1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.540 2008/01/28 18:24:07 fanf2 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
10 When building exim an external PCRE library is now needed -
11 PCRE is a system library on the majority of modern systems.
12 See entry on PCRE_LIBS in EDITME file.
14 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
15 conversation. Added nologin parameter to request.
16 Patch contributed by Kirill Miazine
18 TF/01 Do not log submission mode rewrites if they do not change the address.
20 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
26 TK/01 Add preliminary DKIM support. Currently requires a forked version of
27 ALT-N's libdkim that I have put here:
28 http://duncanthrax.net/exim-experimental/
30 Note to Michael Haardt: I had to rename some vars in sieve.c. They
31 were called 'true' and it seems that C99 defines that as a reserved
32 keyword to be used with 'bool' variable types. That means you could
33 not include C99-style headers which use bools without triggering
34 build errors in sieve.c.
36 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
37 as mailq or other aliases. Changed the --help handling significantly
38 to do whats expected. exim_usage() emits usage/help information.
40 SC/01 Added the -bylocaldomain option to eximstats.
42 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
44 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
46 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
52 PH/01 Another patch from the Sieve maintainer.
54 PH/02 When an IPv6 address is converted to a string for single-key lookup
55 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
56 dots are used instead of colons so that keys in lsearch files need not
57 contain colons. This was done some time before quoting was made available
58 in lsearch files. However, iplsearch files do require colons in IPv6 keys
59 (notated using the quote facility) so as to distinguish them from IPv4
60 keys. This meant that lookups for IP addresses in host lists did not work
61 for iplsearch lookups.
63 This has been fixed by arranging for IPv6 addresses to be expressed with
64 colons if the lookup type is iplsearch. This is not incompatible, because
65 previously such lookups could never work.
67 The situation is now rather anomolous, since one *can* have colons in
68 ordinary lsearch keys. However, making the change in all cases is
69 incompatible and would probably break a number of configurations.
71 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
74 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
75 conversion specification without a maximum field width, thereby enabling
76 a rogue spamd server to cause a buffer overflow. While nobody in their
77 right mind would setup Exim to query an untrusted spamd server, an
78 attacker that gains access to a server running spamd could potentially
79 exploit this vulnerability to run arbitrary code as the Exim user.
81 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
82 $primary_hostname instead of what libspf2 thinks the hosts name is.
84 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
85 a directory entry by the name of the lookup key. Previously, if a
86 symlink pointed to a non-existing file or a file in a directory that
87 Exim lacked permissions to read, a lookup for a key matching that
88 symlink would fail. Now it is enough that a matching directory entry
89 exists, symlink or not. (Bugzilla 503.)
91 PH/03 The body_linecount and body_zerocount variables are now exported in the
94 PH/04 Added the $dnslist_matched variable.
96 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
97 This means they are set thereafter only if the connection becomes
100 PH/06 Added the client_condition to authenticators so that some can be skipped
101 by clients under certain conditions.
103 PH/07 The error message for a badly-placed control=no_multiline_responses left
104 "_responses" off the end of the name.
106 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
108 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
109 (without spaces) instead of just copying the configuration text.
111 PH/10 Added the /noupdate option to the ratelimit ACL condition.
113 PH/11 Added $max_received_linelength.
115 PH/12 Added +ignore_defer and +include_defer to host lists.
117 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
118 way in which PCRE > 7.0 is built.
120 PH/14 Implemented queue_only_load_latch.
122 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
123 MAIL command. The effect was to mangle the value on 64-bit systems.
125 PH/16 Another patch from the Sieve maintainer.
127 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
129 PH/18 If a system quota error occurred while trying to create the file for
130 a maildir delivery, the message "Mailbox is full" was not appended to the
131 bounce if the delivery eventually timed out. Change 4.67/27 below applied
132 only to a quota excession during the actual writing of the file.
134 PH/19 It seems that peer DN values may contain newlines (and other non-printing
135 characters?) which causes problems in log lines. The DN values are now
136 passed through string_printing() before being added to log lines.
138 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
139 and InterBase are left for another time.)
141 PH/21 Added message_body_newlines option.
143 PH/22 Guard against possible overflow in moan_check_errorcopy().
145 PH/23 POSIX allows open() to be a macro; guard against that.
147 PH/24 If the recipient of an error message contained an @ in the local part
148 (suitably quoted, of course), incorrect values were put in $domain and
149 $local_part during the evaluation of errors_copy.
155 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
156 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
159 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
160 issue a MAIL command.
162 PH/02 In an ACL statement such as
164 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
166 if a client was not listed at all, or was listed with a value other than
167 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
168 the condition was not true (as it should be), so access was not denied.
169 The bug was that the ! inversion was incorrectly passed on to the second
170 item. This has been fixed.
172 PH/03 Added additional dnslists conditions == and =& which are different from
173 = and & when the dns lookup returns more than one IP address.
175 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
176 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
178 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
179 FSYNC, which compiles an option called disable_fsync that allows for
180 bypassing fsync(). The documentation is heavily laced with warnings.
182 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
184 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
185 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
186 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
187 including adding "make clean"; (3) Added -fPIC when compiling the test
188 dynamically loaded module, to get rid of a warning.
190 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
191 message fails, move_frozen_messages = true and ignore_bounce_errors_after
192 = 0s. The bug is otherwise harmless.
194 PH/07 There was a bug in the dovecot authenticator such that the value of
195 $auth1 could be overwritten, and so not correctly preserved, after a
196 successful authentication. This usually meant that the value preserved by
197 the server_setid option was incorrect.
199 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
201 PH/09 Installed PCRE release 7.0.
203 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
204 run for batched SMTP input. It is now run at the start of every message
205 in the batch. While fixing this I discovered that the process information
206 (output by running exiwhat) was not always getting set for -bs and -bS
207 input. This is fixed, and it now also says "batched" for BSMTP.
209 PH/11 Added control=no_pipelining.
211 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
212 patch, slightly modified), and move the expansion of helo_data till after
213 the connection is made in the smtp transport (so it can use these
216 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
218 PH/14 Added log_selector = +pid.
220 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
222 PH/16 Add ${if forany and ${if forall.
224 PH/17 Added dsn_from option to vary the From: line in DSNs.
226 PH/18 Flush SMTP output before performing a callout, unless control =
227 no_callout_flush is set.
229 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
230 was true (the default) a successful delivery failed to delete the retry
231 item, thus causing premature timeout of the address. The bug is now
234 PH/20 Added hosts_avoid_pipelining to the smtp transport.
236 PH/21 Long custom messages for fakedefer and fakereject are now split up
237 into multiline reponses in the same way that messages for "deny" and
238 other ACL rejections are.
240 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
241 with slight modification.
243 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
244 draft, changing the syntax and factoring some duplicate code".
246 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
247 for deliveries of the second and subsequent messages over the same SMTP
250 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
251 ${reduce, with only minor "tidies".
253 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
255 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
256 expansion side effects.
258 PH/27 When a message times out after an over-quota error from an Exim-imposed
259 quota, the bounce message says "mailbox is full". This message was not
260 being given when it was a system quota that was exceeded. It now should
263 MH/03 Made $recipients available in local_scan(). local_scan() already has
264 better access to the recipient list through recipients_list[], but
265 $recipients can be useful in postmaster-provided expansion strings.
267 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
268 in the case of a MAIL command with additional options following the
269 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
270 were accidentally chopped off.
272 PH/29 SMTP synchronization checks are implemented when a command is read -
273 there is a check that no more input is waiting when there shouldn't be
274 any. However, for some commands, a delay in an ACL can mean that it is
275 some time before the response is written. In this time, more input might
276 arrive, invalidly. So now there are extra checks after an ACL has run for
277 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
278 pipelining has not been advertised.
280 PH/30 MH's patch to allow iscntrl() characters to be list separators.
282 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
283 returned in the SMTP response when smtp_return_error_details was false.
286 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
287 instead of the C I/O that was originally supplied, because problems were
290 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
291 Exim which did not show up earlier: it was assuming that a call to
292 SSL_CTX_set_info_callback() might give an error value. In fact, there is
293 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
294 was a macro that became an assignment, so it seemed to work. This has
295 changed to a proper function call with a void return, hence the compile
296 error. Exim's code has been fixed.
298 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
301 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
303 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
305 PH/37 If a message is not accepted after it has had an id assigned (e.g.
306 because it turns out to be too big or there is a timeout) there is no
307 "Completed" line in the log. When some messages of this type were
308 selected by exigrep, they were listed as "not completed". Others were
309 picked up by some special patterns. I have improved the selection
310 criteria to be more general.
312 PH/38 The host_find_failed option in the manualroute router can now be set
313 to "ignore", to completely ignore a host whose IP address cannot be
314 found. If all hosts are ignored, the behaviour is controlled by the new
315 host_all_ignored option.
317 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
318 homing or because of multiple MX records with /mx) generated more than
319 one IP address, and the following item turned out to be the local host,
320 all the secondary addresses of the first item were incorrectly removed
321 from the list, along with the local host and any following hosts (which
322 is what is supposed to happen).
324 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
325 whoever called Exim into the -H file. In the case of the daemon it was
326 behaving confusingly. When first started, it used values for whoever
327 started the daemon, but after a SIGHUP it used the Exim user (because it
328 calls itself on a restart). I have changed the code so that it now always
331 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
332 message are rejected with the same error (e.g. no authentication or bad
333 sender address), and a DATA command is nevertheless sent (as can happen
334 with PIPELINING or a stupid MUA), the error message that was given to the
335 RCPT commands is included in the rejection of the DATA command. This is
336 intended to be helpful for MUAs that show only the final error to their
339 PH/42 Another patch from the Sieve maintainer.
341 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
342 Eximstats - Fixed some broken HTML links and added missing column headers
344 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
345 columns for Rejects, Temp Rejects, Ham, and Spam rows.
347 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
349 PH/43 Yet another patch from the Sieve maintainer.
351 PH/44 I found a way to check for a TCP/IP connection going away before sending
352 the response to the final '.' that terminates a message, but only in the
353 case where the client has not sent further data following the '.'
354 (unfortunately, this is allowed). However, in many cases there won't be
355 any further data because there won't be any more messages to send. A call
356 to select() can be used: if it shows that the input is "ready", there is
357 either input waiting, or the socket has been closed. An attempt to read
358 the next input character can distinguish the two cases. Previously, Exim
359 would have sent an OK response which the client would never have see.
360 This could lead to message repetition. This fix should cure that, at
361 least in a lot of common cases.
363 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
364 advertised in response to EHLO.
370 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
371 fixed by 4.65/MH/01 (is this a record?) are fixed:
373 (i) An empty string was always treated as zero by the numeric comparison
374 operators. This behaviour has been restored.
376 (ii) It is documented that the numeric comparison operators always treat
377 their arguments as decimal numbers. This was broken in that numbers
378 starting with 0 were being interpreted as octal.
380 While fixing these problems I realized that there was another issue that
381 hadn't been noticed. Values of message_size_limit (both the global option
382 and the transport option) were treated as octal if they started with 0.
383 The documentation was vague. These values are now always treated as
384 decimal, and I will make that clear in the documentation.
390 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
391 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
394 MH/01 Don't check that the operands of numeric comparison operators are
395 integers when their expansion is in "skipping" mode (fixes bug
396 introduced by 4.64-PH/07).
398 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
399 child addresses, Exim now panics and dies. Previously, because the count
400 is held in a short int, deliveries were likely to be lost. As such a
401 large number of recipients for a single message is ridiculous
402 (performance will be very, very poor), I have chosen to impose a limit
403 rather than extend the field.
409 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
410 leftover -K file (the existence of which was triggered by #402).
411 While we were at it, introduced process PID as part of the -K
412 filename. This should rule out race conditions when creating
415 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
416 processing considerably. Previous code took too long for large mails,
417 triggering a timeout which in turn triggers #401.
419 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
420 in the DK code in transports.c. sendfile() is not really portable,
421 hence the _LINUX specificness.
423 TF/01 In the add_headers option to the mail command in an Exim filter,
424 there was a bug that Exim would claim a syntax error in any
425 header after the first one which had an odd number of characters
428 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
429 callout verification, Exim cached a "reject" for the entire domain. This
430 is correct for most verifications, but it is not correct for a recipient
431 verification with use_sender or use_postmaster set, because in that case
432 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
433 case of MAIL FROM:<> rejection from other early rejections (e.g.
434 rejection of HELO). When verifying a recipient using a non-null MAIL
435 address, the cache is ignored if it shows MAIL FROM:<> rejection.
436 Whatever the result of the callout, the value of the domain cache is
437 left unchanged (for any other kind of callout, getting as far as trying
438 RCPT means that the domain itself is ok).
440 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
443 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
444 manifest itself as EPIPE rather than ECONNECT. When tidying away a
445 session, the daemon ignores ECONNECT errors and logs others; it now
446 ignores EPIPE as well.
448 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
449 (quoted-printable decoding).
451 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
452 later the small subsequent patch to fix an introduced bug.
454 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
456 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
458 PH/08 An error is now given if message_size_limit is specified negative.
460 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
461 to be given (somewhat) arbitrary names.
463 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
466 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
467 miscellaneous code fixes
469 PH/10 Added the log_reject_target ACL modifier to specify where to log
472 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
473 hostname. This is wrong, because it relates to the incoming message (and
474 probably the interface on which it is arriving) and not to the outgoing
475 callout (which could be using a different interface). This has been
476 changed to use the value of the helo_data option from the smtp transport
477 instead - this is what is used when a message is actually being sent. If
478 there is no remote transport (possible with a router that sets up host
479 addresses), $smtp_active_hostname is used.
481 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
482 tweaks were necessary in order to get it to work (see also 21 below):
483 (a) The code assumed that strncpy() returns a negative number on buffer
484 overflow, which isn't the case. Replaced with Exim's string_format()
486 (b) There were several signed/unsigned issues. I just did the minimum
487 hacking in of casts. There is scope for a larger refactoring.
488 (c) The code used strcasecmp() which is not a standard C function.
489 Replaced with Exim's strcmpic() function.
490 (d) The code set only $1; it now sets $auth1 as well.
491 (e) A simple test gave the error "authentication client didn't specify
492 service in request". It would seem that Dovecot has changed its
493 interface. Fortunately there's a specification; I followed it and
494 changed what the client sends and it appears to be working now.
496 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
499 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
500 address A is aliased to B and C, where B exists and C does not. Without
501 -v the output is "A verified" because verification stops after a
502 successful redirection if more than one address is generated. However,
503 with -v the child addresses are also verified. Exim was outputting "A
504 failed to verify" and then showing the successful verification for C,
505 with its parentage. It now outputs "B failed to verify", showing B's
506 parentage before showing the successful verification of C.
508 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
509 look up a TXT record in a specific list after matching in a combined
512 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
513 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
514 they consult the DNS. I had assumed they would set it the way they
515 wanted; and indeed my experiments on Linux seem to show that in some
516 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
517 To be on the safe side, however, I have now made the interface to
518 host_find_byname() similar to host_find_bydns(), with an argument
519 containing the DNS resolver options. The host_find_byname() function now
520 sets these options at its start, just as host_find_bydns() does. The smtp
521 transport options dns_qualify_single and dns_search_parents are passed to
522 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
523 of host_find_byname() use the default settings of RES_DEFNAMES
524 (qualify_single) but not RES_DNSRCH (search_parents).
526 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
527 spool_read_header() do less string testing, by means of a preliminary
528 switch on the second character of optional "-foo" lines. (This is
529 overdue, caused by the large number of possibilities that now exist.
530 Originally there were few.) While I was there, I also converted the
531 str(n)cmp tests so they don't re-test the leading "-" and the first
532 character, in the hope this might squeeze out yet more improvement.
534 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
535 flag allowing group syntax was set by the header_syntax check but not
536 turned off, possible causing trouble later; (2) The flag was not being
537 set at all for the header_verify test, causing "group"-style headers to
538 be rejected. I have now set it in this case, and also caused header_
539 verify to ignore an empty address taken from a group. While doing this, I
540 came across some other cases where the code for allowing group syntax
541 while scanning a header line wasn't quite right (mostly, not resetting
542 the flag correctly in the right place). These bugs could have caused
543 trouble for malformed header lines. I hope it is now all correct.
545 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
546 with the "reply" argument non-NULL. The code, however (which originally
547 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
548 but it didn't always do it. This confused somebody who was copying the
549 code for some other use. I have removed all the tests.
551 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
552 feature that was used to support insecure browsers during the U.S. crypto
553 embargo. It requires special client support, and Exim is probably the
554 only MTA that supported it -- and would never use it because real RSA is
555 always available. This code has been removed, because it had the bad
556 effect of slowing Exim down by computing (never used) parameters for the
557 RSA_EXPORT functionality.
559 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
560 authenticator to fail if there's a tab character in the incoming data
561 (there should never be unless someone is messing about, as it's supposed
562 to be base64-encoded). Also added, on Timo's advice, the "secured" option
563 if the connection is using TLS or if the remote IP is the same as the
564 local IP, and the "valid-client-cert option" if a client certificate has
567 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
568 authenticators. This can be used for authorization after authentication
569 succeeds. (In the case of plaintext, it servers for both authentication
572 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
573 if any retry times were supplied.
575 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
576 connection, where there is no client IP address to check. In this
577 situation, the verify now always succeeds.
579 PH/25 Applied John Jetmore's -Mset patch.
581 PH/26 Added -bem to be like -Mset, but loading a message from a file.
583 PH/27 In a string expansion for a processed (not raw) header when multiple
584 headers of the same name were present, leading whitespace was being
585 removed from all of them, but trailing whitespace was being removed only
586 from the last one. Now trailing whitespace is removed from each header
587 before concatenation. Completely empty headers in a concatenation (as
590 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
591 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
593 PH/29 [Removed. This was a change that I later backed out, and forgot to
594 correct the ChangeLog entry (that I had efficiently created) before
595 committing the later change.]
