would *ever* want, thought I. Needless to
3792 say, somebody hit the limit. I have increased the maximum line length to
3793 20 000 and the maximum data length of concatenated lines to 100 000. I have
3794 also fixed two bugs, because there was no checking on these buffers. Tsk,
3795 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3796 buffer is too small.
3798 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3799 lsearch lookups. Now it does.
3801 46. When parsing a route_list item in a manualroute router, a fixed-length
3802 buffer was used for the list of hosts. I made this 1024 bytes long,
3803 thinking that nobody would ever have a list of hosts that long. Wrong.
3804 Somebody had a whole pile of complicated expansion conditions, and the
3805 string was silently truncated, leading to an expansion error. It turns out
3806 that it is easier to change to an unlimited length (owing to other changes
3807 that have happened since this code was originally written) than to build
3808 structure for giving a limitation error. The length of the item that
3809 expands into the list of hosts is now unlimited.
3811 47. The lsearch lookup could not handle data where the length of text line was
3812 more than 4095 characters. Such lines were truncated, leading to shortened
3813 data being returned. It should now handle lines of any length.
3815 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3816 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3819 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3820 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3821 no longer happens in comment lines. A list of replacements is now placed
3822 at the head of all of the source files, except those whose only change is
3823 to replace PERL_COMMAND in the very first #! line.
3825 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3826 messages on the queue, with a bottom-up merge sort, using code contributed
3827 by Michael Haardt. This should make operations like -bp somewhat faster on
3828 large queues. It won't affect queue runners, except when queue_run_in_order
3831 51. Installed eximstats 1.31 in the distribution.
3833 52. Added support for SRV lookups to the dnslookup router.
3835 53. If an ACL referred to $message_body or $message_body_end, the value was not
3836 reset for any messages that followed in the same SMTP session.
3838 54. The store-handling optimization for building very long strings was not
3839 differentiating between the different store pools. I don't think this
3840 actually made any difference in practice, but I've tidied it.
3842 55. While running the routers to verify a sender address, $sender_address
3843 was still set to the sender address. This is wrong, because when routing to
3844 send a bounce to the sender, it would be empty. Therefore, I have changed
3845 it so that, while verifying a sender address, $sender_address is set to <>.
3846 (There is no change to what happens when verifying a recipient address.)
3848 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3849 target A or AAAA records (if not already returned) without resetting the
3850 qualify_single or search_parents options of the DNS resolver. These are
3851 inappropriate in this case because the targets of MX and SRV records must
3852 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3853 target that, when qualified, matched something in the local domain. These
3854 two options are now turned off when doing these lookups.
3856 57. It seems that at least some releases of Reiserfs (which does not have the
3857 concept of a fixed number of inodes) returns zero and not -1 for the
3858 number of available inodes. This interacted badly with check_spool_inodes,
3859 which assumed that -1 was the "no such thing" setting. What I have done is
3860 to check that the total number of inodes is greater than zero before doing
3861 the test of how many are available.
3863 58. When a "warn" ACL statement has a log_message modifier, the message is
3864 remembered, and not repeated. This is to avoid a lot of repetition when a
3865 message has many recipients that cause the same warning to be written.
3866 Howewer, Exim was preserving the list of already written lines for an
3867 entire SMTP session, which doesn't seem right. The memory is now reset if a
3868 new message is started.
3870 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3871 debugging output unless log_rewrite was also set.
3873 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3874 of (int)(handle) when we know that handle contains (void *)(-1).
3876 61. The Exim daemon panic-logs an error return when it closes the incoming
3877 connection. However "connection reset by peer" seems to be common, and
3878 isn't really an error worthy of noting specially, so that particular error
3881 62. When Exim is trying to find all the local interfaces, it used to panic and
3882 die if the ioctl to get the interface flags failed. However, it seems that
3883 on at least one OS (Solaris 9) it is possible to have an interface that is
3884 included in the list of interfaces, but for which you get a failure error
3885 for this call. This happens when the interface is not "plumbed" into a
3886 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3887 failure of the "get flags" call assumes that the interface is down.
3889 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3890 makes life easier for people who are doing arithmetic on fields extracted
3891 from dates, where you often get leading zeros that should not be
3892 interpreted as octal.
3894 64. Added qualify_domain to the redirect router, to override the global
3897 65. If a pathologically long header line contained very many addresses (the
3898 report of this problem mentioned 10 000) and each of them was rewritten,
3899 Exim could use up a very large amount of memory. (It kept on making new
3900 copies of the header line as it rewrote, and never released the old ones.)
3901 At the expense of a bit more processing, the header rewriting function has
3902 been changed so that it no longer eats memory in this way.
3904 66. The generation of the Received: header has been moved from the time that a
3905 message starts to be received, to the time that it finishes. The timestamp
3906 in the Received: header should now be very close to that of the <= log
3907 line. There are two side-effects of this change:
3909 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3910 logged header lines no longer include the local Received: line, because
3911 it has not yet been created. The same applies to a copy of the message
3912 that is returned to a non-SMTP sender when a message is rejected.
3914 (b) When a filter file is tested using -bf, no additional Received: header
3915 is added to the test message. After some thought, I decided that this
3918 This change does not affect the value of $received_for. It is still set
3919 after address rewriting, but before local_scan() is called.
3921 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3923 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3924 gave an unhelpful panic error message, and a defer error. I have managed to
3925 change this behaviour so that it now rejects any supplied certificate,
3926 which seems right, as the list of acceptable certificates is empty.
3928 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3929 gave an unhelpful defer error. I have not managed to make this reject any
3930 supplied certificates, but the error message it gives is "no certificate
3931 supplied", which is not helpful.
3933 70. exigrep's output now also includes lines that are not associated with any
3934 message, but which match the given pattern. Implemented by a patch from
3935 Martin Sluka, which also tidied up the Perl a bit.
3937 71. Recipient callout verification, like sender verification, was using <> in
3938 the MAIL FROM command. This isn't really the right thing, since the actual
3939 sender may affect whether the remote host accepts the recipient or not. I
3940 have changed it to use the actual sender in the callout; this means that
3941 the cache record is now keyed on a recipient/sender pair, not just the
3942 recipient address. There doesn't seem to be a real danger of callout loops,
3943 since a callout by the remote host to check the sender would use <>.
3944 [SEE ABOVE: changed after hitting problems.]
3946 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3947 temporary errors. However, in the case of such a code being given after
3948 the end of a data transmission (i.e. after ".") Exim was failing to write
3949 a retry record for the message. (Yes, there was some broken host that was
3950 actually sending 8xx at this point.)
3952 73. An unknown lookup type in a host list could cause Exim to panic-die when
3953 the list was checked. (An example that provoked this was putting <; in the
3954 middle of a list instead of at the start.) If this happened during a DATA
3955 ACL check, a -D file could be left lying around. This kind of configuration
3956 error no longer causes Exim to die; instead it causes a defer errror. The
3957 incident is still logged to the main and panic logs.
3959 74. Buglet left over from Exim 3 conversion. The message "too many messages
3960 in one connection" was written to the rejectlog but not the mainlog, except
3961 when address rewriting (yes!) was being logged.
3963 75. Added write_rejectlog option.
3965 76. When a system filter was run not as root (that is, when system_filter_user
3966 was set), the values of the $n variables were not being returned to the
3967 main process; thus, they were not subsequently available in the $sn
3970 77. Added +return_path_on_delivery log selector.
3972 78. A connection timeout was being treated differently from recipients deferred
3973 when testing hosts_max_try with a message that was older than the host's
3974 retry timeout. (The host should not be counted, thus allowing all hosts to
3975 be tried at least once before bouncing.) This may have been the cause of an
3976 occasionally reported bug whereby a message would remain on the queue
3977 longer than the retry timeout, but would be bounced if a delivery was
3978 forced. I say "may" because I never totally pinned down the problem;
3979 setting up timeout/retry tests is difficult. See also the next item.
3981 79. The ultimate address timeout was not being applied to errors that involved
3982 a combination of host plus message (for example, a timeout on a MAIL
3983 command). When an address resolved to a number of possible hosts, and they
3984 were not all tried for each delivery (e.g. because of hosts_max_try), a
3985 message could remain on the queue longer than the retry timeout.
3987 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3990 81. Fixed an obscure SMTP outgoing bug which required at least the following
3991 conditions: (a) there was another message waiting for the same server;
3992 (b) the server returned 5xx to all RCPT commands in the first message so
3993 that the message was not completed; (c) the server dropped the connection
3994 or gave a negative response to the RSET that Exim sends to abort the
3995 transaction. The observed case was a dropped connection after DATA that had
3996 been sent in pipelining mode. That is, the server had advertised PIPELINING
3997 but was not implementing it correctly. The effect of the bug was incorrect
3998 behaviour, such as trying another host, and this could lead to a crash.
4004 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4005 and daemon.c were passed as pointers to ints; they should have been
4006 pointers to socklen_t variables (which are typically unsigned ints).
4008 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4011 3. Fixed a really odd bug that affected only the testing scheme; patching a
4012 certain fixed string in the binary changed the value of another string that
4013 happened to be identical to the end of the original first string.
4015 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4016 name", it returns that address as the IP address. On some operating
4017 systems (e.g. Solaris), it also passes back the IP address string as the
4018 "host name". However, on others (e.g. Linux), it passes back an empty
4019 string. Exim wasn't checking for this, and was changing the host name to an
4020 empty string, assuming it had been canonicized.
4022 5. Although rare, it is permitted to have more than one PTR record for a given
4023 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4024 all the names associated with an address, because they do in Solaris.
4025 However, it seems that they do not in Linux for data that comes from the
4026 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4027 I found this out when I moved to a new Linux workstation and tried to run
4028 the Exim test suite.
4030 To get round this problem I have changed the code so that it now does its
4031 own call to the DNS to look up PTR records when searching for a host name.
4032 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4033 addresses that are only in /etc/hosts are still found.
4035 This behaviour is, however, controlled by an option called host_lookup_
4036 order, which defaults to "bydns:byaddr". If people want to use the other
4037 order, or indeed, just use one or the other means of lookup, they can
4038 specify it in this variable.