597 PH/30 Exim was sometimes attempting to deliver messages that had suffered
598 address errors (4xx response to RCPT) over the same connection as other
599 messages routed to the same hosts. Such deliveries are always "forced",
600 so retry times are not inspected. This resulted in far too many retries
601 for the affected addresses. The effect occurred only when there were more
602 hosts than the hosts_max_try setting in the smtp transport when it had
603 the 4xx errors. Those hosts that it had tried were not added to the list
604 of hosts for which the message was waiting, so if all were tried, there
605 was no problem. Two fixes have been applied:
607 (i) If there are any address or message errors in an SMTP delivery, none
608 of the hosts (tried or untried) are now added to the list of hosts
609 for which the message is waiting, so the message should not be a
610 candidate for sending over the same connection that was used for a
611 successful delivery of some other message. This seems entirely
612 reasonable: after all the message is NOT "waiting for some host".
613 This is so "obvious" that I'm not sure why it wasn't done
614 previously. Hope I haven't missed anything, but it can't do any
615 harm, as the worst effect is to miss an optimization.
617 (ii) If, despite (i), such a delivery is accidentally attempted, the
618 routing retry time is respected, so at least it doesn't keep
619 hammering the server.
621 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
622 in ${readsocket because some servers need this prod.
624 PH/32 Added some extra debug output when updating a wait-xxx database.
626 PH/33 The hint "could be header name not terminated by colon", which has been
627 given for certain expansion errors for a long time, was not being given
628 for the ${if def:h_colon_omitted{... case.
630 PH/34 The spec says: "With one important exception, whenever a domain list is
631 being scanned, $domain contains the subject domain." There was at least
632 one case where this was not true.
634 PH/35 The error "getsockname() failed: connection reset by peer" was being
635 written to the panic log as well as the main log, but it isn't really
636 panic-worthy as it just means the connection died rather early on. I have
637 removed the panic log writing for the ECONNRESET error when getsockname()
640 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
641 runs only) independently of the message's sender address. This meant
642 that, if the 4xx error was in fact related to the sender, a different
643 message to the same recipient with a different sender could confuse
644 things. In particualar, this can happen when sending to a greylisting
645 server, but other circumstances could also provoke similar problems.
646 I have changed the default so that the retry time for these errors is now
647 based a combination of the sender and recipient addresses. This change
648 can be overridden by setting address_retry_include_sender=false in the
651 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
652 remote server are returned as part of bounce messages. This was not
653 happening for LMTP over a pipe (the lmtp transport), but now it is the
654 same for both kinds of LMTP.
656 PH/38 Despite being documented as not happening, Exim was rewriting addresses
657 in header lines that were in fact CNAMEs. This is no longer the case.
659 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
660 and queue runs started by the daemon processed all messages. This has
661 been fixed so that -R and -S can now usefully be given with -q<time>.
663 PH/40 Import PCRE release 6.7 (fixes some bugs).
665 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
667 PH/42 Give an error if -q is specified more than once.
669 PH/43 Renamed the variables $interface_address and $interface_port as
670 $received_ip_address and $received_port, to make it clear that these
671 values apply to message reception, and not to the outgoing interface when
672 a message is delivered. (The old names remain recognized, of course.)
674 PH/44 There was no timeout on the connect() call when using a Unix domain
675 socket in the ${readsocket expansion. There now is.
677 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
678 be meaningful with "accept".
680 SC/01 Eximstats V1.43
681 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
683 SC/02 Eximstats V1.44
684 Use a glob alias rather than an array ref in the generated
685 parser. This improves both readability and performance.
687 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
688 Collect SpamAssassin and rejection statistics.
689 Don't display local sender or destination tables unless
690 there is data to show.
691 Added average volumes into the top table text output.
693 SC/04 Eximstats V1.46
694 Collect data on the number of addresses (recipients)
695 as well as the number of messages.
697 SC/05 Eximstats V1.47
698 Added 'Message too big' to the list of mail rejection
699 reasons (thanks to Marco Gaiarin).
701 SC/06 Eximstats V1.48
702 Mainlog lines which have GMT offsets and are too short to
703 have a flag are now skipped.
705 SC/07 Eximstats V1.49 (Alain Williams)
706 Added the -emptyok flag.
708 SC/08 Eximstats V1.50
709 Fixes for obtaining the IP address from reject messages.
711 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
712 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
713 whitesspace changes from 4.64-PH/27
715 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
718 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
719 are found, allow negative numbers in numeric criteria)
721 JJ/06 exipick.20061117.2, added new $message_body_missing variable
723 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
724 to match changes made in 4.64-PH/43
726 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
728 PH/47 Put in an explicit test for a DNS lookup of an address record where the
729 "domain" is actually an IP address, and force a failure. This locks out
730 those revolvers/nameservers that support "A-for-A" lookups, in
731 contravention of the specifications.
733 PH/48 When a host name was looked up from an IP address, and the subsequent
734 forward lookup of the name timed out, the host name was left in
735 $sender_host_name, contrary to the specification.
737 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
738 restricted to single-key lookups, Exim was not diagnosing an error if
739 * or *@ was used with a query-style lookup.
741 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
743 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
744 long ago, but noone interested enough thought of it. Let's just say that
745 the "1.1" means that there are some new functions that weren't there at
746 some point in the past.
748 PH/51 Error processing for expansion failure of helo_data from an smtp
749 transport during callout processing was broken.
751 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
752 tested/used via the -bh/-bhc/-bs options.
754 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
755 bug, fixed in subsequent PCRE releases).
757 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
758 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
760 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
766 SC/01 Use a glob alias rather than an array ref in eximstats generated
767 parser. This improves both readability and performance.
769 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
770 Don't display local sender or destination tables in eximstats unless
771 there is data to show.
772 Added average volumes into the eximstats top table text output.
774 SC/03 Collect data on the number of addresses (recipients) as well
775 as the number of messages in eximstats.
777 TF/01 Correct an error in the documentation for the redirect router. Exim
778 does (usually) call initgroups() when daemonizing.
780 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
781 with consistent privilege compared to when running as a daemon.
783 TF/03 Note in the spec that $authenticated_id is not set for local
784 submissions from trusted users.
786 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
787 Thanks to Dean Brooks <dean@iglou.com> for the patch.
789 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
790 by adding some example configuration directives to the default
791 configuration file. A little bit of work is required to uncomment the
792 directives and define how usernames and passwords are checked, but
793 there is now a framework to start from.
795 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
796 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
797 without this. I don't know how relevant this is to other LDAP libraries.
799 PH/02 Add the verb name to the "unknown ACL verb" error.
801 PH/03 Magnus Holmgren's patch for filter_prepend_home.
803 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
805 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
806 directory not expanded when it should be if an expanded home directory
807 was set for the address (which is overridden by the transport).
809 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
812 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
813 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
814 because it is too late at that time, and has no effect.
816 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
817 security issue with \' (bugzilla #107). I could not use the
818 PQescapeStringConn() function, because it needs a PGconn value as one of
821 PH/08 When testing addresses using -bt, indicate those final addresses that
822 are duplicates that would not cause an additional delivery. At least one
823 person was confused, thinking that -bt output corresponded to deliveries.
824 (Suppressing duplicates isn't a good idea as you lose the information
825 about possibly different redirections that led to the duplicates.)
827 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
828 systems where poll() doesn't work, in particular OS X.
830 PH/10 Added more information to debugging output for retry time not reached.
832 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
833 operations in malware.c.
835 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
838 PH/13 If write_rejectlog was set false when logging was sent to syslog with
839 syslog_duplication set false, log lines that would normally be written
840 both the the main log and to the reject log were not written to syslog at
843 PH/14 In the default configuration, change the use of "message" in ACL warn
844 statements to "add_header".
846 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
847 not followed by a command (e.g. "seen endif").
849 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
850 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
853 PH/17 Added extra conditions to the default value of delay_warning_condition
857 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
858 { match{$h_precedence:}{(?i)bulk|list|junk} } \
859 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
862 The Auto-Submitted: and various List- headers are standardised, whereas I
863 don't think Precedence: ever was.
865 PH/18 Refactored debugging code in route_finduser() to show more information,
866 in particular, the error code if getpwnam() issues one.
868 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
869 This is apparently needed in addition to the PH/07 change above to avoid
870 any possible encoding problems.
872 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
873 but not after initializing Perl.
875 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
876 output them only if debugging. By default they are written stderr,
877 apparently, which is not desirable.
879 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
882 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
885 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
887 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
888 authenticated or an ident call has been made. Suppress the default
889 values for $authenticated_id and $authenticated_sender (but permit -oMai
890 and -oMas) when testing with -bh.
892 PH/24 Re-jigged the order of the tests in the default configuration so that the
893 tests for valid domains and recipients precede the DNS black list and CSA
894 tests, on the grounds that those ones are more expensive.
896 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
897 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
898 This bug exists in every version of Exim that I still have, right back to
901 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
902 However, an attempt to turn on case-sensitivity in a regex key by
903 including (?-i) didn't work because the subject string was already
904 lowercased, and the effects were non-intuitive. It turns out that a
905 one-line patch can be used to allow (?-i) to work as expected.
911 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
912 other effects) broke the use of negated acl sub-conditions.
914 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
917 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
918 "Deny" causes Exim to reject the incoming connection with a 554 error.
919 Unfortunately, if there is a major crisis, such as a disk failure,
920 tcp-wrappers gives "deny", whereas what one would like would be some
921 kind of temporary error. A kludge has been added to help with this.
922 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
923 554 error is used if errno is still zero or contains ENOENT (which occurs
924 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
927 PH/03 Add -lutil to the default FreeBSD LIBS setting.
929 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
930 errors. Otherwise a message that provokes a temporary error (when other
931 messages do not) can cause a whole host to time out.
933 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
934 addresses were routed directly to files or pipes from a redirect router.
935 File deliveries just didn't batch; pipe deliveries might have suffered
938 PH/06 A failure to get a lock for a hints database would erroneously always say
939 "Failed to get write lock", even when it was really a read lock.
941 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
942 of 0600. This has been changed to use the value of the lockfile_mode
943 option (which defaults to 0600).
945 PH/08 Applied small patch from the Sieve maintainer.
947 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
948 folder from quota calculations, a direct delivery into this folder messed
949 up the contents of the maildirsize file. This was because the regex was
950 used only to exclude .Trash (or whatever) when the size of the mailbox
951 was calculated. There was no check that a delivery was happening into an
952 excluded directory. This bug has been fixed by ignoring all quota
953 processing for deliveries into excluded directories.
955 PH/10 Added the maildirfolder_create_regex option to appendfile.
961 PH/01 The code for finding all the local interface addresses on a FreeBSD
962 system running IPv6 was broken. This may well have applied to all BSD
963 systems, as well as to others that have similar system calls. The broken
964 code found IPv4 interfaces correctly, but gave incorrect values for the
965 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
966 that it would not match correctly against @[] and not recognize the IPv6
969 PH/02 The ipliteral router was not recognizing addresses of the form user@
970 [ipv6:....] because it didn't know about the "ipv6:" prefix.
972 PH/03 Added disable_ipv6.
974 PH/04 Changed $reply_address to use the raw form of the headers instead of the
975 decoded form, because it is most often used to construct To: headers
976 lines in autoreplies, and the decoded form may well be syntactically
977 invalid. However, $reply_address has leading white space removed, and all
978 newlines turned into spaces so that the autoreply transport does not
981 PH/05 If group was specified without a user on a router, and no group or user
982 was specified on a transport, the group from the router was ignored.
984 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
985 for visible compile-time settings that can be used to change these
986 numbers, for those that want even more. Backwards compatibility with old
987 spool files has been maintained. However, going back to a previous Exim
988 release will lost any variables that are in spool files.
990 PH/07 Two small changes when running in the test harness: increase delay when
991 passing a TCP/IP connection to a new process, in case the original
992 process has to generate a bounce, and remove special handling of
993 127.0.0.2 (sic), which is no longer necessary.
995 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
996 be the same on different OS.
998 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1001 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1002 whether --show-vars was specified or not
1004 JJ/02 exipick: Added support for new ACL variable spool format introduced
1007 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1008 syntactically invalid From: or Reply-to: line, and a filter used this to
1009 generate an autoreply, and therefore failed to obtain an address for the
1010 autoreply, Exim could try to deliver to a non-existent relative file
1011 name, causing unrelated and misleading errors. What now happens is that
1012 it logs this as a hard delivery error, but does not attempt to create a
1015 PH/11 The exinext utility has a -C option for testing purposes, but although
1016 the given file was scanned by exinext itself; it wasn't being passed on
1017 when Exim was called.
1019 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1020 an end-of-file indication when reading a command response.
1022 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1023 compiled. In many other places in Exim, IPv6 addresses are always
1024 recognized, so I have changed this. It also means that IPv4 domain
1025 literals of the form [IPV4:n.n.n.n] are now always recognized.
1027 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1028 used if the router is not running as root, for example, when verifying at
1029 ACL time, or when using -bh. The debugging output from this situation was
1030 non-existent - all you got was a failure to exec. I have made two
1033 (a) Failures to set uid/gid, the current directory, or a process leader
1034 in a subprocess such as that created by queryprogram now generate
1035 suitable debugging ouput when -d is set.
1037 (b) The queryprogram router detects when it is not running as root,
1038 outputs suitable debugging information if -d is set, and then runs
1039 the subprocess without attempting to change uid/gid.
1041 PH/15 Minor change to Makefile for building test_host (undocumented testing
1044 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1045 additional section of a DNS packet that returns MX or SRV records.
1046 Instead, it always explicitly searches for A/AAAA records. This avoids
1047 major problems that occur when a DNS server includes only records of one
1048 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1049 fixed another bug: if SRV records were looked up and the corresponding
1050 address records were *not* found in the additional section, the port
1051 values from the SRV records were lost.
1053 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1054 using the correct key (the original address) when searching the retry
1055 rules in order to find which one to use for generating the retry hint.
1057 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1058 adding the default one. Similarly, if it contains a Reply-To: header, the
1059 errors_reply_to option, if set, is not used.
1061 PH/19 When calculating a retry time, Exim used to measure the "time since
1062 failure" by looking at the "first failed" field in the retry record. Now
1063 it does not use this if it is later than than the arrival time of the
1064 message. Instead it uses the arrival time. This makes for better
1065 behaviour in cases where some deliveries succeed, thus re-setting the
1066 "first failed" field. An example is a quota failure for a huge message
1067 when small messages continue to be delivered. Without this change, the
1068 "time since failure" will always be short, possible causing more frequent
1069 delivery attempts for the huge message than are intended.
1070 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1072 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1073 $1, $2, $3) because the numerical variables can be reset during some
1074 expansion items (e.g. "match"), thereby losing the authentication data.
1076 PH/21 Make -bV show the size of off_t variables so that the test suite can
1077 decide whether to run tests for quotas > 2G.
1079 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1080 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1081 filecount value is greater than 2G or if a quota value is greater than 2G
1082 on a system where the size of off_t is not greater than 4, a panic error
1085 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1086 never match. The debug and -bh output now contains an explicit error
1087 message indicating a malformed IPv4 address or mask.
1089 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1090 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1091 PH/23 above applies.
1093 PH/25 Do not write to syslog when running in the test harness. The only
1094 occasion when this arises is a failure to open the main or panic logs
1095 (for which there is an explicit test).
1097 PH/26 Added the /no_tell option to "control=freeze".
1099 PH/27 If a host name lookup failed very early in a connection, for example, if
1100 the IP address matched host_lookup and the reverse lookup yielded a name
1101 that did not have a forward lookup, an error message of the form "no IP
1102 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1103 could be logged. Now it outputs the IP address instead of "NULL".
1105 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1106 allows the sender and the authenticated sender to be set when
1107 submitting a message from within Exim. Since child_open_exim() is
1108 documented for local_scan(), the new function should be too.
1110 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1111 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1112 results in an empty string is now treated as unset.
1114 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1116 PH/31 Added sender_verify_fail logging option.
1118 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1119 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1120 tidied the source and removed it altogether.
1122 PH/33 When a queue run was abandoned because the load average was too high, a
1123 log line was always written; now it is written only if the queue_run log
1124 selector is set. In addition, the log line for abandonment now contains
1125 information about the queue run such as the pid. This is always present
1126 in "start" and "stop" lines but was omitted from the "abandon" line.
1128 PH/34 Omit spaces between a header name and the colon in the error message that
1129 is given when verify = headers_syntax fails (if there are lots of them,
1130 the message gets confusing).
1132 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1133 names, as there are now some PTR records that contain slashes. This check
1134 is only to protect against broken name servers that fall over on strange
1135 characters, so the fact that it applies to all lookups doesn't matter.
1137 PH/36 Now that the new test suite is complete, we can remove some of the
1138 special code in Exim that was needed for the old test suite. For example,
1139 sorting DNS records because real resolvers return them in an arbitrary
1140 order. The new test suite's fake resolver always returns records in the
1143 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1144 bounces) except when queue_only is set, to avoid logging races between
1145 the different processes.
1147 PH/38 Panic-die if .include specifies a non-absolute path.
1149 PH/39 A tweak to the "H" retry rule from its user.
1151 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1152 a label. They prevented compilation on older perls.
1154 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1155 a warning to be raised on newish perls.
1157 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1158 on queue. Changes to match documented behaviour of showing count of
1159 messages matching specified criteria.
1161 PH/40 Changed the default ident timeout from 30s to 5s.
1163 PH/41 Added support for the use of login_cap features, on those BSD systems
1164 that have them, for controlling the resources used by pipe deliveries.
1166 PH/42 The content-scanning code uses fopen() to create files in which to put
1167 message data. Previously it was not paying any attention to the mode of
1168 the files. Exim runs with umask(0) because the rest of the code creates
1169 files with open(), and sets the required mode explicitly. Thus, these
1170 files were ending up world-writeable. This was not a big issue, because,
1171 being within the spool directory, they were not world-accessible. I have
1172 created a function called modefopen, which takes an additional mode
1173 argument. It sets umask(777), creates the file, chmods it to the required
1174 mode, then resets the umask. All the relevant calls to fopen() in the
1175 content scanning code have been changed to use this function.
1177 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1178 to 24 hours. This avoids potential overflow problems when processing G
1179 and H retry rules. I suspect nobody ever tinkers with this value.