4040 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4041 some operating systems, this comes back from gethostbyaddr() as an empty
4042 string, and this is what Exim used to test for. However, it seems that in
4043 other systems, "." is yielded. Exim now tests for this case too.
4045 7. The values of check_spool_space and check_log_space are now held internally
4046 as a number of kilobytes instead of an absolute number of bytes. If a
4047 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4048 kilobyte. This means that much larger values can be stored.
4050 8. Exim monitor: an attempt to get the action menu when not actually pointing
4051 at a message produces an empty menu entitled "No message selected". This
4052 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4053 no entries in it ("Shell widget menu has zero width and/or height"). So I
4054 have added a single, blank menu entry in this case.
4056 9. Added ${quote_local_part.
4058 10. MIME decoding is now applied to the contents of Subject: header lines when
4061 11. Now that a reference to $sender_host_address automatically causes a reverse
4062 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4063 host lookup before query-style lookups in lists that might use this
4064 variable. This has therefore been abolished, and the "net-" prefix is no
4065 longer necessary for query-style lookups.
4067 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4068 have been a typo for LFLAGS, so it has been changed.
4070 13. The install script calls Exim with "-C /dev/null" in order to find the
4071 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4072 to be output. Howeve, since Exim outputs its version number before the
4073 error, it didn't break the script. It just looked ugly. I fixed this by
4074 always allowing "-C /dev/null" if the caller is root.
4076 14. Ignore overlarge ACL variable number when reading spool file - insurance
4077 against a later release with more variables having written the file.
4079 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4080 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4081 this, as well as the form without the "IPv6" on the front.
4083 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4084 OS/Makefile-Darwin file.
4086 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4087 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4089 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4090 variable. However, some OS use ints for inodes. I've added cast to long int
4091 to get rid of the compiler warning.
4093 19. I had forgotten to lock out "/../" in configuration file names when
4094 ALT_CONFIG_PREFIX was set.
4096 20. Routers used for verification do not need to specify transports. However,
4097 if such a router generated a host list, and callout was configured, Exim
4098 crashed, because it could not find a port number from the (non-existent)
4099 transport. It now assumes port 25 in this circumstance.
4101 21. Added the -t option to exigrep.
4103 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4104 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4105 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4106 them is set, LOOKUP_LSEARCH is forced.
4108 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4110 24. Added sender and host information to the "rejected by local_scan()" log
4111 line; previously there was no indication of these.
4113 25. Added .include_if_exists.
4115 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4116 Linux. It turns out that not all file systems support this. Apparently some
4117 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4118 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4119 not supported on the file descriptor, is now ignored when Exim is trying to
4120 sync a directory. This applies only to Linux.
4122 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4124 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4125 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4128 29. Added support for maildirsize files from supplied patch (modified a bit).
4130 30. The use of :fail: followed by an empty string could lead Exim to respond to
4131 sender verification failures with (e.g.):
4133 550 Verification failed for <xxx>
4134 550 Sender verify failed
4136 where the first response line was missing the '-' that indicates it is not
4137 the final line of the response.
4139 31. The loop for finding the name of the user that called Exim had a hardwired
4140 limit of 10; it now uses the value of finduser_retries, which is used for
4141 all other user lookups.
4143 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4146 33. Exim was neglecting to zero errno before one call of strtol() when
4147 expanding a string and expecting an integer value. On some systems this
4148 resulted in spurious "integer overflow" errors. Also, it was casting the
4149 result into an int without checking.
4151 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4152 did not work. The code looks as if it has *never* worked, though it appears
4153 to have been documented since at least releast 1.62. I have made it work.
4155 35. The "timeout_DNS" error in retry rules, also documented since at least
4156 1.62, also never worked. As it isn't clear exactly what this means, and
4157 clearly it isn't a major issue, I have abolished the feature by treating it
4158 as "timeout", and writing a warning to the main and panic logs.
4160 36. The display of retry rules for -brt wasn't always showing the error code
4163 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4164 timeout_connect_A, timeout_connect_MX.
4166 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4167 to the empty sender.
4169 39. The daemon was not analysing the content of -oX till after it had closed
4170 stderr and disconnected from the controlling terminal. This meant that any
4171 syntax errors were only noted on the panic log, and the return code from
4172 the command was 0. By re-arranging the code a little, I've made the
4173 decoding happen first, so such errors now appear on stderr, and the return
4174 code is 1. However, the actual setting up of the sockets still happens in
4175 the disconnected process, so errors there are still only recorded on the
4178 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4179 connections (as happens on some IP stacks) was logged at start up time as
4180 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4181 it from "IPv6 and IPv4", which means that two separate sockets are being
4184 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4185 says whether AF_INET or AF_INET6 was passed as an argument.
4187 42. Exiwhat output was messed up when time zones were included in log
4190 43. Exiwhat now gives more information about the daemon's listening ports,
4191 and whether -tls-on-connect was used.
4193 44. The "port" option of the smtp transport is now expanded.
4195 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4196 silently ignored. Now an error message is written to the main and panic
4199 46. There's a new ACL modifier called "logwrite" which writes to a log file
4200 as soon as it is encountered.
4202 47. Added $local_user_uid and $local_user_gid at routing time.
4204 48. Exim crashed when trying to verify a sender address that was being
4207 49. Exim was recognizing only a space character after ".include". It now also
4208 recognizes a tab character.
4210 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4211 extracting the relevant information from the specification. The man page no
4212 longer contains scrambled data for the -d option, and I've added a section
4213 at the front about calling Exim under different names.
4215 51. Added "extra_headers" argument to the "mail" command in filter files.
4217 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4220 53. Installed eximstats 1.29.
4222 54. Added transport_filter_timeout as a generic transport option.
4224 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4225 Cc: header lines. This was required by RFC 822, but it not required by RFC
4228 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4229 incoming messages that did not have them. Now it does so only if the
4230 message originates locally, that is, if there is no associated remote host
4231 address. When Resent- header lines are present, this applies to the Resent-
4232 lines rather than the non-Resent- lines.
4234 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4235 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4237 58. Messages for configuration errors now include the name of the main
4238 configuration file - useful now that there may be more than one file in a
4239 list (.included file names were always shown).
4241 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4242 for those rare installations that do not start the daemon as root or run it
4243 setuid root. I've cut out the call to initgroups() if the daemon is not
4246 60. The Exim user and group can now be bound into the binary as text strings
4247 that are looked up at the start of Exim's processing.
4249 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4251 62. Added $mailstore_basename variable.
4253 63. Installed patch to sieve.c from Michael Haardt.
4255 64. When Exim failed to open the panic log after failing to open the main log,
4256 the original message it was trying to log was written to stderr and debug
4257 output, but if they were not available (the usual case in production), it
4258 was lost. Now it is written to syslog before the two lines that record the
4259 failures to open the logs.
4261 65. Users' Exim filters run in subprocesses under the user's uid. It is
4262 possible for a "deliver" command or an alias in a "personal" command to
4263 provoke an address rewrite. If logging of address rewriting is configured,
4264 this fails because the process is not running as root or exim. There may be
4265 a better way of dealing with this, but for the moment (because 4.30 needs
4266 to be released), I have disabled address rewrite logging when running a
4267 filter in a non-root, non-exim process.
4273 1. The buildconfig auxiliary program wasn't quoting the value set for
4274 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4275 not defined. This bug was masked in 4.22 by the effect that was fixed in
4278 2. Some messages that were rejected after a message id was allocated were
4279 shown as "incomplete" by exigrep. It no longer does this for messages that
4280 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4282 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4283 have allow_domain_literals set, the ID did not get logged in the <= line.
4284 Domain literals are now always recognized in Message-ID: header lines.
4286 4. The first argument for a ${extract expansion item is the key name or field
4287 number. Leading and trailing spaces in this item were not being ignored,
4288 causing some misleading effects.
4290 5. When deliver_drop_privilege was set, single queue runner processes started
4291 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4292 same command in the process it spins off) were not dropping privilege.
4294 6. When the daemon running as "exim" started a queue runner, it always
4295 re-executed Exim in the spun-off process. This is a waste of effort when
4296 deliver_drop_privilege is set. The new process now just calls the
4297 queue-runner function directly.
4303 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4306 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4307 ignored. Though the use of -oP was forcing the writing of a pid file, it
4308 was always written to the default place.
4310 3. If the message "no IP address found for host xxxx" is generated during
4311 incoming verification, it is now followed by identification of the incoming
4312 connection (so you can more easily find what provoked it).
4314 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4316 5. Added some features to "harden" Exim a bit more against certain attacks:
4318 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4319 be put in Local/Makefile. This is like the never_users runtime option,
4320 but it cannot be overridden. The default setting is "root".
4322 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4323 prefix string with which any file named in a -C command line option
4326 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4327 is retained for -C and -D only if the caller of Exim is root. Without
4328 it, the exim user may also use -C and -D and retain privilege.
4330 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4331 command line option is disabled.
4333 6. Macro names set by the -D option must start with an upper case letter, just
4334 like macro names defined in the configuration file.
4336 7. Added "dereference=" facility to LDAP.
4338 8. Two instances of the typo "uknown" in the source files are fixed.
4340 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4341 the Configure-Makefile script screwed up while processing it.
4343 10. Incorporated PCRE 4.4.
4345 11. The SMTP synchronization check was not operating right at the start of an
4346 SMTP session. For example, it could not catch a HELO sent before the client
4347 waited for the greeting. There is now a check for outstanding input at the
4348 point when the greeting is written. Because of the duplex, asynchronous
4349 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4350 way, but not yet received, when the check is performed.
4352 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4353 on TCP/IP sockets, because this apparently causes some broken clients to
4356 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4357 unchanged) from the Cygwin maintainer.
4359 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4360 was supported instead of testing for mbx. Effectively a typo.
4362 15. The spa authenticator server code was not checking that the input it
4363 received was valid base64.
4365 16. The debug output line for the "set" modifier in ACLs was not showing the
4366 name of the variable that was being set.
4368 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4370 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4371 lookup on its own. Something else was needed to trigger the lookup. For
4372 example, a match in host_lookup or the need for a host name in a host list.
4373 Now, if $sender_host_name is referenced and the host name has not yet been
4374 looked up, a lookup is performed. If the lookup fails, the variable remains
4375 empty, and $host_lookup_failed is set to "1".