1181 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1183 PH/45 When the plaintext authenticator is running as a client, the server's
1184 challenges are checked to ensure they are valid base64 strings. By
1185 default, the authentication attempt is cancelled if an invalid string is
1186 received. Setting client_ignore_invalid_base64 true ignores these errors.
1187 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1188 they are received. Thus, the responses can be made to depend on the
1189 challenges. If an invalid string is ignored, an empty string is placed in
1192 PH/46 Messages that are created by the autoreply transport now contains a
1193 References: header, in accordance with RFCs 2822 and 3834.
1195 PH/47 Added authenticated_sender_force to the smtp transport.
1197 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1199 PH/49 Installed latest patch from the Sieve maintainer.
1201 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1202 was also set, the appendfile transport was unnecessarily scanning a
1203 directory of message files (e.g. for maildir delivery) to find the count
1204 of files (along with the size), even though it did not need this
1205 information. It now does the scan only if it needs to find either the
1206 size of the count of files.
1208 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1210 PH/52 Two bugs concerned with error handling when the smtp transport is
1213 (i) Exim was not creating retry information for temporary errors given
1214 for individual recipients after the DATA command when the smtp transport
1215 was used in LMTP mode. This meant that they could be retried too
1216 frequently, and not timed out correctly.
1218 (ii) Exim was setting the flag that allows error details to be returned
1219 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1220 recipients that were returned after the DATA command.
1222 PH/53 This is related to PH/52, but is more general: for any failing address,
1223 when detailed error information was permitted to be returned to the
1224 sender, but the error was temporary, then after the final timeout, only
1225 "retry timeout exceeded" was returned. Now it returns the full error as
1226 well as "retry timeout exceeded".
1228 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1229 do this, and (what is worse) MTAs that accept it.
1231 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1232 will now be deprecated.
1234 PH/56 New os.c-cygwin from the Cygwin maintainer.
1236 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1237 formats (previously only available in exim formats via -bpr, -bpru,
1238 and -bpra. Now also available in native and exiqgrep formats)
1240 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1241 with very large, slow to parse queues
1243 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1245 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1247 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1248 responses to authentication challenges, though it was showing the
1249 challenges; (ii) I've removed the CR characters from the debug output for
1252 PH/58 Allow for the insertion of a newline as well as a space when a string
1253 is turned into more than one encoded-word during RFC 2047 encoding. The
1254 Sieve code now uses this.
1256 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1257 data_4xx, lost_connection, tls_required.
1259 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1260 message was being sent as an SMTP response.
1262 PH/61 Add -l and -k options to exicyclog.
1264 PH/62 When verifying, if an address was redirected to one new address, so that
1265 verification continued, and the new address failed or deferred after
1266 having set something in $address_data, the value of $address_data was not
1267 passed back to the ACL. This was different to the case when no
1268 redirection occurred. The value is now passed back in both cases.
1270 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1271 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1272 use login_cap.h, so on its own it isn't the distinguishing feature. The
1273 new name refers directly to the setclassresources() function.
1275 PH/65 Added configuration files for NetBSD3.
1277 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1279 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1280 is preferred over IPv4.
1282 PH/68 The bounce_return_message and bounce_return_body options were not being
1283 honoured for bounces generated during the reception of non-SMTP messages.
1284 In particular, this applied to messages rejected by the ACL. This bug has
1285 been fixed. However, if bounce_return_message is true and bounce_return_
1286 body is false, the headers that are returned for a non-SMTP message
1287 include only those that have been read before the error was detected.
1288 (In the case of an ACL rejection, they have all been read.)
1290 PH/69 The HTML version of the specification is now built in a directory called
1291 spec_html instead of spec.html, because the latter looks like a path with
1292 a MIME-type, and this confuses some software.
1294 PH/70 Catch two compiler warnings in sieve.c.
1296 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1297 function verify_get_ident() calls ip_connect() to connect a socket, but
1298 if the "connect()" function timed out, ip_connect() used to close the
1299 socket. However, verify_get_ident() also closes the socket later, and in
1300 between Exim writes to the log, which may get opened at this point. When
1301 the socket was closed in ip_connect(), the log could get the same file
1302 descriptor number as the socket. This naturally causes chaos. The fix is
1303 not to close the socket in ip_connect(); the socket should be closed by
1304 the function that creates it. There was only one place in the code where
1305 this was missing, in the iplookup router, which I don't think anybody now
1306 uses, but I've fixed it anyway.
1308 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1309 well as to direct DNS lookups. Otherwise the handling of names in host
1310 lists is inconsistent and therefore confusing.
1316 PH/01 Two changes to the default runtime configuration:
1318 (1) Move the checks for relay_from_hosts and authenticated clients from
1319 after to before the (commented out) DNS black list checks.
1321 (2) Add control=submission to the relay_from_hosts and authenticated
1322 clients checks, on the grounds that messages accepted by these
1323 statements are most likely to be submissions.
1325 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1327 (1) Generate an error if the third argument for the ${prvs expansion is
1330 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1333 (3) Reset the variables that are obtained from the first argument of
1334 ${prvscheck and used in the second argument before leaving the code,
1335 because their memory is reclaimed, so using them afterwards may do
1338 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1339 one (it's much easier than Tom thought :-).
1341 (5) Because of (4), we can now allow for the use of $prvscheck_result
1342 inside the third argument.
1344 PH/03 For some reason, the default setting of PATH when running a command from
1345 a pipe transport was just "/usr/bin". I have changed it to
1348 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1349 anything to be listed in the output from -bV.
1351 PH/05 When a filter generated an autoreply, the entire To: header line was
1352 quoted in the delivery log line, like this:
1354 => >A.N.Other <ano@some.domain> <original@ddress> ...
1356 This has been changed so that it extracts the operative address. There
1357 may be more than one such address. If so, they are comma-separated, like
1360 => >ano@some.domain,ona@other.domain <original@ddress> ...
1362 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1363 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1364 not being looked up in the rDNS to get a host name, Exim was showing the
1365 IP address twice in Received: lines, even though the IP addresses were
1366 identical. For example:
1368 Received: from [1.2.3.4] (helo=[1.2.3.4])
1370 However, if the real host name was known, it was omitting the HELO data
1371 if it matched the actual IP address. This has been tidied up so that it
1372 doesn't show the same IP address twice.
1374 PH/07 When both +timestamp and +memory debugging was on, the value given by
1375 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1376 called by the debug printing, thereby overwriting the timestamp buffer.
1377 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1379 PH/08 When the original message was included in an autoreply transport, it
1380 always said "this is a copy of the message, including all the headers",
1381 even if body_only or headers_only was set. It now gives an appropriate
1384 PH/09 Applied a patch from the Sieve maintainer which:
1386 o fixes some comments
1387 o adds the (disabled) notify extension core
1388 o adds some debug output for the result of if/elsif tests
1389 o points to the current vacation draft in the documentation
1390 and documents the missing references header update
1394 o fixes a bug in processing the envelope test (when testing
1395 multiple envelope elements, the last element determinted the
1398 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1399 Electronic Mail") by including:
1401 Auto-submitted: auto-generated
1403 in the messages that it generates (bounce messages and others, such as
1404 warnings). In the case of bounce messages for non-SMTP mesages, there was
1405 also a typo: it was using "Auto_submitted" (underscore instead of
1406 hyphen). Since every message generated by Exim is necessarily in response
1407 to another message, thes have all been changed to:
1409 Auto-Submitted: auto-replied
1411 in accordance with these statements in the RFC:
1413 The auto-replied keyword:
1415 - SHOULD be used on messages sent in direct response to another
1416 message by an automatic process,
1418 - MUST NOT be used on manually-generated messages,
1420 - MAY be used on Delivery Status Notifications (DSNs) and Message
1421 Disposition Notifications (MDNs),
1423 - MUST NOT be used on messages generated by automatic or periodic
1424 processes, except for messages which are automatic responses to
1427 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1428 to the default Received: header definition.
1430 PH/12 Added log selector acl_warn_skipped (default on).
1432 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1433 variables because (a) they are in the wrong storage pool and (b) even if
1434 they were copied, it wouldn't work properly because of the caching.
1436 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1437 checking when decoding. Apparently there are clients that generate
1438 overlong encoded strings. Why am I not surprised?
1440 PH/15 If the first argument of "${if match_address" was not empty, but did not
1441 contain an "@" character, Exim crashed. Now it writes a panic log message
1442 and treats the condition as false.
1444 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1446 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1447 extension "envelope-auth". The code is finished and in agreement with
1448 other implementations, but there is no documentation so far and in fact,
1449 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1450 not changing the active code.
1452 Print executed "if" and "elsif" statements when debugging is used. This
1453 helps a great deal to understand what a filter does.
1455 Document more things not specified clearly in RFC3028. I had all this
1456 sorted out, when out of a sudden new issues came to my mind. Oops."
1458 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1461 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1462 canonical form (as documented). However, after a host name lookup from
1463 the IP address, check_host() was doing a simple string comparison with
1464 addresses acquired from the DNS when checking that the found name did
1465 have the original IP as one of its addresses. Since any found IPv6
1466 addresses are likely to be in abbreviated form, the comparison could
1467 fail. Luckily, there already exists a function for doing the comparison
1468 by converting both addresses to binary, so now that is used instead of
1469 the text comparison.
1471 PH/20 There was another similar case to PH/19, when a complete host name was
1472 given in a host list; looking up its IP address could give an abbreviated
1473 form, whereas the current host's name might or might not be abbreviated.
1474 The same fix has been applied.
1480 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1481 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1484 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1485 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1487 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1489 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1490 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1491 Also, for "Auto-Submitted", treat anything other than "no" as
1492 non-personal, in accordance with RFC 3834. (Previously it treated
1493 anything starting "auto-" as non-personal.)
1495 TF/01 The control=submission/name=... option had a problem with syntax
1496 errors if the name included a slash character. The /name= option
1497 now slurps the rest of the string, so it can include any characters
1498 but it must come last in the list of options (after /sender_retain
1501 PH/05 Some modifications to the interface to the fake nameserver for the new
1509 TK/01 Added the "success_on_redirect" address verification option. See
1510 NewStuff for rationale and an example.
1512 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1514 PH/02 Patch to exigrep to allow it to work on syslog lines.
1516 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1517 fread() to skip over the body file's header line, because in Cygwin the
1518 header line is locked and is inaccessible.
1520 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1521 co-exist for some time) to make it clear that it is the Exim ID that is
1522 referenced, not the Message-ID: header line.
1524 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1525 string_format() function, because snprintf() does not exist on all
1528 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1529 ${stat: expansion item.
1531 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1532 protocol synchronization error", to keep the pedants happy.
1534 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1535 well as for IRIX systems, when gcc is being used. See the host.c source
1538 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1540 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1543 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1544 transport and to the smtp transport in LMTP mode.
1546 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1548 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1549 run a filter in a subprocess. This could lead to confusion in subsequent
1550 lookups in the parent process. There should also be a search_tidyup() at
1551 the end of the subprocess.
1553 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1554 only if the host matched helo_try_verify_hosts, which caused the
1555 verification to occur when the EHLO/HELO command was issued. The ACL just
1556 tested the remembered result. Now, if a previous verification attempt has
1557 not happened, "verify = helo" does it there and then.
1559 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1561 TK/03 Fix log output including CR from clamd.
1563 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1564 exist provoked a memory error which could cause a segfault.
1566 PH/15 Installed PCRE 6.2
1568 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1570 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1571 of the problem. Specifically, suggested +O2 rather than +O1 for the
1574 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1576 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1577 record, the port was not correctly logged when the outgoing_port log
1578 selector was set (it logged the transort's default port).
1580 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1581 fallback_hosts, and "hosts" in the smtp transport.
1583 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1584 host errors such as "Connection refused".
1586 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1587 authentication with radiusclient 0.4.9:
1589 - Error returned from rc_read_config was caught wrongly
1590 - Username/password not passed on to radius server due to wrong length.
1592 The presumption is that some radiusclient API changes for 4.51/PH/17
1593 were not taken care of correctly. The code is still untested by me (my
1594 Linux distribution still has 0.3.2 of radiusclient), but it was
1595 contributed by a Radius user.
1597 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1598 expanding the "port" option of the smtp transport.
1600 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1601 while reading a MIME header. Thanks to Tom Hughes for a patch.
1603 PH/24 Include config.h inside local_scan.h so that configuration settings are
1606 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1607 that in an ACL for RCPT (for example), you can examine exactly what was
1610 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1611 commands, but it was not correctly comparing the address with the actual
1612 client host address. Thus, it would show the EHLO address in Received:
1613 header lines when this was not necessary.
1615 PH/27 Added the % operator to ${eval:}.
1617 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1618 it should be ignoring failures (because with -C, for example, it has lost
1619 privilege). It wasn't ignoring creation failures other than "already
1622 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1625 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1626 because an input line was too long, either on its own, or by virtue of
1627 too many continuations, the temporary file was not being removed, and the
1628 return code was incorrect.
1630 PH/31 Missing "BOOL" in function definition in filtertest.c.
1632 PH/32 Applied Sieve patches from the maintainer.
1634 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1636 PH/33 Added "verify = not_blind".
1638 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1639 Local/Makefile (with some defaults set). These are used in built scripts
1640 such as exicyclog, but they have never been used in the exim_install
1641 script (though there are many overriding facilities there). I have
1642 arranged that the exim_install script now takes note of these two
1645 PH/35 Installed configuration files for Dragonfly.
1647 PH/36 When a locally submitted message by a trusted user did not contain a
1648 From: header, and the sender address was obtained from -f or from an SMTP
1649 MAIL command, and the trusted user did not use -F to supply a sender
1650 name, $originator_name was incorrectly used when constructing a From:
1651 header. Furthermore, $originator_name was used for submission mode
1652 messages from external hosts without From: headers in a similar way,
1653 which is clearly wrong.
1655 PH/37 Added control=suppress_local_fixups.
1657 PH/38 When log_selector = +received_sender was set, and the addition of the
1658 sender made the log line's construction buffer exactly full, or one byte
1659 less than full, an overflow happened when the terminating "\n" was
1662 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1663 when the result of a list match is failure because a DNS lookup failed.
1665 PH/40 RM_COMMAND is now used in the building process.
1667 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1668 the "build-* directories that it finds.
1670 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1671 address in a domain literal was a prefix of an interface address.
1673 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1674 when verifying a sender address, unless rewrite_headers is false.
1676 PH/44 Wrote a long comment about why errors_to addresses are verified as
1677 recipients, not senders.
1679 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1680 the ratelimit ACL was added.
1682 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1684 PH/46 Added extra information about PostgreSQL errors to the error string.
1686 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1687 avoiding the need to install special zones in a real server. This is
1688 backwards compatible; if it can't find the fake resolver, it drops back.
1689 Thus, both old and new test suites can be run.
1691 TF/02 Added util/ratelimit.pl
1693 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1694 clock is set back in time.
1696 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1697 Candler <B.Candler@pobox.com>.
1699 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1700 OK for addresses that are the result of redirections.
1702 PH/48 A number of further additions for the benefit of the new test suite,
1703 including a fake gethostbyname() that interfaces to the fake DNS resolver
1706 TF/06 The fix for widen_domains has also been applied to qualify_single and
1707 search_parents which are the other dnslookup options that can cause
1710 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1713 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1715 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1716 service" if it encountered an MX record with an empty target hostname.
1717 The message is now "an MX or SRV record indicated no SMTP service".
1719 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1720 if the DNS of the sending site is misconfigured. This is quite a
1721 common situation. This change restores the behaviour of treating a
1722 helo verification defer as a failure.
1724 PH/51 If self=fail was set on a router, the bounce message did not include the
1725 actual error message.
1731 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1733 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1734 command itself ended in error, the underlying message about the transport
1735 filter timeout was being overwritten with the pipe command error. Now the
1736 underlying error message should be appended to the second error message.
1738 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1740 PH/02 Reduce the amount of output that "make" produces by default. Full output
1741 can still be requested.
1743 PH/03 The warning log line about a condition test deferring for a "warn" verb
1744 was being output only once per connection, rather than after each
1745 occurrence (because it was using the same function as for successful
1746 "warn" verbs). This seems wrong, so I have changed it.
1748 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1749 it should not have, which might have caused a crash in the right
1750 circumstances, but probably never did.
1752 PH/04 Installed a modified version of Tony Finch's patch to make submission
1753 mode fix the return path as well as the Sender: header line, and to
1754 add a /name= option so that you can make the user's friendly name appear
1757 TF/03 Added the control = fakedefer ACL modifier.
1759 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1760 Mark Lowes for thorough testing.
1762 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1764 TK/03 Merged latest SRS patch from Miles Wilton.
1766 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1767 with the definition in sysexits.h (which is #included earlier).
1768 Fortunately, Exim does not actually use EX_OK. The code used to try to
1769 preserve the sysexits.h value, by assumimg that macro definitions were
1770 scanned for macro replacements. I have been disabused of this notion,
1771 so now the code just undefines EX_OK before #including unistd.h.
1773 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1774 in the smtp transport. When a block could not be written in a single
1775 write() function, the timeout was being re-applied to each part-write.
1776 This seems wrong - if the receiver was accepting one byte at a time it
1777 would take for ever. The timeout is now adjusted when this happens. It
1778 doesn't have to be particularly precise.
1780 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1781 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1783 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1784 without a domain if the check to <postmaster@domain> fails.
1786 SC/01 Eximstats: added -xls and the ability to specify output files
1787 (patch written by Frank Heydlauf).
1789 SC/02 Eximstats: use FileHandles for outputing results.
1791 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1793 SC/04 Eximstats: fixed display of large numbers with -nvr option
1795 SC/05 Eximstats: fixed merging of reports with empty tables.
1797 SC/06 Eximstats: added the -include_original_destination flag
1799 SC/07 Eximstats: removed tabs and trailing whitespace.
1801 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1803 TK/06 MBOX spool code: Add real "From " MBOX separator line
1804 so the .eml file is really in mbox format (even though
1805 most programs do not really care). Patch from Alex Miller.
1807 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1808 The latter is generated from $received_to and is only set if the
1809 message has one envelope recipient. SA can use these headers,
1810 obviously out-of-the-box. Patch from Alex Miller.