4377 19. Added "eqi" as a case-independent comparison operator.
4379 20. The saslauthd authentication condition could segfault if neither service
4380 nor realm was specified.
4382 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4383 error message that was logged was misleading, and incoming SMTP
4384 connections were dropped. The message is now more accurate, and temporary
4385 errors are given to SMTP connections.
4387 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4388 (except RSET) with a 421 error, until QUIT is received. However, it was
4389 failing to send a response to QUIT.
4391 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4392 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4393 ACL is now run after the helo_try_verify_hosts code.
4395 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4396 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4397 case-independent, so other case variants are also recognized). Apparently
4398 some systems use these upper case variants.
4400 25. If more than two messages were waiting for the same host, and a transport
4401 filter was specified for the transport, Exim sent two messages over the
4402 same TCP/IP connection, and then failed with "socket operation on non-
4403 socket" when it tried to send the third.
4405 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4407 27. The extern definition of crypt16() in expand.c was not being excluded when
4408 the OS had its own crypt16() function.
4410 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4411 as a preferred synonym for return_size_limit, both as an option and as an
4414 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4416 30. Changed the default configuration ACL to relax the local part checking rule
4417 for addresses that are not in any local domains. For these addresses,
4418 slashes and pipe symbols are allowed within local parts, but the sequence
4419 /../ is explicitly forbidden.
4421 31. SPA server authentication was not clearing the challenge buffer before
4424 32. log_message in a "warn" ACL statement was writing to the reject log as
4425 well as to the main log, which contradicts the documentation and doesn't
4426 seem right (because no rejection is happening). So I have stopped it.
4428 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4429 However, I am unable to do any testing of this.
4431 34. Fixed an infelicity in the appendfile transport. When checking directories
4432 for a mailbox, to see if any needed to be created, it was accidentally
4433 using path names with one or more superfluous leading slashes; tracing
4434 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4436 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4437 contents are added to the log line that is written for every discarded
4438 recipient. (Previously a log_message setting was ignored.)
4440 36. The ${quote: operator now quotes the string if it is empty.
4442 37. The install script runs exim in order to find its version number. If for
4443 some reason other than non-existence or emptiness, which it checks, it
4444 could not run './exim', it was installing it with an empty version number,
4445 i.e. as "exim-". This error state is now caught, and the installation is
4448 38. An argument was missing from the function that creates an error message
4449 when Exim fails to connect to the socket for saslauthd authentication.
4450 This could cause Exim to crash, or give a corrupted message.
4452 39. Added isip, isip4, and isip6 to ${if conditions.
4454 40. The ACL variables $acl_xx are now saved with the message, and can be
4455 accessed later in routers, transports, and filters.
4457 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4458 strings in the file are not string-expanded.
4460 42. If a MAIL command specified a SIZE value that was too large to fit into an
4461 int variable, the check against message_size_limit failed. Such values are
4462 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4463 day this will have to be increased, but I don't think I want to be around
4464 when emails are that large.
4471 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4472 iconv() is not standard in FreeBSD.
4474 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4475 IPv6 enabled. The observed symptom was a segmentation fault on return from
4476 the function os_common_find_running_interfaces() in src/os.c.
4478 3. In the check_special_case() function in daemon.c I had used "errno" as an
4479 argument name, which causes warnings on some systems. This was basically a
4480 typo, since it was named "eno" in the comments!
4482 4. The code that waits for the clock to tick (at a resolution of some fraction
4483 of a second) so as to ensure message-id uniqueness was always waiting for
4484 at least one whole tick, when it could have waited for less. [This is
4485 almost certainly not relevant at current processor speeds, where it is
4486 unlikely to ever wait at all. But we try to future-proof.]
4488 5. The function that sleeps for a time interval that includes fractions of a
4489 second contained a race. It did not block SIGALRM between setting the
4490 timer, and suspending (a couple of lines later). If the interval was short
4491 and the sigsuspend() was delayed until after it had expired, the suspension
4492 never ended. On busy systems this could lead to processes getting stuck for
4495 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4496 process, before it forks any delivery processes. The open lookup caching
4497 mechanism meant that the open file or database connection was passed into
4498 the delivery process. The problem was that delivery processes always tidy
4499 up cached lookup data. This could cause a problem for the next delivery
4500 process started by the queue runner, because the external queue runner
4501 process does not know about the closure. So the next delivery process
4502 still has data in the lookup cache. In the case of a file lookup, there was
4503 no problem because closing a file descriptor in a subprocess doesn't affect
4504 the parent. However, if the lookup was caching a connection to a database,
4505 the connection was closed, and the second delivery process was likely to
4506 see errors such as "PGSQL: query failed: server closed the connection
4507 unexpectedly". The problem has been fixed by closing all cached lookups
4508 in a queue runner before running a delivery process.
4510 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4511 seem to have the "const" qualifier which it has on other OS. I've
4514 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4515 *of the same type* (client or server) with the same public name that an
4516 error should be diagnosed.
4518 9. When Exim looked up a host name for an IP address, but failed to find the
4519 original IP address when looking up the host name (a safety check), it
4520 output the message "<ip address> does not match any IP for NULL", which was
4521 confusing, to say the least. The bug was that the host name should have
4522 appeared instead of "NULL".
4524 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4525 user that is built into the binary, and the -C or -D options is used, root
4526 privilege is dropped before the configuration file is read. In addition,
4527 logging is switched to stderr instead of the normal log files. If the
4528 configuration then re-defines the Exim user, the unprivileged environment
4529 is probably not what is expected, so Exim logs a panic warning message (but
4532 However, if deliver_drop_privilege is set, the unprivileged state may well
4533 be exactly what is intended, so the warning has been cut out in that case,
4534 and Exim is allowed to try to write to its normal log files.
4540 1. smtp_return_error_details was not giving details for temporary sender
4541 or receiver verification errors.
4543 2. Diagnose a configuration error if two authenticators have the same public
4546 3. Exim used not to create the message log file for a message until the first
4547 delivery attempt. This could be confusing when incoming messages were held
4548 for policy or load reasons. The message log file is now created at the time
4549 the message is received, and an initial "Received" line is written to it.
4551 4. The automatically generated man page for command line options had a minor
4552 bug that caused no ill effects; however, a more serious problem was that
4553 the procedure for building the man page automatically didn't always
4554 operate. Consequently, release 4.20 contains an out-of-date version. This
4555 shouldn't happen again.
4557 5. When building Exim with embedded Perl support, the script that builds the
4558 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4559 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4561 6. The freeze_tell option was not being used for messages that were frozen on
4562 arrival, either by an ACL or by local_scan().
4564 7. Added the smtp_incomplete_transaction log selector.
4566 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4567 was accepting AUTH without a new EHLO.
4569 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4570 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4573 10. Logging of TCP/IP connections (when configured) now happens in the main
4574 daemon process instead of the child process, so that the TCP/IP connection
4575 count is more accurate (but it can never be perfect).
4577 11. The use of "drop" in a nested ACL was not being handled correctly in the
4578 outer ACL. Now, if condition failure induced by the nested "drop" causes
4579 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4580 or "require"), the connection is dropped.
4582 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4583 that yield "discard" can now be used with an "accept" or a "discard" verb,
4584 but an error is generated for any others (because I can't see a useful way
4585 to define what should happen).
4587 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4588 are now processed in the same way as lines in the Exim configuration file.
4589 In particular, continuation lines are supported.
4591 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4593 15. Added -ti meaning -t -i.
4595 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4596 domains, and warn by logging if others are found.
4598 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4599 returned from GIFCONF when Exim is trying to find the list of interfaces on
4600 a host. The code in os.c has been modified to copy each ifreq to an aligned
4601 structure in all cases.
4603 Also, in some cases, the returned ifreq's were being copied to a 'struct
4604 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4605 means the last couple of bytes of an IPv6 address could be chopped if the
4606 ifreq contained only a normal sockaddr (14 bytes storage).
4608 18. Named domain lists were not supported in the hosts_treat_as_local option.
4609 An entry such as +xxxx was not recognized, and was treated as a literal
4612 19. Ensure that header lines added by a DATA ACL are included in the reject log
4613 if the ACL subsequently rejects the message.
4615 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4616 MD5 (which is deprecated).
4618 21. When testing a filter file using -bf, Exim was writing a message when it
4619 took the sender from a "From " line in the message, but it was not doing so
4620 when it took $return_path from a Return-Path: header line. It now does.
4622 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4623 with a valid header line field name (a series of printing characters
4624 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4626 23. Changed "disc" in the source to "disk" to conform to the documentation and
4627 the book and for uniformity.
4629 24. Ignore Sendmail's -Ooption=value command line item.
4631 25. When execve() failed while trying to run a command in a pipe transport,
4632 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4633 could be confused with a return value of 69 from the command itself. This
4634 has been changed to 127, the value the shell returns if it is asked to run
4635 a non-existent command. The wording for the related log line suggests a
4636 non-existent command as the problem.
4638 26. If received_header_text expands to an empty string, do not add a Received:
4639 header line to the message. (Well, it adds a token one on the spool, but
4640 marks it "old" so that it doesn't get used or transmitted.)
4642 27. Installed eximstats 1.28 (addition of -nt option).
4644 28. There was no check for failure on the call to getsockname() in the daemon
4645 code. This can fail if there is a shortage of resources on the system, with
4646 ENOMEM, for example. A temporary error is now given on failure.
4648 29. Contrary to the C standard, it seems that in some environments, the
4649 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4650 program. Exim now does this explicitly; it affects the formatting of
4651 timestamps using strftime().
4653 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4654 complaints. I've now initialized all the variables, to avoid this.
4656 32. Header lines added by a system filter were not being "seen" during
4657 transport-time rewrites.
4659 33. The info_callback() function passed to OpenSSL is set up with type void
4660 (*)(SSL *, int, int), as described somewhere. However, when calling the
4661 function (actually a macro) that sets it up, the type void(*)() is
4662 expected. I've put in a cast to prevent warnings from picky compilers.
4664 34. If a DNS black list lookup found a CNAME record, but there were no A
4665 records associated with the domain it pointed at, Exim crashed.