1812 PH/08 The ${def test on a variable was returning false if the variable's
1813 value was "0", contrary to what the specification has always said!
1814 The result should be true unless the variable is empty.
1816 PH/09 The syntax error of a character other than { following "${if
1817 def:variable_name" (after optional whitespace) was not being diagnosed.
1818 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1819 accidental colon was present, for example, could give incorrect results.
1821 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1822 result is used (not including appendfile, where other changes are about
1825 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1826 This involved changing a lot of size variables from int to off_t. It
1827 should work with maildirs and everything.
1829 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1830 spamd dying while we are connected to it.
1832 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1835 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1836 function for BDB 4.3.
1838 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1840 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1841 the sources for PCRE have been re-arranged and more files are now
1844 PH/15 The code I had for printing potentially long long variables in PH/11
1845 above was not the best (it lost precision). The length of off_t variables
1846 is now inspected at build time, and an appropriate printing format (%ld
1847 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1848 to be "long long int" or "long int". This is needed for the internal
1849 formatting function string_vformat().
1851 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1852 the configuration file to be ":syslog", then the script "guesses" where
1853 the logs files are, rather than using the compiled in default. In our
1854 case the guess is not the same as the compiled default, so the script
1855 suddenly stopped working when I started to use syslog. The patch checks
1856 to see if log_file_path is "". If so, it attempts to read it from exim
1857 with no configuration file to get the compiled in version, before it
1858 falls back to the previous guessing code."
1860 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1861 implementing BATV in an Exim configuration. See NewStuff for the gory
1864 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1865 Makefile that are specific to HP-UX.
1867 PH/18 If the "use_postmaster" option was set for a recipient callout together
1868 with the "random" option, the postmaster address was used as the MAIL
1869 FROM address for the random test, but not for the subsequent recipient
1870 test. It is now used for both.
1872 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1873 patch removes a few documentation additions to RFC 3028, because the
1874 latest draft now contains them. It adds the new en;ascii-case comparator
1875 and a new error check for 8bit text in MIME parts. Comparator and
1876 require names are now matched exactly. I enabled the subaddress
1877 extension, but it is not well tested yet (read: it works for me)."
1879 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1880 rework some of the code of TK/09 above to avoid the hardwired use of
1881 "%lld" and "long long". Replaced the call to snprintf() with a call to
1884 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1885 records point to non-existent hosts", "retry timeout exceeded", and
1886 "retry time not reached for any host after a long failure period".
1888 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1889 experimental DomainKeys support:
1891 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1892 (2) On an error such as an illegally used "control", the wrong name for
1893 the control was given.
1895 These problems did NOT occur unless DomainKeys support was compiled.
1897 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1899 PH/24 Added ${if match_ip condition.
1901 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1902 databases so that it will be absolutely obvious if a crash occurs in the
1903 DB library. This is a regular occurrence (often caused by mis-matched
1906 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1907 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1908 picked up on a user's system that detects such things. There doesn't seem
1909 to be a gcc warning option for this - only an attribute that has to be
1910 put on the function's prototype. It seems that in Fedora Core 4 they have
1911 set this on a number of new functions. No doubt there will be more in due
1914 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1915 specify a transport. However, if an address that was verified by such a
1916 router was the subject of a callout, Exim crashed because it tried to
1917 read the rcpt_include_affixes from the non-existent transport. Now it
1918 just assumes that the setting of that option is false. This bug was
1919 introduced by 4.51/PH/31.
1921 PH/28 Changed -d+all to exclude +memory, because that information is very
1922 rarely of interest, but it makes the output a lot bigger. People tend to
1923 do -d+all out of habit.
1925 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1926 code in os-type was giving problems when libc.so lives in lib64, like on
1929 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1930 aren't the modern standard, and it seems that some systems' include files
1931 don't always have them. Exim was already checking for some of the newer
1932 ones like T_AAAA, and defining it itself. I've added checks for all the
1933 record types that Exim uses.
1935 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1936 not automatically generating a new one, as it is supposed to. This
1937 prevented TLS from working. If the file did exist, but contained invalid
1938 data, a new version was generated, as expected. It was only the case of a
1939 non-existent file that was broken.
1941 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1942 with a change in libdomainkeys > 0.64.
1944 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1945 from DNS. If the selector record carries the flag, it now has
1946 precedence over the domain-wide flag.
1948 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1950 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1951 the use of a port name that isn't defined in /etc/services) occurred, the
1952 message was deferred as in a normal delivery, and thus remained on the
1953 spool, instead of being failed because of the mua_wrapper setting. This
1954 is now fixed, and I tidied up some of the mua_wrapper messages at the
1957 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1958 the messages in a hash of arrays rather than using individual hashes.
1959 This is a bit cleaner and results in dramatic memory savings, albeit
1960 at a slight CPU cost.
1962 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1963 as requested by Marc Sherman.
1965 SC/10 Eximstats: added histograms for user specified patterns as requested
1968 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1970 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1971 fopen() in the content-scanning modules that did not already have it.
1977 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1978 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1980 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1982 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1984 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1985 to test Sieve filters that use "vacation".
1987 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1988 that changes the way the GnuTLS parameters are stored in the cache file.
1989 The new format can be generated externally. For backward compatibility,
1990 if the data in the cache doesn't make sense, Exim assumes it has read an
1991 old-format file, and it generates new data and writes a new file. This
1992 means that you can't go back to an older release without removing the
1995 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1996 work if there are any delivery delays because "one_time" forces the
1997 parent to be marked "delivered", so its unseen clone is never tried
1998 again. For this reason, Exim now forbids the simultaneous setting of
2001 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2002 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2003 entry if you want to know the details. The fix, however, neglected to
2004 consider the case when local delivery batching is involved. The test for
2005 "previously delivered" was not happening when checking to see if an
2006 address could be batched with a previous (undelivered) one; under
2007 certain circumstances this could lead to multiple deliveries to the same
2010 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2011 in its include files, and this causes problems building Exim.
2013 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2014 header_syntax) but Exim was just ignoring anything given after a slash.
2015 In particular, this caused confusion with an attempt to use "verify =
2016 reverse_host_lookup/defer_ok". An error is now given when options are
2017 supplied for verify items that do not have them. (Maybe reverse_host_
2018 lookup should have a defer_ok option, but that's a different point.)
2020 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2021 defined by RFC 821) to 2048, because there were problems with some AUTH
2022 commands, and RFC 1869 says the size should be increased for extended
2023 SMTP commands that take arguments.
2025 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2028 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2029 "unknown" error; now it says that the functionality isn't in the binary.
2031 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2032 an address' error message when a string expansion fails (syntax or
2033 whatever). Otherwise the password may appear in the log. Following change
2034 PH/42 below, there is no longer a chance of it appearing in a bounce
2037 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2039 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2040 of its addresses was ever tried. (Bugzilla bug #2.)
2042 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2043 the result incorrectly in the debug output. (It correctly added a newline
2044 to what was transported.)
2046 TF/01 Added $received_time.
2048 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2049 commented out examples of how to interface to a virus scanner and to
2050 SpamAssassin. Also added commented examples of av_scanner and
2051 spamd_address settings.
2053 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2054 and controls are allowed in which ACLs. There were a couple of minor
2055 errors. Some of the entries in the conditions table (which is a table of
2056 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2057 negation of where the condition IS allowed.
2059 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2061 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2062 header file does not have a version number, so I've had to invent a new
2063 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2064 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2065 radiusclient), but it was contributed by a Radius user.
2067 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2068 files or directories, for OpenSSL.
2070 PH/19 When an Exim process that is running as root has to create an Exim log
2071 file, it does so in a subprocess that runs as exim:exim so as to get the
2072 ownership right at creation (otherwise, other Exim processes might see
2073 the file with the wrong ownership). There was no test for failure of this
2074 fork() call, which would lead to the process getting stuck as it waited
2075 for a non-existent subprocess. Forks do occasionally fail when resources
2076 run out. I reviewed all the other calls to fork(); they all seem to check
2079 PH/20 When checking for unexpected SMTP input at connect time (before writing
2080 the banner), Exim was not dealing correctly with a non-positive return
2081 from the read() function. If the client had disconnected by this time,
2082 the result was a log entry for a synchronization error with an empty
2083 string after "input=" when read() returned zero. If read() returned -1
2084 (an event I could not check), uninitialized data bytes were printed.
2085 There were reports of junk text (parts of files, etc) appearing after
2088 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2090 PH/22 Added support for macro redefinition, and (re)definition in between
2091 driver and ACL definitions.
2093 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2094 forgetting to use the resulting value; it was using the unexpanded value.
2096 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2097 hadn't been configured. The fix is from Juergen Kreileder, who
2098 understands it better than I do:
2100 "Here's what I see happening with three configured cyrus_sasl
2101 authenticators configured (plain, login, cram-md5):
2103 On startup auth_cyrus_sasl_init() gets called for each of these.
2104 This means three calls to sasl_listmech() without a specified mech_list.
2105 => SASL tests which mechs of all available mechs actually work
2106 => three warnings about OTP not working
2107 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2109 With the patch, sasl_listmech() also gets called three times. But now
2110 SASL's mech_list option is set to the server_mech specified in the the
2111 authenticator. Or in other words, the answer from sasl_listmech()
2112 gets limited to just the mech you're testing for (which is different
2114 => the return list contains just 'plain' or 'login', 'cram-md5' or
2115 nothing depending on the value of ob->server_mech.
2117 I've just tested the patch: Authentication still works fine,
2118 unavailable mechs specified in the exim configuration are still
2119 caught, and the auth.log warnings about OTP are gone."
2121 PH/25 When debugging is enabled, the contents of the command line are added
2122 to the debugging output, even when log_selector=+arguments is not
2125 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2126 answer is "GNU", and only if the return is "GNU/something" is the answer
2129 PH/27 $acl_verify_message is now set immediately after the failure of a
2130 verification in an ACL, and so is available in subsequent modifiers. In
2131 particular, the message can be preserved by coding like this:
2133 warn !verify = sender
2134 set acl_m0 = $acl_verify_message
2136 Previously, $acl_verify_message was set only while expanding "message"
2137 and "log_message" when a very denied access.
2139 PH/28 Modified OS/os.c-Linux with
2141 -#ifndef OS_LOAD_AVERAGE
2142 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2144 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2145 nomenclature these days.)
2147 PH/29 Installed patch from the Sieve maintainer that adds the options
2148 sieve_useraddress and sieve_subaddress to the redirect router.
2150 PH/30 In these circumstances:
2151 . Two addresses routed to the same list of hosts;
2152 . First host does not offer TLS;
2153 . First host accepts first address;
2154 . First host gives temporary error to second address;
2155 . Second host offers TLS and a TLS session is established;
2156 . Second host accepts second address.
2157 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2158 and peerdn, if requested) that were in fact used only for the second
2161 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2162 attention to any local part prefix or suffix that was matched by the
2163 router that accepted the address. It now behaves in the same way as it
2164 does for delivery: the affixes are removed from the local part unless
2165 rcpt_include_affixes is set on the transport.
2167 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2168 timeout during the DATA phase of an incoming message.
2170 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2171 applied a patch sanctioned by the Sieve maintainer.
2173 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2174 the uid or gid is negative. A case of a negative gid caused this to be
2175 noticed. The fix allows for either to be negative.
2177 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2178 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2181 PH/36 The change PH/12 above was broken. Fixed it.
2183 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2184 the grounds that routing the same address twice would always produce the
2185 same answer. This might have been true once, but it is certainly no
2186 longer true now. Routing a child address may depend on the previous
2187 routing that produced that child. Some complicated redirection strategies
2188 went wrong when messages had multiple recipients, and made Exim's
2189 behaviour dependent on the order in which the addresses were given.
2191 I have moved the duplicate checking until after the routing is complete.
2192 Exim scans the addresses that are assigned to local and remote
2193 transports, and removes any duplicates. This means that more work will be
2194 done, as duplicates will always all be routed, but duplicates are
2195 presumably rare, so I don't expect this is of any significance.
2197 For deliveries to pipes, files, and autoreplies, the duplicate checking
2198 still happens during the routing process, since they are not going to be
2201 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2202 It corrects a timeout issue with spamd. This is Ian's comment: "The
2203 background is that sometimes spamd either never reads data from a
2204 connection it has accepted, or it never writes response data. The exiscan
2205 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2206 blindly assumes that writes won't block so it may never time out."
2208 PH/39 Allow G after quota size as well as K and M.
2210 PH/40 The value set for $authenticated_id in an authenticator may not contain
2211 binary zeroes or newlines because the value is written to log lines and
2212 to spool files. There was no check on this. Now the value is run through
2213 the string_printing() function so that such characters are converted to
2214 printable escape sequences.
2216 PH/41 $message_linecount is a new variable that contains the total number of
2217 lines in the message. Compare $body_linecount, which is the count for the
2220 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2221 bounce and delay warning messages, except in certain special cases, which
2224 (a) An SMTP error message from a remote host;
2225 (b) A message specified in a :fail: redirection;
2226 (c) A message specified in a "fail" command in a system filter;
2227 (d) A message specified in a FAIL return from the queryprogram router;
2228 (e) A message specified by the cannot_route_message router option.
2230 In these cases only, Exim does include the error details in bounce and
2231 warning messages. There are also a few cases where bland messages such
2232 as "unrouteable address" or "local delivery error" are given.
2234 PH/43 $value is now also set for the "else" part of a ${run expansion.
2236 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2237 being worked on, but at least Exim now implements the latest version to
2240 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2241 process to complete was treated as a delivery failure, a timeout while
2242 writing the message to the pipe was logged, but erroneously treated as a
2243 successful delivery. Such timeouts include transport filter timeouts. For
2244 consistency with the overall process timeout, these timeouts are now
2245 treated as errors, giving rise to delivery failures by default. However,
2246 there is now a new Boolean option for the pipe transport called
2247 timeout_defer, which, if set TRUE, converts the failures into defers for
2248 both kinds of timeout. A transport filter timeout is now identified in
2251 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2252 systems where "make" and "gmake" are different, calling "gmake" at top
2253 level broke things. I've arranged for the value of $(MAKE) to be passed
2254 from the Makefile to this script so that it can call the same version of
2258 A note about Exim versions 4.44 and 4.50
2259 ----------------------------------------
2261 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2262 changes of various kinds. As a consequence, a big documentation update was
2263 needed. This delayed the release for rather longer than seemed good, especially
2264 in the light of a couple of (minor) security issues. Therefore, the changes
2265 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2266 release. So 4.44 and 4.50 are in effect two different branches that both start
2269 I have left the 4.50 change log unchanged; it contains all the changes since
2270 4.43. The change log for 4.44 is below; many of its items are identical to
2271 those for 4.50. This seems to be the most sensible way to preserve the
2272 historical information.
2278 1. Minor wording change to the doc/README.SIEVE file.
2280 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2281 computation of the current number of files was incorrect.
2283 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2284 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2287 4. Give more explanation in the error message when the command for a transport
2288 filter fails to execute.
2290 5. There are several places where Exim runs a non-Exim command in a
2291 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2292 was being done only for the command run by the queryprogram router. It is
2293 now done for all such subprocesses. The other cases are: ${run, transport
2294 filters, and the commands run by the lmtp and pipe transports.
2296 6. Added CONFIGURE_GROUP build-time option.
2298 7. Some older OS have a limit of 256 on the maximum number of file
2299 descriptors. Exim was using setrlimit() to set 1000 as a large value
2300 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2301 systems. I've change it so that if it can't get 1000, it tries for 256.
2303 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2304 was an oversight, and furthermore, ever since the addition of extra
2305 controls (e.g. 4.43/32), the checks on when to allow different forms of
2306 "control" were broken. There should now be diagnostics for all cases when a
2307 control that does not make sense is encountered.
2309 9. Added the /retain_sender option to "control=submission".
2311 10. $recipients is now available in the predata ACL (oversight).
2313 11. Tidy the search cache before the fork to do a delivery from a message
2314 received from the command line. Otherwise the child will trigger a lookup
2315 failure and thereby defer the delivery if it tries to use (for example) a
2316 cached ldap connection that the parent has called unbind on.
2318 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2319 of $address_data from the recipient verification was clobbered by the
2320 sender verification.
2322 13. The value of address_data from a sender verification is now available in
2323 $sender_address_data in subsequent conditions in the ACL statement.
2325 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2327 15. Added a new option "connect=<time>" to callout options, to set a different
2330 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2331 was its contents. (It was OK if the option was not defined at all.)
2333 17. A "Completed" log line is now written for messages that are removed from
2334 the spool by the -Mrm option.
2336 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2337 information about exactly what failed.
2339 19. Added -dd to debug only the daemon process.
2341 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2342 handles timeouts, both on the server side and network timeouts. Renamed the
2343 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2345 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2346 It is now set to "smtps".
2348 22. $host_address is now set to the target address during the checking of
2349 ignore_target_hosts.
2351 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2352 being passed; this would have caused $sender_host_name to have been used if
2353 matching the list had actually called for a host name (not very likely,
2354 since this list is usually IP addresses). A host name is now passed as
2357 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2358 code that specifies a non-restarting handler (typically sigaction() in
2359 modern systems) in an attempt to fix a rare and obscure crash bug.
2361 25. Narrowed the window for a race in the daemon that could cause it to ignore
2362 SIGCHLD signals. This is not a major problem, because they are used only to
2363 wake it up if nothing else does.
2365 26. A malformed maildirsize file could cause Exim to calculate negative values
2366 for the mailbox size or file count. Odd effects could occur as a result.
2367 The maildirsize information is now recalculated if the size or filecount
2370 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2371 support for a long time. Removed HAVE_SYS_VFS_H.
2373 28. Installed the latest version of exipick from John Jetmore.
2375 29. In an address list, if the pattern was not a regular expression, an empty
2376 subject address (from a bounce message) matched only if the pattern was an
2377 empty string. Non-empty patterns were not even tested. This was the wrong
2378 because it is perfectly reasonable to use an empty address as part of a
2379 database query. An empty address is now tested by patterns that are
2380 lookups. However, all the other forms of pattern expect the subject to
2381 contain a local part and a domain, and therefore, for them, an empty
2382 address still always fails if the pattern is not itself empty.