4667 35. If a DNS black list lookup returned more than one A record, Exim ignored
4668 all but the first. It now scans all returned addresses if a particular IP
4669 value is being sought. In this situation, the contents of the
4670 $dnslist_value variable are a list of all the addresses, separated by a
4673 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4674 to accept a host name and all its aliases if the forward lookup for any of
4675 them yielded the IP address of the incoming connection. Now it accepts only
4676 those names whose forward lookup yields the correct IP address. Any other
4677 names are discarded. This closes a loophole whereby a rogue DNS
4678 administrator could create reverse DNS records to break through a
4679 wildcarded host restriction in an ACL.
4681 37. If a user filter or a system filter that ran in a subprocess used any of
4682 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4683 the wrong values were passed to the pipe command ($thisaddress had the
4684 value of $0, $0 had the value of $1, etc). This bug was introduced by
4685 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4687 38. Improved the line breaking for long SMTP error messages from ACLs.
4688 Previously, if there was no break point between 40 and 75 characters, Exim
4689 left the rest of the message alone. Two changes have been made: (a) I've
4690 reduced the minimum length to 35 characters; (b) if it can't find a break
4691 point between 35 and 75 characters, it looks ahead and uses the first one
4692 that it finds. This may give the occasional overlong line, but at least the
4693 remaining text gets split now.
4695 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4696 file descriptors might be low, and that setting 1000 would always raise it.
4697 It turns out that in some environments, the limit is already over 1000 and
4698 that lowering it causes trouble. So now Exim takes care not to decrease it.
4700 40. When delivering a message, the value of $return_path is set to $sender_
4701 address at the start of routing (routers may change the value). By an
4702 oversight, this default was not being set up when an address was tested by
4703 -bt or -bv, which affected the outcome if any router or filter referred to
4706 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4707 log only when "message" or "log_message" are set. However, if one of the
4708 conditions was an address verification, or a call to a nested ACL, the
4709 messages generated by the underlying test were being passed through. This
4710 no longer happens. The underlying message is available in $acl_verify_
4711 message for both "message" and "log_message" expansions, so it can be
4712 passed through if needed.
4714 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4715 new expansion $bh_ to give the encoded byte string without charset
4716 translation. Translation happens only if iconv() is available; HAVE_ICONV
4717 indicates this at build time. HEADERS_CHARSET gives the charset to
4718 translate to; headers_charset can change it in the configuration, and
4719 "headers charset" can change it in an individual filter file.
4721 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4722 that creates RFC 2047 encoded "words" labels them as that charset instead
4723 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4724 expansion operator; (ii) when Exim creates a From: line for a local
4725 message; (iii) when a header line is rewritten to include a "phrase" part.
4727 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4728 buggy, causing it to skip the first lines of messages whose message ID
4729 ended in 'D'. This would not have bitten before Exim release 4.14, because
4730 message IDs were unlikely to end in 'D' before then. The effect was to have
4731 incorrect size information for certain domains.
4733 45. #include "config.h" was missing at the start of the crypt16.c module. This
4734 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4737 46. If there was a timeout during a "random" callout check, Exim treated it as
4738 a failure of the random address, and carried on sending RSET and the real
4739 address. If the delay was just some slowness somewhere, the response to the
4740 original RCPT would be taken as a response to RSET and so on, causing
4741 mayhem of various kinds.
4743 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4744 when I implemented it. It didn't allow for the fact that some option values
4745 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4746 the right test for positive values.
4748 48. Domain names in DNS records are case-independent. Exim always looks them up
4749 in lower case. Some resolvers return domain names in exactly the case they
4750 appear in the zone file, that is, they may contain uppercase letters. Not
4751 all resolvers do this - some return always lower case. Exim was treating a
4752 change of case by a resolver as a change of domain, similar to a widening
4753 of a domain abbreviation. This triggered its re-routing code and so it was
4754 trying to route what was effectively the same domain again. This normally
4755 caused routing to fail (because the router wouldn't handle the domain
4756 twice). Now Exim checks for this case specially, and just changes the
4757 casing of the domain that it ultimately uses when it transmits the message
4760 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4763 50. If a filter generated a file delivery with a non-absolute name (possible if
4764 no home directory exists for the router), the forbid_file option was not
4767 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4768 the existing equality matching.
4770 52. Exim was using ints instead of ino_t variables in some places where it was
4771 dealing with inode numbers.
4773 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4774 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4775 called TMPDIR, and if it finds it is different, it changes its value.
4777 54. The smtp_printf() function is now made available to local_scan() so
4778 additional output lines can be written before returning. There is also an
4779 smtp_fflush() function to enable the detection of a dropped connection.
4780 The variables smtp_input and smtp_batched_input are exported to
4783 55. Changed the default runtime configuration: the message "Unknown user"
4784 has been removed from the ACL, and instead placed on the localuser router,
4785 using the cannot_route_message feature. This means that any verification
4786 failures that generate their own messages won't get overridden. Similarly,
4787 the "Unrouteable address" message that was in the ACL for unverifiable
4788 relay addresses has also been removed.
4790 56. Added hosts_avoid_esmtp to the smtp transport.
4792 57. The exicyclog script was not checking for the esoteric option
4793 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4794 will work only if exicyclog is run under the appropriate euid.