2384 30. Exim went into a mad DNS loop when attempting to do a callout where the
2385 host was specified on an smtp transport, and looking it up yielded more
2386 than one IP address.
2388 31. Re-factored the code for checking spool and log partition space into a
2389 function that finds that data and another that does the check. The former
2390 is then used to implement four new variables: $spool_space, $log_space,
2391 $spool_inodes, and $log_inodes.
2393 32. The RFC2047 encoding function was originally intended for short strings
2394 such as real names; it was not keeping to the 75-character limit for
2395 encoded words that the RFC imposes. It now respects the limit, and
2396 generates multiple encoded words if necessary. To be on the safe side, I
2397 have increased the buffer size for the ${rfc2047: expansion operator from
2400 33. It is now permitted to omit both strings after an "if" condition; if the
2401 condition is true, the result is "true". As before, when the second string
2402 is omitted, a false condition yields an empty string. This makes it less
2403 cumbersome to write custom ACL and router conditions.
2405 34. Failure to deliver a bounce message always caused it to be frozen, even if
2406 there was an errors_to setting on the router. The errors_to setting is now
2409 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2410 canonical form (fully expanded) before being placed in
2411 $sender_host_address.
2413 36. The table in the code that translates DNS record types into text (T_A to
2414 "A" for instance) was missing entries for NS and CNAME. It is just possible
2415 that this could have caused confusion if both these types were looked up
2416 for the same domain, because the text type is used as part of Exim's
2417 per-process caching. But the chance of anyone hitting this buglet seems
2420 37. The dnsdb lookup has been extended in a number of ways.
2422 (1) There is a new type, "zns", which walks up the domain tree until it
2423 finds some nameserver records. It should be used with care.
2425 (2) There is a new type, "mxh", which is like "mx" except that it returns
2426 just the host names, not the priorities.
2428 (3) It is now possible to give a list of domains (or IP addresses) to be
2429 looked up. The behaviour when one of the lookups defers can be
2430 controlled by a keyword.
2432 (4) It is now possible to specify the separator character for use when
2433 multiple records are returned.
2435 38. The dnslists ACL condition has been extended: it is now possible to supply
2436 a list of IP addresses and/or domains to be looked up in a particular DNS
2439 39. Added log_selector=+queue_time_overall.
2441 40. When running the queue in the test harness, wait just a tad after forking a
2442 delivery process, to get repeatability of debugging output.
2444 41. Include certificate and key file names in error message when GnuTLS fails
2445 to set them up, because the GnuTLS error message doesn't include the name
2446 of the failing file when there is a problem reading it.
2448 42. Allow both -bf and -bF in the same test run.
2450 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2452 44. The "Exiscan patch" is now merged into the mainline Exim source.
2454 45. Sometimes the final signoff response after QUIT could fail to get
2455 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2456 < 0 before doing a fflush(). This bug looks as though it goes back to the
2457 introduction of TLS in release 3.20, but "sometimes" must have been rare
2458 because the tests only now provoked it.
2460 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2461 (this can affect the format of dates).
2463 47. exim_tidydb, when checking for the continued existence of a message for
2464 which it has found a message-specific retry record, was not finding
2465 messages that were in split spool directories. Consequently, it was
2466 deleting retry records that should have stayed in existence.
2468 48. Steve fixed some bugs in eximstats.
2470 49. The SPA authentication driver was not abandoning authentication and moving
2471 on to the next authenticator when an expansion was forced to fail,
2472 contradicting the general specification for all authenticators. Instead it
2473 was generating a temporary error. It now behaves as specified.
2475 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2476 (the order specifies the preference for clients). The order is now AES256,
2477 AES128, 3DES, ARCFOUR128.
2479 51. Small patch to Sieve code - explicitly set From: when generating an
2482 52. Exim crashed if a remote delivery caused a very long error message to be
2483 recorded - for instance if somebody sent an entire SpamAssassin report back
2484 as a large number of 550 error lines. This bug was coincidentally fixed by
2485 increasing the size of one of Exim's internal buffers (big_buffer) that
2486 happened as part of the Exiscan merge. However, to be on the safe side, I
2487 have made the code more robust (and fixed the comments that describe what
2490 53. Now that there can be additional text after "Completed" in log lines (if
2491 the queue_time_overall log selector is set), a one-byte patch to exigrep
2492 was needed to allow it to recognize "Completed" as not the last thing in
2495 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2496 patch that reportedly fixes this has been added. I am not expert enough to
2497 create a test for it. This is what the patch creator wrote:
2499 "I found a little strange behaviour of ldap code when working with
2500 Windows 2003 AD Domain, where users was placed in more than one
2501 Organization Units. When I tried to give exim partial DN, the exit code
2502 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2503 But simultaneously result of request was absolutely normal ldap result,
2504 so I produce this patch..."
2506 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2507 REFERENCE, so I have modified the code to exclude the patch when that macro
2510 55. Some experimental protocols are using DNS PTR records for new purposes. The
2511 keys for these records are domain names, not reversed IP addresses. The
2512 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2513 leaves it alone. Component reversal etc. now happens only for IP addresses.
2516 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2518 57. Double the size of the debug message buffer (to 2048) so that more of very
2519 long debug lines gets shown.
2521 58. The exicyclog utility now does better if the number of log files to keep
2522 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2524 59. Two changes related to the smtp_active_hostname option:
2526 (1) $smtp_active_hostname is now available as a variable.
2527 (2) The default for smtp_banner uses $smtp_active_hostname instead
2528 of $primary_hostname.
2530 60. The host_aton() function is supposed to be passed a string that is known
2531 to be a valid IP address. However, in the case of IPv6 addresses, it was
2532 not checking this. This is a hostage to fortune. Exim now panics and dies
2533 if the condition is not met. A case was found where this could be provoked
2534 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2535 components; fortuitously, this particular loophole had already been fixed
2536 by change 4.50/55 above.
2538 If there are any other similar loopholes, the new check in host_aton()
2539 itself should stop them being exploited. The report I received stated that
2540 data on the command line could provoke the exploit when Exim was running as
2541 exim, but did not say which command line option was involved. All I could
2542 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2543 running as the user.
2546 61. There was a buffer overflow vulnerability in the SPA authentication code
2547 (which came originally from the Samba project). I have added a test to the
2548 spa_base64_to_bits() function which I hope fixes it.
2551 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2552 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2554 63. The daemon start-up calls getloadavg() while still root for those OS that
2555 need the first call to be done as root, but it missed one case: when
2556 deliver_queue_load_max is set with deliver_drop_privilege. This is
2557 necessary for the benefit of the queue runner, because there is no re-exec
2558 when deliver_drop_privilege is set.
2560 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2561 This has been fixed.
2563 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2564 was in use, was not putting the data itself into the right store pool;
2565 consequently, it could be overwritten for a subsequent message in the same
2566 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2569 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2571 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2572 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2573 the function were treating the return as a boolean value, which happened to
2574 work because 0=false and not-0=true, but is not correct code.
2576 68. The host_aton() function was not handling scoped IPv6 addresses (those
2577 with, for example, "%eth0" on the end) correctly.
2579 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2580 negated items (that is, ~something) in unsigned ints. Some compilers
2581 apparently mutter when there is no cast.
2583 70. If an address verification called from an ACL failed, and did not produce a
2584 user-specific message (i.e. there was only a "system" message), nothing was
2585 put in $acl_verify_message. In this situation, it now puts the system
2588 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2589 session; change 4.31/43 added the unwanted input to the log line - except
2590 that it did not do this in the start of session case. It now does.
2592 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2593 This is wrong and can cause the other end to generate a synchronization
2594 error if it is another Exim or anything else that does the synchronization
2595 check. A QUIT command is no longer sent after a timeout.
2597 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2598 during host lookups.
2600 74. The defer_ok option of callout verification was not working if it was used
2601 when verifying addresses in header lines, that is, for this case:
2603 verify = header_sender/callout=defer_ok
2605 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2606 those file descriptors could be used for SMTP connections. If anything
2607 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2608 could be sent to the SMTP client, causing chaos. The daemon now opens
2609 stdin, stdout, and stderr to /dev/null when it puts itself into the
2612 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2613 log by default. The user can override this with suitable Perl magic.
2615 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2616 the log message when discard triggers, was not working for the DATA ACL or
2617 for the non-SMTP ACL.
2619 78. Error message wording change in sieve.c.
2621 79. If smtp_accept_max_per_host was set, the number of connections could be
2622 restricted to fewer than expected, because the daemon was trying to set up
2623 a new connection before checking whether the processes handling previous
2624 connections had finished. The check for completed processes is now done
2625 earlier. On busy systems, this bug wouldn't be noticed because something
2626 else would have woken the daemon, and it would have reaped the completed
2629 80. If a message was submitted locally by a user whose login name contained one
2630 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2631 It caused a spool format error. I have fixed the spool reading code. A
2632 related problem was that the "from" clause in the Received: line became
2633 illegal because of the space(s). It is now covered by ${quote_local_part.
2635 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2638 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2640 83. Patch from Sieve maintainer to fix unterminated string problem in
2641 "vacation" handling.
2643 84. Some minor changes to the Linux configuration files to help with other
2644 OS variants using glibc.
2646 85. One more patch for Sieve to update vacation handling to latest spec.
2649 ----------------------------------------------------
2650 See the note above about the 4.44 and 4.50 releases.
2651 ----------------------------------------------------
2657 1. Change 4.43/35 introduced a bug that caused file counts to be
2658 incorrectly computed when quota_filecount was set in an appendfile
2661 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2662 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2665 3. Give more explanation in the error message when the command for a transport
2666 filter fails to execute.
2668 4. There are several places where Exim runs a non-Exim command in a
2669 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2670 was being done only for the command run by the queryprogram router. It is
2671 now done for all such subprocesses. The other cases are: ${run, transport
2672 filters, and the commands run by the lmtp and pipe transports.
2674 5. Some older OS have a limit of 256 on the maximum number of file
2675 descriptors. Exim was using setrlimit() to set 1000 as a large value
2676 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2677 systems. I've change it so that if it can't get 1000, it tries for 256.
2679 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2680 was an oversight, and furthermore, ever since the addition of extra
2681 controls (e.g. 4.43/32), the checks on when to allow different forms of
2682 "control" were broken. There should now be diagnostics for all cases when a
2683 control that does not make sense is encountered.
2685 7. $recipients is now available in the predata ACL (oversight).
2687 8. Tidy the search cache before the fork to do a delivery from a message
2688 received from the command line. Otherwise the child will trigger a lookup
2689 failure and thereby defer the delivery if it tries to use (for example) a
2690 cached ldap connection that the parent has called unbind on.
2692 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2693 of $address_data from the recipient verification was clobbered by the
2694 sender verification.
2696 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2697 was its contents. (It was OK if the option was not defined at all.)
2699 11. A "Completed" log line is now written for messages that are removed from
2700 the spool by the -Mrm option.
2702 12. $host_address is now set to the target address during the checking of
2703 ignore_target_hosts.
2705 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2706 being passed; this would have caused $sender_host_name to have been used if
2707 matching the list had actually called for a host name (not very likely,
2708 since this list is usually IP addresses). A host name is now passed as
2711 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2712 code that specifies a non-restarting handler (typically sigaction() in
2713 modern systems) in an attempt to fix a rare and obscure crash bug.
2715 15. Narrowed the window for a race in the daemon that could cause it to ignore
2716 SIGCHLD signals. This is not a major problem, because they are used only to
2717 wake it up if nothing else does.
2719 16. A malformed maildirsize file could cause Exim to calculate negative values
2720 for the mailbox size or file count. Odd effects could occur as a result.
2721 The maildirsize information is now recalculated if the size or filecount
2724 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2725 support for a long time. Removed HAVE_SYS_VFS_H.
2727 18. Updated exipick to current release from John Jetmore.
2729 19. Allow an empty sender to be matched against a lookup in an address list.
2730 Previously the only cases considered were a regular expression, or an
2733 20. Exim went into a mad DNS lookup loop when doing a callout where the
2734 host was specified on the transport, if the DNS lookup yielded more than
2737 21. The RFC2047 encoding function was originally intended for short strings
2738 such as real names; it was not keeping to the 75-character limit for
2739 encoded words that the RFC imposes. It now respects the limit, and
2740 generates multiple encoded words if necessary. To be on the safe side, I
2741 have increased the buffer size for the ${rfc2047: expansion operator from
2744 22. Failure to deliver a bounce message always caused it to be frozen, even if
2745 there was an errors_to setting on the router. The errors_to setting is now
2748 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2749 canonical form (fully expanded) before being placed in
2750 $sender_host_address.
2752 24. Updated eximstats to version 1.33
2754 25. Include certificate and key file names in error message when GnuTLS fails
2755 to set them up, because the GnuTLS error message doesn't include the name
2756 of the failing file when there is a problem reading it.
2758 26. Expand error message when OpenSSL has problems setting up cert/key files.
2761 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2762 (this can affect the format of dates).
2764 28. exim_tidydb, when checking for the continued existence of a message for
2765 which it has found a message-specific retry record, was not finding
2766 messages that were in split spool directories. Consequently, it was
2767 deleting retry records that should have stayed in existence.
2769 29. eximstats updated to version 1.35
2770 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2771 1.35 - bugfix such that pie charts by volume are generated correctly
2773 30. The SPA authentication driver was not abandoning authentication and moving
2774 on to the next authenticator when an expansion was forced to fail,
2775 contradicting the general specification for all authenticators. Instead it
2776 was generating a temporary error. It now behaves as specified.
2778 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2779 (the order specifies the preference for clients). The order is now AES256,
2780 AES128, 3DES, ARCFOUR128.
2782 31. Small patch to Sieve code - explicitly set From: when generating an
2785 32. Exim crashed if a remote delivery caused a very long error message to be
2786 recorded - for instance if somebody sent an entire SpamAssassin report back
2787 as a large number of 550 error lines. This bug was coincidentally fixed by
2788 increasing the size of one of Exim's internal buffers (big_buffer) that
2789 happened as part of the Exiscan merge. However, to be on the safe side, I
2790 have made the code more robust (and fixed the comments that describe what
2793 33. Some experimental protocols are using DNS PTR records for new purposes. The
2794 keys for these records are domain names, not reversed IP addresses. The
2795 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2796 leaves it alone. Component reversal etc. now happens only for IP addresses.
2799 34. The host_aton() function is supposed to be passed a string that is known
2800 to be a valid IP address. However, in the case of IPv6 addresses, it was
2801 not checking this. This is a hostage to fortune. Exim now panics and dies
2802 if the condition is not met. A case was found where this could be provoked
2803 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2804 components; fortuitously, this particular loophole had already been fixed
2805 by change 4.50/55 or 4.44/33 above.
2807 If there are any other similar loopholes, the new check in host_aton()
2808 itself should stop them being exploited. The report I received stated that
2809 data on the command line could provoke the exploit when Exim was running as
2810 exim, but did not say which command line option was involved. All I could
2811 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2812 running as the user.
2815 35. There was a buffer overflow vulnerability in the SPA authentication code
2816 (which came originally from the Samba project). I have added a test to the
2817 spa_base64_to_bits() function which I hope fixes it.
2820 36. The daemon start-up calls getloadavg() while still root for those OS that
2821 need the first call to be done as root, but it missed one case: when
2822 deliver_queue_load_max is set with deliver_drop_privilege. This is
2823 necessary for the benefit of the queue runner, because there is no re-exec
2824 when deliver_drop_privilege is set.
2826 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2827 was in use, was not putting the data itself into the right store pool;
2828 consequently, it could be overwritten for a subsequent message in the same
2829 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2832 38. Sometimes the final signoff response after QUIT could fail to get
2833 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2834 < 0 before doing a fflush(). This bug looks as though it goes back to the
2835 introduction of TLS in release 3.20, but "sometimes" must have been rare
2836 because the tests only now provoked it.
2842 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2843 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2844 Now it returns an integer. A number of calls were still expecting a T/F
2845 return. Fortuitously, in all cases, the tests worked in OK situations,
2846 which is the norm. However, things would have gone wrong on any write
2847 failures on the smtp file descriptor. This function is used when sending
2848 messages over SMTP and also when doing verify callouts.
2850 2. When Exim is called to do synchronous delivery of a locally submitted
2851 message (the -odf or -odi options), it no longer closes stderr before doing
2854 3. Implemented the mua_wrapper option.
2856 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2858 5. Implemented the functions header_remove(), header_testname(),
2859 header_add_at_position(), and receive_remove_recipient(), and exported them
2862 6. If an ACL "warn" statement specified the addition of headers, Exim already
2863 inserted X-ACL-Warn: at the start if there was no header name. However, it
2864 was not making this test for the second and subsequent header lines if
2865 there were newlines in the string. This meant that an invalid header could
2866 be inserted if Exim was badly configured.
2868 7. Allow an ACL "warn" statement to add header lines at the start or after all
2869 the Received: headers, as well as at the end.
2871 8. Added the rcpt_4xx retry error code.
2873 9. Added postmaster_mailfrom=xxx to callout verification option.
2875 10. Added mailfrom=xxxx to the callout verification option, for verify=
2878 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2879 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2881 12. Inserted some casts to stop certain compilers warning when using pointer
2882 differences as field lengths or precisions in printf-type calls (mostly
2883 affecting debugging statements).
2885 13. Added optional readline() support for -be (dynamically loaded).
2887 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2888 same clock tick as a message's arrival, so that its received time was the
2889 same as the "first fail" time on the retry record, and that message
2890 remained on the queue past the ultimate address timeout, every queue runner
2891 would try a delivery (because it was past the ultimate address timeout) but
2892 after another failure, the ultimate address timeout, which should have then
2893 bounced the address, did not kick in. This was a "< instead of <=" error;
2894 in most cases the first failure would have been in the next clock tick
2895 after the received time, and all would be well.
2897 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2898 being recognized when the domain list was tested by the match_domain
2899 condition in an expansion string.
2901 16. Added the ${str2b64: operator.
2903 17. Exim was always calling setrlimit() to set a large limit for the number of
2904 processes, without checking whether the existing limit was already
2905 adequate. (It did check for the limit on file descriptors.) Furthermore,
2906 errors from getrlimit() and setrlimit() were being ignored. Now they are
2907 logged to the main and panic logs, but Exim does carry on, to try to do its
2908 job under whatever limits there are.
2910 18. Imported PCRE 5.0.
2912 19. Trivial typo in log message " temporarily refused connection" (the leading
2915 20. If the log selector return_path_on_delivery was set and an address was
2916 redirected to /dev/null, the delivery process crashed because it assumed
2917 that a return path would always be set for a "successful" delivery. In this
2918 case, the whole delivery is bypassed as an optimization, and therefore no
2921 21. Internal re-arrangement: the function for sending a challenge and reading
2922 a response while authentication was assuming a zero-terminated challenge
2923 string. It's now changed to take a pointer and a length, to allow for
2924 binary data in such strings.
2926 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2928 23. Exim was not respecting finduser_retries when seeking the login of the
2929 uid under which it was called; it was always trying 10 times. (The default
2930 setting of finduser_retries is zero.) Also, it was sleeping after the final
2931 failure, which is pointless.
2933 24. Implemented tls_on_connect_ports.
2935 25. Implemented acl_smtp_predata.
2937 26. If the domain in control=submission is set empty, Exim assumes that the
2938 authenticated id is a complete email address when it generates From: or
2939 Sender: header lines.
2941 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2942 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2943 chown and chgrp in /bin and hostname in /usr/bin.
2945 28. Exim was keeping the "process log" file open after each use, just as it
2946 does for the main log. This opens the possibility of it remaining open for
2947 long periods when the USR1 signal hits a daemon. Occasional processlog
2948 errors were reported, that could have been caused by this. Anyway, it seems
2949 much more sensible not to leave this file open at all, so that is what now
2952 29. The long-running daemon process does not normally write to the log once it
2953 has entered its main loop, and it closes the log before doing so. This is
2954 so that log files can straightforwardly be renamed and moved. However,
2955 there are a couple of unusual error situations where the daemon does write
2956 log entries, and I had neglected to close the log afterwards.
2958 30. The text of an SMTP error response that was received during a remote
2959 delivery was being truncated at 512 bytes. This is too short for some of
2960 the long messages that one sometimes sees. I've increased the limit to
2963 31. It is now possible to make retry rules that apply only when a message has a
2964 specific sender, in particular, an empty sender.
2966 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2967 it possible to be selective about when SMTP synchronization is enforced.
2969 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2971 32. Implemented hosts_connection_nolog.
2973 33. Added an ACL for QUIT.
2975 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2978 35. Added mailbox_size and mailbox_filecount to appendfile.
2980 36. Added control = no_multiline_responses to ACLs.
2982 37. There was a bug in the logic of the code that waits for the clock to tick
2983 in the case where the clock went backwards by a substantial amount such
2984 that the microsecond fraction of "now" was more than the microsecond
2985 fraction of "then" (but the whole seconds number was less).
2987 38. Added support for the libradius Radius client library this is found on
2988 FreeBSD (previously only the radiusclient library was supported).
2994 1. When certain lookups returned multiple values in the form name=value, the
2995 quoting of the values was not always being done properly. Specifically:
2996 (a) If the value started with a double quote, but contained no whitespace,
2998 (b) If the value contained whitespace other than a space character (i.e.
2999 tabs or newlines or carriage returns) it was not quoted.
3000 This fix has been applied to the mysql and pgsql lookups by writing a
3001 separate quoting function and calling it from the lookup code. The fix
3002 should probably also be applied to nisplus, ibase and oracle lookups, but
3003 since I cannot test any of those, I have not disturbed their existing code.
3005 2. A hit in the callout cache for a specific address caused a log line with no
3006 reason for rejecting RCPT. Now it says "Previous (cached) callout
3007 verification failure".
3009 3. There was an off-by-one bug in the queryprogram router. An over-long
3010 return line was truncated at 256 instead of 255 characters, thereby
3011 overflowing its buffer with the terminating zero. As well as fixing this, I
3012 have increased the buffer size to 1024 (and made a note to document this).
3014 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3015 when Exim is waiting for an SMTP response from a remote server, Exim
3016 restarts its select() call on the socket, thereby resetting its timeout.
3017 This is not a problem when such interrupts are rare. Somebody set up a cron
3018 job to run exiwhat every 2 minutes, which is less than the normal select()
3019 timeout (5 or 10 minutes). This meant that the select() timeout never
3020 kicked in because it was always reset. I have fixed this by comparing the
3021 time when an interrupt arrives with the time at the start of the first call
3022 to select(). If more time than the timeout has elapsed, the interrupt is
3023 treated as a timeout.
3025 5. Some internal re-factoring in preparation for the addition of Sieve
3026 extensions (by MH). In particular, the "personal" test is moved to a
3027 separate function, and given an option for scanning Cc: and Bcc: (which is
3028 not set for Exim filters).
3030 6. When Exim created an email address using the login of the caller as the
3031 local part (e.g. when creating a From: or Sender: header line), it was not
3032 quoting the local part when it contained special characters such as @.
3034 7. Installed new OpenBSD configuration files.
3036 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3037 try to make them clearer.
3039 9. Callout options, other than the timeout value, were being ignored when
3040 verifying sender addresses in header lines. For example, when using
3042 verify = header_sender/callout=no_cache
3044 the cache was (incorrectly) being used.
3046 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3047 only the Cygwin environment.
3049 11. When return_path_on_delivery was set as a log selector, if different remote
3050 addresses in the same message used different return paths and parallel
3051 remote delivery occurred, the wrong values would sometimes be logged.
3052 (Whenever a remote delivery process finished, the return path value from
3053 the most recently started remote delivery process was logged.)
3055 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3056 lines when AUTH and/or TLS are in use. This is the "received protocol"
3057 field. Exim used to use "asmtp" for authenticated SMTP, without any
3058 indication (in the protocol name) for TLS use. Now it follows the RFC and
3059 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3060 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3061 names appear in log lines as well as in Received: header lines.
3063 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3064 extensions, and comparison tests, and to fix some bugs.
3066 14. Changes to the "personal" filter test:
3068 (1) The test was buggy in that it was just doing the equivalent of
3069 "contains" tests on header lines. For example, if a user's address was
3070 anne@some.where, the "personal" test would incorrectly be true for
3072 To: susanne@some.where
3074 This test is now done by extracting each address from the header in turn,
3075 and checking the entire address. Other tests that are part of "personal"
3076 are now done using regular expressions (for example, to check local parts
3077 of addresses in From: header lines).
3079 (2) The list of non-personal local parts in From: addresses has been
3080 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3081 taken from the Sieve specification recommendations.
3083 (3) If the message contains any header line starting with "List-" it is
3084 treated as non-personal.
3086 (4) The test for "circular" in the Subject: header line has been removed
3087 because it now seems ill-conceived.
3089 15. Minor typos in src/EDITME comments corrected.
3091 16. Installed latest exipick from John Jetmore.
3093 17. If headers_add on a router specified a text string that was too long for
3094 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3095 of string_sprintf() is now avoided.
3097 18. $message_body_size was not set (it was always zero) when running the DATA
3098 ACL and the local_scan() function.
3100 19. For the "mail" command in an Exim filter, no default was being set for
3101 the once_repeat time, causing a random time value to be used if "once" was
3102 specified. (If the value happened to be <= 0, no repeat happened.) The
3103 default is now 0s, meaning "never repeat". The "vacation" command was OK
3104 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3105 (I found it when inspecting the code).
3107 20. There is now an overall timeout for performing a callout verification. It
3108 defaults to 4 times the callout timeout, which applies to individual SMTP
3109 commands during the callout. The overall timeout applies when there is more
3110 than one host that can be tried. The timeout is checked before trying the
3111 next host. This prevents very long delays if there are a large number of
3112 hosts and all are timing out (e.g. when the network connections are timing
3113 out). The value of the overall timeout can be changed by specifying an
3114 additional sub-option for "callout", called "maxwait". For example:
3116 verify = sender/callout=5s,maxwait=20s
3118 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3119 to the end before writing, but this should make it even safer).
3121 22. Exim was forgetting that it had advertised PIPELINING for the second and
3122 subsequent messages on an SMTP connection. It was also not resetting its
3123 memory on STARTTLS and an internal HELO.
3125 23. When Exim logs an SMTP synchronization error within a session, it now
3126 records whether PIPELINING has been advertised or not.
3128 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3129 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3130 rather than long int.
3132 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3134 26. Added the never_mail option to autoreply.
3140 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3141 crash if the getsockname() call failed; this can happen if a connection is
3142 closed very soon after it is established. The problem was simply in the
3143 order in which certain operations were done, causing Exim to try to write
3144 to the SMTP stream before it had set up the file descriptor. The bug has
3145 been fixed by making things happen in the correct order.
3151 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3152 before the connection was closed, thus losing the rejection response.
3154 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3155 some early Solaris releases, but causes trouble in current releases where
3156 socklen_t is defined.
3158 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3161 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3164 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3165 that releasing the top of it at the end releases what was used for sub-
3166 expansions (unless the block got too big). However, discard this block if
3167 the first thing is a variable or header, so that we can use its block when
3168 it is dynamic (useful for very large $message_headers, for example).
3170 6. Lookups now cache *every* query, not just the most recent. A new, separate
3171 store pool is used for this. It can be recovered when all lookup caches are
3172 flushed. Lookups now release memory at the end of their result strings.
3173 This has involved some general refactoring of the lookup sources.
3175 7. Some code has been added to the store_xxx() functions to reduce the amount
3176 of flapping under certain conditions.
3178 8. log_incoming_interface used to affect only the <= reception log lines. Now
3179 it causes the local interface and port to be added to several more SMTP log
3180 lines, for example "SMTP connection from", and rejection lines.
3182 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3184 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3186 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3187 could be overwritten at the end of the current message (or the start of a
3188 new message if it was set in a HELO ACL). The value is now preserved for
3189 the duration of the SMTP connection.
3191 12. If a transport had a headers_rewrite setting, and a matching header line
3192 contained an unqualified address, that address was qualified, even if it
3193 did not match any rewriting rules. The underlying bug was that the values
3194 of the flags that permit the existence of unqualified sender and recipient
3195 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3196 non-local messages, and by -bnq for local messages) were not being
3197 preserved with the message after it was received.
3199 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3200 "next input=" as part of the text comprising the host identity instead of
3201 the correct text. The code was using the same buffer for two different
3202 strings. However, depending on which order the printing function evaluated
3203 its arguments, the bug did not always show up. Under Linux, for example, my
3204 test suite worked just fine.
3206 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3207 doesn't work with some older versions of Perl. It has been changed to "my",
3208 which in any case is probably the better facility to use.
3210 15. A really picky compiler found some instances of statements for creating
3211 error messages that either had too many or two few arguments for the format
3214 16. The size of the buffer for calls to the DNS resolver has been increased
3215 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3216 for addresses that have a lot of PTR records. This alleviates a problem; it
3217 does not fully solve it.
3219 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3220 buffer now truncates the list and logs the incident, which is the same
3221 action as happens when Exim is looking up a host name and its aliases.
3222 Previously in this situation something unpredictable would happen;
3223 sometimes it was "internal error: store_reset failed".
3225 18. If a server dropped the connection unexpectedly when an Exim client was
3226 using GnuTLS and trying to read a response, the client delivery process
3227 crashed while trying to generate an error log message.
3229 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3230 string, for example:
3232 warn message = H1: something\nH2: something
3234 the text was added as a single header line from Exim's point of view
3235 though it ended up OK in the delivered message. However, searching for the
3236 second and subsequent header lines using $h_h2: did not work. This has been
3237 fixed. Similarly, if a system filter added multiple headers in this way,
3238 the routers could not see them.
3240 20. Expanded the error message when iplsearch is called with an invalid key to
3241 suggest using net-iplsearch in a host list.
3243 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3244 ACLs are no longer actually imposed (and a message to that effect is
3247 22. If a "gecos" field in a passwd entry contained escaped characters, in
3248 particular, if it contained a \" sequence, Exim got it wrong when building
3249 a From: or a Sender: header from that name. A second bug also caused
3250 incorrect handling when an unquoted " was present following a character
3251 that needed quoting.
3253 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3254 was not being matched caselessly.
3256 24. Arranged for all hyphens in the exim.8 source to be escaped with
3259 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3260 properly. Recipient callout cache records were still being keyed to include
3261 the sender, even when use_sender was set false. This led to far more
3262 callouts that were necessary. The sender is no longer included in the key
3263 when use_sender is false.
3265 26. Added "control = submission" modifier to ACLs.
3267 27. Added the ${base62d: operator to decode base 62 numbers.
3269 28. dnsdb lookups can now access SRV records.
3271 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3272 the configuration file.
3274 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3275 (-v) mode. This makes the output for a verbose queue run more intelligible.
3277 31. Added a use_postmaster feature to recipient callouts.
3279 32. Added the $body_zerocount variable, containing the number of binary zero
3280 bytes in the message body.
3282 33. The time of last modification of the "new" subdirectory is now used as the
3283 "mailbox time last read" when there is a quota error for a maildir
3286 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3288 35. Added +ignore_unknown as a special item in host lists.
3290 36. Code for decoding IPv6 addresses in host lists is now included, even if
3291 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3292 address was recognized as an IP address, but was then not correctly decoded
3293 into binary, causing unexpected and incorrect effects when compared with
3300 1. Very minor rewording of debugging text in manualroute to say "list of
3301 hosts" instead of "hostlist".
3303 2. If verify=header_syntax was set, and a header line with an unqualified
3304 address (no domain) and a large number of spaces between the end of the
3305 name and the colon was received, the reception process suffered a buffer
3306 overflow, and (when I tested it) crashed. This was caused by some obsolete
3307 code that should have been removed. The fix is to remove it!
3309 3. When running in the test harness, delay a bit after writing a bounce
3310 message to get a bit more predictability in the log output.
3312 4. Added a call to search_tidyup() just before forking a reception process. In
3313 theory, someone could use a lookup in the expansion of smtp_accept_max_
3314 per_host which, without the tidyup, could leave open a database connection.
3316 5. Added the variables $recipient_data and $sender_data which get set from a
3317 lookup success in an ACL "recipients" or "senders" condition, or a router
3318 "senders" option, similar to $domain_data and $local_part_data.
3320 6. Moved the writing of debug_print from before to after the "senders" test
3323 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3324 problems for message scanning, either using a data ACL, or using
3325 local_scan() because the Received: header was not generated till after they
3326 were called (in order to set the time as the time of reception completion).
3327 I have revised the way this works. The header is now generated after the
3328 body is received, but before the ACL or local_scan() are called. After they
3329 are run, the timestamp in the header is updated.
3335 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3336 before starting a queue runner without re-exec. This happened only when
3337 deliver_drop_privilege was set or when the Exim user was set to root. The
3338 effect of the bug was that timeouts during subsequent deliveries caused
3339 crashes instead of being properly handled. The handler is now left at its
3340 default (and expected) setting.
3342 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3343 message, again when deliver_drop_privilege is set or Exim is run as root.
3344 The bug described in (1) was not present in this case, but the tidying up
3345 of the other signals was missing. I have made the two cases consistent.
3347 3. The ignore_target_hosts setting on a manualroute router was being ignored
3348 for hosts that were looked up using the /MX notation.
3350 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3353 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3354 operated on the sender address. After changing the $sender_address to <>
3355 for the sender address verify, Exim was re-instated it as the original
3356 (before rewriting) address, but remembering that it had rewritten it, so it
3357 wasn't rewriting it again. This bug also had the effect of breaking the
3358 sender address verification caching when the sender address was rewritten.
3360 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3361 This has been changed so that if the ip literal address matches
3362 ignore_target_hosts, the router declines.
3364 7. Added expansion conditions match_domain, match_address, and match_local_
3365 part (NOT match_host).
3367 8. The placeholder for the Received: header didn't have a length field set.
3369 9. Added code to Exim itself and to exim_lock to test for a specific race
3370 condition that could lead to file corruption when using MBX delivery. The
3371 issue is with the lockfile that is created in /tmp. If this file is removed
3372 after a process has opened it but before that process has acquired a lock,
3373 there is the potential for a second process to recreate the file and also
3374 acquire a lock. This could lead to two Exim processes writing to the file
3375 at the same time. The added code performs the same test as UW imapd; it
3376 checks after acquiring the lock that its file descriptor still refers to
3377 the same named file.
3379 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3380 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3383 11. Added the smtp_active_hostname option. If used, this will typically be made
3384 to depend on the incoming interface address. Because $interface_address is
3385 not set up until the daemon has forked a reception process, error responses
3386 that can happen earlier (such as "too many connections") no longer contain
3389 12. If an expansion in a condition on a "warn" statement fails because a lookup
3390 defers, the "warn" statement is abandoned, and the next ACL statement is
3391 processed. Previously this caused the whole ACL to be aborted.
3393 13. Added the iplsearch lookup type.
3395 14. Added ident_timeout as a log selector.
3397 15. Added tls_certificate_verified as a log selector.
3399 16. Added a global option tls_require_ciphers (compare the smtp transport
3400 option of the same name). This controls incoming TLS connections.
3402 17. I finally figured out how to make tls_require_ciphers do a similar thing
3403 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3404 before starting the TLS session.
3406 18. Tabs are now shown as \t in -bP output.
3408 19. If the log selector return_path_on_delivery was set, Exim crashed when
3409 bouncing a message because it had too many Received: header lines.
3411 20. If two routers both had headers_remove settings, and the first one included
3412 a superfluous trailing colon, the final name in the first list and the
3413 first name in the second list were incorrectly joined into one item (with a
3414 colon in the middle).
3420 1. Added -C and -D options to the exinext utility, mainly to make it easier
3421 to include in the automated testing, but these could be helpful when
3422 multiple configurations are in use.
3424 2. The exinext utility was not formatting the output nicely when there was
3425 an alternate port involved in the retry record key, nor when there was a
3426 message id as well (for retries that were specific to a specific message
3427 and a specific host). It was also confused by IPv6 addresses, because of
3428 the additional colons they contain. I have fixed the IPv4 problem, and
3429 patched it up to do a reasonable job for IPv6.
3431 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3432 delivery, the log line now contains "pipelined" if PIPELINING was used.
3434 4. An SMTP transport process used to panic and die if the bind() call to set
3435 an explicit outgoing interface failed. This has been changed; it is now
3436 treated in the same way as a connect() failure.
3438 5. A reference to $sender_host_name in the part of a conditional expansion
3439 that was being skipped was still causing a DNS lookup. This no longer
3442 6. The def: expansion condition was not recognizing references to header lines
3443 that used bh_ and bheader_.
3445 7. Added the _cache feature to named lists.
3447 8. The code for checking quota_filecount in the appendfile transport was
3448 allowing one more file than it should have been.
3450 9. For compatibility with Sendmail, the command line option
3458 and sets the incoming protocol and host name (for trusted callers). The
3459 host name and its colon can be omitted when only the protocol is to be set.
3460 Note the Exim already has two private options, -pd and -ps, that refer to
3461 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3462 "s", but I don't think that's a major issue.
3464 10. A number of refactoring changes to the code, none of which should affect
3467 (a) The number of logging options was getting close to filling up the
3468 32-bit word that was used as a bit map. I have split them into two classes:
3469 those that are passed in the argument to log_write(), and those that are
3470 only ever tested independently outside of that function. These are now in
3471 separate 32-bit words, so there is plenty of room for expansion again.
3472 There is no change in the user interface or the logging behaviour.
3474 (b) When building, for example, log lines, the code previously used a
3475 macro that called string_cat() twice, in order to add two strings. This is
3476 not really sufficiently general. Furthermore, there was one instance where
3477 it was actually wrong because one of the argument was used twice, and in
3478 one call a function was used. (As it happened, calling the function twice
3479 did not affect the overall behaviour.) The macro has been replaced by a
3480 function that can join an arbitrary number of extra strings onto a growing
3483 (c) The code for expansion conditions now uses a table and a binary chop
3484 instead of a serial search (which was left over from when there were very
3485 few conditions). Also, it now recognizes conditions like "pam" even when
3486 the relevant support is not compiled in: a suitably worded error message is
3487 given if an attempt is made to use such a condition.
3489 11. Added ${time_interval:xxxxx}.
3491 12. A bug was causing one of the ddress fields not to be passed back correctly
3492 from remote delivery subprocesses. The field in question was not being
3493 subsequently used, so this caused to problems in practice.
3495 13. Added new log selectors queue_time and deliver_time.
3497 14. Might have fixed a bug in maildirsizefile handling that threw up
3498 "unexpected character" debug warnings, and recalculated the data
3499 unnecessarily. In any case, I expanded the warning message to give more
3502 15. Added the message "Restricted characters in address" to the statements in
3503 the default ACL that block characters like @ and % in local parts.
3505 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3506 Three changes have been made:
3508 (a) There was a serious bug; a negative response to MAIL caused the whole
3509 recipient domain to be cached as invalid, thereby blocking all messages
3510 to all local parts at the same domain, from all senders. This bug has
3511 been fixed. The domain is no longer cached after a negative response to
3512 MAIL if the sender used is not empty.
3514 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3517 (c) A new callout option, "use_sender" has been added for people who want
3518 the modified behaviour.
3524 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3527 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3528 indeed breaks things for older releases.
3530 3. Added additional logging to the case where there is a problem reading data
3531 from a filter that is running in a subprocess using a pipe, in order to
3532 try to track down a specific problem.
3534 4. Testing facility fudge: when running in the test harness and attempting
3535 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3536 getting "No route to host". Convert this to a timeout.
3538 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3541 6. Some OS don't have socklen_t but use size_t instead. This affects the
3542 fifth argument of getsockopt() amongst other things. This is now
3543 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3544 can be set for individual OS. I have set it for SunOS5, OSF1, and
3545 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3546 some earlier ones do not.
3548 7. Change 4.30/15 was not doing the test caselessly.
3550 8. The standard form for an IPv6 address literal was being rejected by address
3551 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3552 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3553 this, as well as the form without the "IPv6" on the front (but only when
3554 address literals are enabled, of course).
3556 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3558 10. Exim crashed if a message with an empty sender address specified by -f
3559 encountered a router with an errors_to setting. This could be provoked only
3560 by a command such as
3564 where an empty string was supplied; "<>" did not hit this bug.
3566 11. Installed PCRE release 4.5.
3568 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3569 remained set. It is now erased.
3571 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3572 times from message ids (which are base 36 rather than the normal 62).
3574 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3575 were being counted as actual protocol errors, and logged if the log
3576 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3577 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3578 and DATA following a set of rejected RCPTs do not count as protocol errors.
3579 In other words, Exim assumes they were pipelined, though this may not
3580 actually be the case. Of course, in all cases the client gets an
3581 appropriate error code.
3583 15. If a lookup fails in an ACL condition, a message about the failure may
3584 be available; it is used if testing the ACL cannot continue, because most
3585 such messages specify what the cause of the deferral is. However, some
3586 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3587 that caused an old message to be retained and used if a later statement
3588 caused a defer, replacing the real cause of the deferral.
3590 16. If an IP address had so many PTR records that the DNS lookup buffer
3591 was not large enough to hold them, Exim could crash while trying to process
3592 the truncated data. It now detects and logs this case.
3594 17. Further to 4.21/58, another change has been made: if (and only if) the
3595 first line of a message (the first header line) ends with CRLF, a bare LF
3596 in a subsequent header line has a space inserted after it, so as not to
3597 terminate the header.
3599 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3600 unnecessarily, used atoi() instead of strtol(), and didn't check the
3601 termination when getting file sizes from file names by regex.
3603 19. Completely re-implemented the support for maildirsize files, in the light
3604 of a number of problems with the previous contributed implementation
3605 (4.30/29). In particular:
3607 . If the quota is zero, the maildirsize file is maintained, but no quota is
3610 . If the maildir directory does not exist, it is created before any attempt
3611 to write a maildirsize file.
3613 . The quota value in the file is just a cache; if the quota is changed in
3614 the transport, the new value overrides.
3616 . A regular expression is available for excluding directories from the
3619 20. The autoreply transport checks the characters in options that define the
3620 message's headers; it allows continued headers, but it was checking with
3621 isspace() after an embedded newline instead of explicitly looking for a
3624 21. If all the "regular" hosts to which an address was routed had passed their
3625 expiry times, and had not reached their retry times, the address was
3626 bounced, even if fallback hosts were defined. Now Exim should go on to try
3629 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3630 equivalent code in the SMTP transport. Some hosts send humungous responses
3631 to HELO/EHLO, more than 1024 it seems.
3633 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3634 gives compiler warnings in some environments. I've now done it "properly",
3637 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3638 (because of problems with the built-in one) was declared to return uschar *
3639 instead of char *, causing compiler failure.
3641 25. Fixed a file descriptor leak when processing alias/forward files.
3643 26. Fixed a minor format string issue in dbfn.c.
3645 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3647 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3648 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3651 29. When a sender address is verified, it is cached, to save repeating the test
3652 when there is more than one recipient in a message. However, when the
3653 verification involves a callout, it is possible for different callout
3654 options to be set for different recipients. It is too complicated to keep
3655 track of this in the cache, so now Exim always runs a verification when a
3656 callout is required, relying on the callout cache for the optimization.
3657 The overhead is duplication of the address routing, but this should not be
3660 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3661 to be verified with callout=postmaster, and the main callout worked but the
3662 postmaster check failed, the verification correctly failed. However, if a
3663 subsequent RCPT command asked for sender verification *without* the
3664 postmaster check, incorrect caching caused this verification also to fail,
3667 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3668 it was not caching the DNS options (qualify_single, search_parents) that
3669 were used when the lookup failed. A subsequent lookup with different
3670 options therefore always gave the same answer, though there were cases
3671 where it should not have. (Example: a "domains = !$mx_any" option on a
3672 dnslookup router: the "domains" option is always processed without any
3673 widening, but the router might have qualify_single set.) Now Exim uses the
3674 cached value only when the same options are set.
3676 32. Added John Jetmore's "exipick" utility to the distribution.
3678 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3679 than a timeout (e.g. a certificate is required, and is not provided), an
3680 Exim server now closes the connection immediately. Previously it waited for
3681 the client to close - but if the client is SSL, it seems that they each
3682 wait for each other, leading to a delay before one of them times out.
3684 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3685 maintained 0.8.x compatibility because I don't think many are using it, and
3686 it is clearly obsolete.
3688 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3691 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3692 client certificate was expired. A simple patch fixes this, though I don't
3693 understand the full logic of why the verify callback is called multiple
3696 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3697 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3698 which causes problems with some clients (such as the Certicom SSL Plus
3699 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3700 disables the coutermeasure allowing Eudora to connect."
3702 38. Exim was not checking that a write() to a log file succeeded. This could
3703 lead to Bad Things if a log got too big, in particular if it hit a file
3704 size limit. Exim now panics and dies if it cannot write to a log file, just
3705 as it does if it cannot open a log file.
3707 39. Modified OS/Makefile-Linux so that it now contains
3709 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3711 The two -D definitions ensure that Exim is compiled with large file
3712 support, which makes it possible to handle log files that are bigger than
3715 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3716 instance) a domain was checked against a named list that involved a lookup,
3717 causing $domain_data to be set, then another domain was checked against the
3718 same list, then the first domain was re-checked, the value of $domain_data
3719 after the final check could be wrong. In particular, if the second check
3720 failed, it could be set empty. This bug probably also applied to
3723 41. The strip_trailing_dot option was not being applied to the address given
3724 with the -f command-line option.
3726 42. The code for reading a message's header from the spool was incrementing
3727 $received_count, but never initializing it. This meant that the value was
3728 incorrect (doubled) while delivering a message in the same process in which
3729 it was received. In the most common configuration of Exim, this never
3730 happens - a fresh exec is done - but it can happen when
3731 deliver_drop_privilege is set.
3733 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3734 it now includes up to 150 characters of the unexpected data in the log
3737 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3738 and building data strings. The size of both of these buffers was 10 000
3739 bytes - far larger than anybody would *ever* want, thought I. Needless to
3740 say, somebody hit the limit. I have increased the maximum line length to
3741 20 000 and the maximum data length of concatenated lines to 100 000. I have
3742 also fixed two bugs, because there was no checking on these buffers. Tsk,
3743 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3744 buffer is too small.
3746 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3747 lsearch lookups. Now it does.
3749 46. When parsing a route_list item in a manualroute router, a fixed-length
3750 buffer was used for the list of hosts. I made this 1024 bytes long,
3751 thinking that nobody would ever have a list of hosts that long. Wrong.
3752 Somebody had a whole pile of complicated expansion conditions, and the
3753 string was silently truncated, leading to an expansion error. It turns out
3754 that it is easier to change to an unlimited length (owing to other changes
3755 that have happened since this code was originally written) than to build
3756 structure for giving a limitation error. The length of the item that
3757 expands into the list of hosts is now unlimited.
3759 47. The lsearch lookup could not handle data where the length of text line was
3760 more than 4095 characters. Such lines were truncated, leading to shortened
3761 data being returned. It should now handle lines of any length.
3763 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3764 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3767 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3768 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3769 no longer happens in comment lines. A list of replacements is now placed
3770 at the head of all of the source files, except those whose only change is
3771 to replace PERL_COMMAND in the very first #! line.
3773 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3774 messages on the queue, with a bottom-up merge sort, using code contributed
3775 by Michael Haardt. This should make operations like -bp somewhat faster on
3776 large queues. It won't affect queue runners, except when queue_run_in_order
3779 51. Installed eximstats 1.31 in the distribution.
3781 52. Added support for SRV lookups to the dnslookup router.
3783 53. If an ACL referred to $message_body or $message_body_end, the value was not
3784 reset for any messages that followed in the same SMTP session.
3786 54. The store-handling optimization for building very long strings was not
3787 differentiating between the different store pools. I don't think this
3788 actually made any difference in practice, but I've tidied it.
3790 55. While running the routers to verify a sender address, $sender_address
3791 was still set to the sender address. This is wrong, because when routing to
3792 send a bounce to the sender, it would be empty. Therefore, I have changed
3793 it so that, while verifying a sender address, $sender_address is set to <>.
3794 (There is no change to what happens when verifying a recipient address.)
3796 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3797 target A or AAAA records (if not already returned) without resetting the
3798 qualify_single or search_parents options of the DNS resolver. These are
3799 inappropriate in this case because the targets of MX and SRV records must
3800 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3801 target that, when qualified, matched something in the local domain. These
3802 two options are now turned off when doing these lookups.
3804 57. It seems that at least some releases of Reiserfs (which does not have the
3805 concept of a fixed number of inodes) returns zero and not -1 for the
3806 number of available inodes. This interacted badly with check_spool_inodes,
3807 which assumed that -1 was the "no such thing" setting. What I have done is
3808 to check that the total number of inodes is greater than zero before doing
3809 the test of how many are available.
3811 58. When a "warn" ACL statement has a log_message modifier, the message is
3812 remembered, and not repeated. This is to avoid a lot of repetition when a
3813 message has many recipients that cause the same warning to be written.
3814 Howewer, Exim was preserving the list of already written lines for an
3815 entire SMTP session, which doesn't seem right. The memory is now reset if a
3816 new message is started.
3818 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3819 debugging output unless log_rewrite was also set.
3821 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3822 of (int)(handle) when we know that handle contains (void *)(-1).
3824 61. The Exim daemon panic-logs an error return when it closes the incoming
3825 connection. However "connection reset by peer" seems to be common, and
3826 isn't really an error worthy of noting specially, so that particular error
3829 62. When Exim is trying to find all the local interfaces, it used to panic and
3830 die if the ioctl to get the interface flags failed. However, it seems that
3831 on at least one OS (Solaris 9) it is possible to have an interface that is
3832 included in the list of interfaces, but for which you get a failure error
3833 for this call. This happens when the interface is not "plumbed" into a
3834 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3835 failure of the "get flags" call assumes that the interface is down.
3837 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3838 makes life easier for people who are doing arithmetic on fields extracted
3839 from dates, where you often get leading zeros that should not be
3840 interpreted as octal.
3842 64. Added qualify_domain to the redirect router, to override the global
3845 65. If a pathologically long header line contained very many addresses (the
3846 report of this problem mentioned 10 000) and each of them was rewritten,
3847 Exim could use up a very large amount of memory. (It kept on making new
3848 copies of the header line as it rewrote, and never released the old ones.)
3849 At the expense of a bit more processing, the header rewriting function has
3850 been changed so that it no longer eats memory in this way.
3852 66. The generation of the Received: header has been moved from the time that a
3853 message starts to be received, to the time that it finishes. The timestamp
3854 in the Received: header should now be very close to that of the <= log
3855 line. There are two side-effects of this change:
3857 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3858 logged header lines no longer include the local Received: line, because
3859 it has not yet been created. The same applies to a copy of the message
3860 that is returned to a non-SMTP sender when a message is rejected.
3862 (b) When a filter file is tested using -bf, no additional Received: header
3863 is added to the test message. After some thought, I decided that this
3866 This change does not affect the value of $received_for. It is still set
3867 after address rewriting, but before local_scan() is called.
3869 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3871 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3872 gave an unhelpful panic error message, and a defer error. I have managed to
3873 change this behaviour so that it now rejects any supplied certificate,
3874 which seems right, as the list of acceptable certificates is empty.
3876 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3877 gave an unhelpful defer error. I have not managed to make this reject any
3878 supplied certificates, but the error message it gives is "no certificate
3879 supplied", which is not helpful.
3881 70. exigrep's output now also includes lines that are not associated with any
3882 message, but which match the given pattern. Implemented by a patch from
3883 Martin Sluka, which also tidied up the Perl a bit.
3885 71. Recipient callout verification, like sender verification, was using <> in
3886 the MAIL FROM command. This isn't really the right thing, since the actual
3887 sender may affect whether the remote host accepts the recipient or not. I
3888 have changed it to use the actual sender in the callout; this means that
3889 the cache record is now keyed on a recipient/sender pair, not just the
3890 recipient address. There doesn't seem to be a real danger of callout loops,
3891 since a callout by the remote host to check the sender would use <>.
3892 [SEE ABOVE: changed after hitting problems.]
3894 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3895 temporary errors. However, in the case of such a code being given after
3896 the end of a data transmission (i.e. after ".") Exim was failing to write
3897 a retry record for the message. (Yes, there was some broken host that was
3898 actually sending 8xx at this point.)
3900 73. An unknown lookup type in a host list could cause Exim to panic-die when
3901 the list was checked. (An example that provoked this was putting <; in the
3902 middle of a list instead of at the start.) If this happened during a DATA
3903 ACL check, a -D file could be left lying around. This kind of configuration
3904 error no longer causes Exim to die; instead it causes a defer errror. The
3905 incident is still logged to the main and panic logs.
3907 74. Buglet left over from Exim 3 conversion. The message "too many messages
3908 in one connection" was written to the rejectlog but not the mainlog, except
3909 when address rewriting (yes!) was being logged.
3911 75. Added write_rejectlog option.
3913 76. When a system filter was run not as root (that is, when system_filter_user
3914 was set), the values of the $n variables were not being returned to the
3915 main process; thus, they were not subsequently available in the $sn
3918 77. Added +return_path_on_delivery log selector.
3920 78. A connection timeout was being treated differently from recipients deferred
3921 when testing hosts_max_try with a message that was older than the host's
3922 retry timeout. (The host should not be counted, thus allowing all hosts to
3923 be tried at least once before bouncing.) This may have been the cause of an
3924 occasionally reported bug whereby a message would remain on the queue
3925 longer than the retry timeout, but would be bounced if a delivery was
3926 forced. I say "may" because I never totally pinned down the problem;
3927 setting up timeout/retry tests is difficult. See also the next item.
3929 79. The ultimate address timeout was not being applied to errors that involved
3930 a combination of host plus message (for example, a timeout on a MAIL
3931 command). When an address resolved to a number of possible hosts, and they
3932 were not all tried for each delivery (e.g. because of hosts_max_try), a
3933 message could remain on the queue longer than the retry timeout.
3935 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3938 81. Fixed an obscure SMTP outgoing bug which required at least the following
3939 conditions: (a) there was another message waiting for the same server;
3940 (b) the server returned 5xx to all RCPT commands in the first message so
3941 that the message was not completed; (c) the server dropped the connection
3942 or gave a negative response to the RSET that Exim sends to abort the
3943 transaction. The observed case was a dropped connection after DATA that had
3944 been sent in pipelining mode. That is, the server had advertised PIPELINING
3945 but was not implementing it correctly. The effect of the bug was incorrect
3946 behaviour, such as trying another host, and this could lead to a crash.
3952 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3953 and daemon.c were passed as pointers to ints; they should have been
3954 pointers to socklen_t variables (which are typically unsigned ints).
3956 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3959 3. Fixed a really odd bug that affected only the testing scheme; patching a
3960 certain fixed string in the binary changed the value of another string that
3961 happened to be identical to the end of the original first string.
3963 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3964 name", it returns that address as the IP address. On some operating
3965 systems (e.g. Solaris), it also passes back the IP address string as the
3966 "host name". However, on others (e.g. Linux), it passes back an empty
3967 string. Exim wasn't checking for this, and was changing the host name to an
3968 empty string, assuming it had been canonicized.
3970 5. Although rare, it is permitted to have more than one PTR record for a given
3971 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3972 all the names associated with an address, because they do in Solaris.
3973 However, it seems that they do not in Linux for data that comes from the
3974 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3975 I found this out when I moved to a new Linux workstation and tried to run
3976 the Exim test suite.
3978 To get round this problem I have changed the code so that it now does its
3979 own call to the DNS to look up PTR records when searching for a host name.
3980 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3981 addresses that are only in /etc/hosts are still found.
3983 This behaviour is, however, controlled by an option called host_lookup_
3984 order, which defaults to "bydns:byaddr". If people want to use the other
3985 order, or indeed, just use one or the other means of lookup, they can
3986 specify it in this variable.
3988 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3989 some operating systems, this comes back from gethostbyaddr() as an empty
3990 string, and this is what Exim used to test for. However, it seems that in
3991 other systems, "." is yielded. Exim now tests for this case too.
3993 7. The values of check_spool_space and check_log_space are now held internally
3994 as a number of kilobytes instead of an absolute number of bytes. If a
3995 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3996 kilobyte. This means that much larger values can be stored.
3998 8. Exim monitor: an attempt to get the action menu when not actually pointing
3999 at a message produces an empty menu entitled "No message selected". This
4000 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4001 no entries in it ("Shell widget menu has zero width and/or height"). So I
4002 have added a single, blank menu entry in this case.
4004 9. Added ${quote_local_part.
4006 10. MIME decoding is now applied to the contents of Subject: header lines when
4009 11. Now that a reference to $sender_host_address automatically causes a reverse
4010 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4011 host lookup before query-style lookups in lists that might use this
4012 variable. This has therefore been abolished, and the "net-" prefix is no
4013 longer necessary for query-style lookups.
4015 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4016 have been a typo for LFLAGS, so it has been changed.
4018 13. The install script calls Exim with "-C /dev/null" in order to find the
4019 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4020 to be output. Howeve, since Exim outputs its version number before the
4021 error, it didn't break the script. It just looked ugly. I fixed this by
4022 always allowing "-C /dev/null" if the caller is root.
4024 14. Ignore overlarge ACL variable number when reading spool file - insurance
4025 against a later release with more variables having written the file.
4027 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4028 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4029 this, as well as the form without the "IPv6" on the front.
4031 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4032 OS/Makefile-Darwin file.
4034 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4035 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4037 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4038 variable. However, some OS use ints for inodes. I've added cast to long int
4039 to get rid of the compiler warning.
4041 19. I had forgotten to lock out "/../" in configuration file names when
4042 ALT_CONFIG_PREFIX was set.
4044 20. Routers used for verification do not need to specify transports. However,
4045 if such a router generated a host list, and callout was configured, Exim
4046 crashed, because it could not find a port number from the (non-existent)
4047 transport. It now assumes port 25 in this circumstance.
4049 21. Added the -t option to exigrep.
4051 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4052 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4053 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4054 them is set, LOOKUP_LSEARCH is forced.
4056 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4058 24. Added sender and host information to the "rejected by local_scan()" log
4059 line; previously there was no indication of these.
4061 25. Added .include_if_exists.
4063 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4064 Linux. It turns out that not all file systems support this. Apparently some
4065 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4066 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4067 not supported on the file descriptor, is now ignored when Exim is trying to
4068 sync a directory. This applies only to Linux.
4070 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4072 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4073 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4076 29. Added support for maildirsize files from supplied patch (modified a bit).
4078 30. The use of :fail: followed by an empty string could lead Exim to respond to
4079 sender verification failures with (e.g.):
4081 550 Verification failed for <xxx>
4082 550 Sender verify failed
4084 where the first response line was missing the '-' that indicates it is not
4085 the final line of the response.
4087 31. The loop for finding the name of the user that called Exim had a hardwired
4088 limit of 10; it now uses the value of finduser_retries, which is used for
4089 all other user lookups.
4091 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4094 33. Exim was neglecting to zero errno before one call of strtol() when
4095 expanding a string and expecting an integer value. On some systems this
4096 resulted in spurious "integer overflow" errors. Also, it was casting the
4097 result into an int without checking.
4099 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4100 did not work. The code looks as if it has *never* worked, though it appears
4101 to have been documented since at least releast 1.62. I have made it work.
4103 35. The "timeout_DNS" error in retry rules, also documented since at least
4104 1.62, also never worked. As it isn't clear exactly what this means, and
4105 clearly it isn't a major issue, I have abolished the feature by treating it
4106 as "timeout", and writing a warning to the main and panic logs.
4108 36. The display of retry rules for -brt wasn't always showing the error code
4111 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4112 timeout_connect_A, timeout_connect_MX.
4114 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4115 to the empty sender.
4117 39. The daemon was not analysing the content of -oX till after it had closed
4118 stderr and disconnected from the controlling terminal. This meant that any
4119 syntax errors were only noted on the panic log, and the return code from
4120 the command was 0. By re-arranging the code a little, I've made the
4121 decoding happen first, so such errors now appear on stderr, and the return
4122 code is 1. However, the actual setting up of the sockets still happens in
4123 the disconnected process, so errors there are still only recorded on the
4126 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4127 connections (as happens on some IP stacks) was logged at start up time as
4128 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4129 it from "IPv6 and IPv4", which means that two separate sockets are being
4132 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4133 says whether AF_INET or AF_INET6 was passed as an argument.
4135 42. Exiwhat output was messed up when time zones were included in log
4138 43. Exiwhat now gives more information about the daemon's listening ports,
4139 and whether -tls-on-connect was used.
4141 44. The "port" option of the smtp transport is now expanded.
4143 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4144 silently ignored. Now an error message is written to the main and panic
4147 46. There's a new ACL modifier called "logwrite" which writes to a log file
4148 as soon as it is encountered.
4150 47. Added $local_user_uid and $local_user_gid at routing time.
4152 48. Exim crashed when trying to verify a sender address that was being
4155 49. Exim was recognizing only a space character after ".include". It now also
4156 recognizes a tab character.
4158 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4159 extracting the relevant information from the specification. The man page no
4160 longer contains scrambled data for the -d option, and I've added a section
4161 at the front about calling Exim under different names.
4163 51. Added "extra_headers" argument to the "mail" command in filter files.
4165 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4168 53. Installed eximstats 1.29.
4170 54. Added transport_filter_timeout as a generic transport option.
4172 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4173 Cc: header lines. This was required by RFC 822, but it not required by RFC
4176 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4177 incoming messages that did not have them. Now it does so only if the
4178 message originates locally, that is, if there is no associated remote host
4179 address. When Resent- header lines are present, this applies to the Resent-
4180 lines rather than the non-Resent- lines.
4182 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4183 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4185 58. Messages for configuration errors now include the name of the main
4186 configuration file - useful now that there may be more than one file in a
4187 list (.included file names were always shown).
4189 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4190 for those rare installations that do not start the daemon as root or run it
4191 setuid root. I've cut out the call to initgroups() if the daemon is not
4194 60. The Exim user and group can now be bound into the binary as text strings
4195 that are looked up at the start of Exim's processing.
4197 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4199 62. Added $mailstore_basename variable.
4201 63. Installed patch to sieve.c from Michael Haardt.
4203 64. When Exim failed to open the panic log after failing to open the main log,
4204 the original message it was trying to log was written to stderr and debug
4205 output, but if they were not available (the usual case in production), it
4206 was lost. Now it is written to syslog before the two lines that record the
4207 failures to open the logs.
4209 65. Users' Exim filters run in subprocesses under the user's uid. It is
4210 possible for a "deliver" command or an alias in a "personal" command to
4211 provoke an address rewrite. If logging of address rewriting is configured,
4212 this fails because the process is not running as root or exim. There may be
4213 a better way of dealing with this, but for the moment (because 4.30 needs
4214 to be released), I have disabled address rewrite logging when running a
4215 filter in a non-root, non-exim process.
4221 1. The buildconfig auxiliary program wasn't quoting the value set for
4222 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4223 not defined. This bug was masked in 4.22 by the effect that was fixed in
4226 2. Some messages that were rejected after a message id was allocated were
4227 shown as "incomplete" by exigrep. It no longer does this for messages that
4228 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4230 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4231 have allow_domain_literals set, the ID did not get logged in the <= line.
4232 Domain literals are now always recognized in Message-ID: header lines.
4234 4. The first argument for a ${extract expansion item is the key name or field
4235 number. Leading and trailing spaces in this item were not being ignored,
4236 causing some misleading effects.
4238 5. When deliver_drop_privilege was set, single queue runner processes started
4239 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4240 same command in the process it spins off) were not dropping privilege.
4242 6. When the daemon running as "exim" started a queue runner, it always
4243 re-executed Exim in the spun-off process. This is a waste of effort when
4244 deliver_drop_privilege is set. The new process now just calls the
4245 queue-runner function directly.
4251 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4254 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4255 ignored. Though the use of -oP was forcing the writing of a pid file, it
4256 was always written to the default place.
4258 3. If the message "no IP address found for host xxxx" is generated during
4259 incoming verification, it is now followed by identification of the incoming
4260 connection (so you can more easily find what provoked it).
4262 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4264 5. Added some features to "harden" Exim a bit more against certain attacks:
4266 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4267 be put in Local/Makefile. This is like the never_users runtime option,
4268 but it cannot be overridden. The default setting is "root".
4270 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4271 prefix string with which any file named in a -C command line option
4274 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4275 is retained for -C and -D only if the caller of Exim is root. Without
4276 it, the exim user may also use -C and -D and retain privilege.
4278 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4279 command line option is disabled.
4281 6. Macro names set by the -D option must start with an upper case letter, just
4282 like macro names defined in the configuration file.
4284 7. Added "dereference=" facility to LDAP.
4286 8. Two instances of the typo "uknown" in the source files are fixed.
4288 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4289 the Configure-Makefile script screwed up while processing it.
4291 10. Incorporated PCRE 4.4.
4293 11. The SMTP synchronization check was not operating right at the start of an
4294 SMTP session. For example, it could not catch a HELO sent before the client
4295 waited for the greeting. There is now a check for outstanding input at the
4296 point when the greeting is written. Because of the duplex, asynchronous
4297 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4298 way, but not yet received, when the check is performed.
4300 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4301 on TCP/IP sockets, because this apparently causes some broken clients to
4304 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4305 unchanged) from the Cygwin maintainer.
4307 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4308 was supported instead of testing for mbx. Effectively a typo.
4310 15. The spa authenticator server code was not checking that the input it
4311 received was valid base64.
4313 16. The debug output line for the "set" modifier in ACLs was not showing the
4314 name of the variable that was being set.
4316 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4318 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4319 lookup on its own. Something else was needed to trigger the lookup. For
4320 example, a match in host_lookup or the need for a host name in a host list.
4321 Now, if $sender_host_name is referenced and the host name has not yet been
4322 looked up, a lookup is performed. If the lookup fails, the variable remains
4323 empty, and $host_lookup_failed is set to "1".
4325 19. Added "eqi" as a case-independent comparison operator.
4327 20. The saslauthd authentication condition could segfault if neither service
4328 nor realm was specified.
4330 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4331 error message that was logged was misleading, and incoming SMTP
4332 connections were dropped. The message is now more accurate, and temporary
4333 errors are given to SMTP connections.
4335 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4336 (except RSET) with a 421 error, until QUIT is received. However, it was
4337 failing to send a response to QUIT.
4339 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4340 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4341 ACL is now run after the helo_try_verify_hosts code.
4343 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4344 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4345 case-independent, so other case variants are also recognized). Apparently
4346 some systems use these upper case variants.
4348 25. If more than two messages were waiting for the same host, and a transport
4349 filter was specified for the transport, Exim sent two messages over the
4350 same TCP/IP connection, and then failed with "socket operation on non-
4351 socket" when it tried to send the third.
4353 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4355 27. The extern definition of crypt16() in expand.c was not being excluded when
4356 the OS had its own crypt16() function.
4358 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4359 as a preferred synonym for return_size_limit, both as an option and as an
4362 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4364 30. Changed the default configuration ACL to relax the local part checking rule
4365 for addresses that are not in any local domains. For these addresses,
4366 slashes and pipe symbols are allowed within local parts, but the sequence
4367 /../ is explicitly forbidden.
4369 31. SPA server authentication was not clearing the challenge buffer before
4372 32. log_message in a "warn" ACL statement was writing to the reject log as
4373 well as to the main log, which contradicts the documentation and doesn't
4374 seem right (because no rejection is happening). So I have stopped it.
4376 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4377 However, I am unable to do any testing of this.
4379 34. Fixed an infelicity in the appendfile transport. When checking directories
4380 for a mailbox, to see if any needed to be created, it was accidentally
4381 using path names with one or more superfluous leading slashes; tracing
4382 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4384 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4385 contents are added to the log line that is written for every discarded
4386 recipient. (Previously a log_message setting was ignored.)
4388 36. The ${quote: operator now quotes the string if it is empty.
4390 37. The install script runs exim in order to find its version number. If for
4391 some reason other than non-existence or emptiness, which it checks, it
4392 could not run './exim', it was installing it with an empty version number,
4393 i.e. as "exim-". This error state is now caught, and the installation is
4396 38. An argument was missing from the function that creates an error message
4397 when Exim fails to connect to the socket for saslauthd authentication.
4398 This could cause Exim to crash, or give a corrupted message.
4400 39. Added isip, isip4, and isip6 to ${if conditions.
4402 40. The ACL variables $acl_xx are now saved with the message, and can be
4403 accessed later in routers, transports, and filters.
4405 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4406 strings in the file are not string-expanded.
4408 42. If a MAIL command specified a SIZE value that was too large to fit into an
4409 int variable, the check against message_size_limit failed. Such values are
4410 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4411 day this will have to be increased, but I don't think I want to be around
4412 when emails are that large.
4419 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4420 iconv() is not standard in FreeBSD.
4422 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4423 IPv6 enabled. The observed symptom was a segmentation fault on return from
4424 the function os_common_find_running_interfaces() in src/os.c.
4426 3. In the check_special_case() function in daemon.c I had used "errno" as an
4427 argument name, which causes warnings on some systems. This was basically a
4428 typo, since it was named "eno" in the comments!
4430 4. The code that waits for the clock to tick (at a resolution of some fraction
4431 of a second) so as to ensure message-id uniqueness was always waiting for
4432 at least one whole tick, when it could have waited for less. [This is
4433 almost certainly not relevant at current processor speeds, where it is
4434 unlikely to ever wait at all. But we try to future-proof.]
4436 5. The function that sleeps for a time interval that includes fractions of a
4437 second contained a race. It did not block SIGALRM between setting the
4438 timer, and suspending (a couple of lines later). If the interval was short
4439 and the sigsuspend() was delayed until after it had expired, the suspension
4440 never ended. On busy systems this could lead to processes getting stuck for
4443 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4444 process, before it forks any delivery processes. The open lookup caching
4445 mechanism meant that the open file or database connection was passed into
4446 the delivery process. The problem was that delivery processes always tidy
4447 up cached lookup data. This could cause a problem for the next delivery
4448 process started by the queue runner, because the external queue runner
4449 process does not know about the closure. So the next delivery process
4450 still has data in the lookup cache. In the case of a file lookup, there was
4451 no problem because closing a file descriptor in a subprocess doesn't affect
4452 the parent. However, if the lookup was caching a connection to a database,
4453 the connection was closed, and the second delivery process was likely to
4454 see errors such as "PGSQL: query failed: server closed the connection
4455 unexpectedly". The problem has been fixed by closing all cached lookups
4456 in a queue runner before running a delivery process.
4458 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4459 seem to have the "const" qualifier which it has on other OS. I've
4462 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4463 *of the same type* (client or server) with the same public name that an
4464 error should be diagnosed.
4466 9. When Exim looked up a host name for an IP address, but failed to find the
4467 original IP address when looking up the host name (a safety check), it
4468 output the message "<ip address> does not match any IP for NULL", which was
4469 confusing, to say the least. The bug was that the host name should have
4470 appeared instead of "NULL".
4472 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4473 user that is built into the binary, and the -C or -D options is used, root
4474 privilege is dropped before the configuration file is read. In addition,
4475 logging is switched to stderr instead of the normal log files. If the
4476 configuration then re-defines the Exim user, the unprivileged environment
4477 is probably not what is expected, so Exim logs a panic warning message (but
4480 However, if deliver_drop_privilege is set, the unprivileged state may well
4481 be exactly what is intended, so the warning has been cut out in that case,
4482 and Exim is allowed to try to write to its normal log files.
4488 1. smtp_return_error_details was not giving details for temporary sender
4489 or receiver verification errors.
4491 2. Diagnose a configuration error if two authenticators have the same public
4494 3. Exim used not to create the message log file for a message until the first
4495 delivery attempt. This could be confusing when incoming messages were held
4496 for policy or load reasons. The message log file is now created at the time
4497 the message is received, and an initial "Received" line is written to it.
4499 4. The automatically generated man page for command line options had a minor