4796 58. Following a discussion on the list, the rules by which Exim recognises line
4797 endings on incoming messages have been changed. The -dropcr and drop_cr
4798 options are now no-ops, retained only for backwards compatibility. The
4799 following line terminators are recognized: LF CRLF CR. However, special
4800 processing applies to CR:
4802 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4803 nor a local message in the state where . is a terminator.
4805 (ii) If a bare CR is encountered in a header line, an extra space is added
4806 after the line terminator so as not to end the header. The reasoning
4807 behind this is that bare CRs in header lines are most likely either
4808 to be mistakes, or people trying to play silly games.
4810 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4811 was being incorrectly given as 18 bytes larger than it should have been.
4812 This is a VOB (very old bug).
4814 60. This may never have affected anything current, but just in case it has:
4815 When the local host is found other than at the start of a list of hosts,
4816 the local host, those with the same MX, and any that follow, are discarded.
4817 When the list in question was part of a longer list of hosts, the following
4818 hosts (not currently being processed) were also being discarded. This no
4819 longer happens. I'm not sure if this situation could ever has previously
4822 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4825 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4826 header. This is something that is recommended in a new Internet Draft, and
4827 is something that is documented as being done by Sendmail. There are two
4828 possible values. For messages generated by the autoreply transport, Exim
4831 Auto-Submitted: auto-replied
4833 whereas for all other generated messages (e.g. bounces) it adds
4835 Auto-Submitted: auto-generated
4837 63. The "personal" condition in filters now includes a test for the
4838 Auto-Submitted: header. If it contains the string "auto-" the message it
4839 not considered personal.
4841 64. Added rcpt_include_affixes as a generic transport option.
4843 65. Added queue_only_override (default true).
4845 66. Added the syslog_duplication option.
4847 67. If what should have been the first header line of a message consisted of
4848 a space followed by a colon, Exim was mis-interpreting it as a header line.
4849 It isn't of course - it is syntactically invalid and should therefore be
4850 treated as the start of the message body. The misbehaviour could have
4851 caused a number of strange effects, including loss of data in subsequent
4852 header lines, and spool format errors.
4854 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4855 client host had authenticated. This control can now be exercised by an ACL
4856 for more flexibility.
4858 69. By default, callouts do not happen when testing with -bh. There is now a
4859 variant, -bhc, which does actually run the callout code, including
4860 consulting and updating the callout cache.
4862 70. Added support for saslauthd authentication, courtesy of Alexander
4865 71. If statvfs() failed on the spool or log directories while checking their
4866 size for availability, Exim confusingly gave the error "space shortage".
4867 Furthermore, in debugging mode it crashed with a floating point exception.
4868 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4869 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4870 serious problem, Exim now writes to the main and panic logs when this
4871 happens, with details of the failure. It then refuses to accept the
4872 incoming message, giving the message "spool directory problem" or "log
4873 directory problem" with a 421 code for SMTP messages.
4875 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4876 0, 1, and 2 exist, because some OS complain for execs without them (see
4877 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4878 descriptors. However, the code omitted to check that the open succeeded,
4879 causing mysterious errors if for some reason the permissions on /dev/null
4880 got screwed. Now Exim writes a message to the main and panic logs, and
4881 bombs out if it can't open /dev/null.
4883 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4884 interact so that it is all more flexible. It is supposed to remain
4885 backwards compatible. Also added extra_local_interfaces.
4887 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4888 to bomb out with an assertion failure - to the client this appears as a
4889 connection drop. This problem occurs in the part of the code that was taken
4890 from the Samba project. Fortunately, the assertion is in a very simple
4891 function, so I have fixed this by reproducing the function inline in the
4892 one place where it is called, and arranging for authentication to fail
4893 instead of killing the process with assert().
4895 75. The SPA client code was not working when the server requested OEM rather
4896 than Unicode encoding.
4898 76. Added code to make require_files with a specific uid setting more usable in
4899 the case where statting the file as root fails - usually a non-root-mounted
4900 NFS file system. When this happens and the failure is EACCES, Exim now
4901 forks a subprocess and does the per-uid checking as the relevant uid.
4903 77. Added process_log_path.
4905 78. If log_file_path was not explicitly set, a setting of check_log_space or
4906 check_log_inodes was ignored.
4908 79. If a space check for the spool or log partitions fails, the incident is now
4909 logged. Of course, in the latter case the data may get lost...
4911 80. Added the %p formatting code to string_format() so that it can be used to
4912 print addresses in debug_print(). Adjusted all the address printing in the
4913 debugging in store.c to use %p rather than %d.
4915 81. There was a concern that a line of code in smtp_in.c could overflow a
4916 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4917 initially expressed, the concern was not well-founded, because trailing
4918 spaces are removed early. However, if the trailing spaces were followed by
4919 a NULL, they did not get removed, so the overflow was possible. Two fixes
4922 (a) I re-wrote the offending code in a cleaner fashion.
4923 (b) If an incoming SMTP command contains a NULL character, it is rejected
4926 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4927 runs initgroups(), so that if the Exim user is in any additional groups,
4928 they will be used during message reception.
4934 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob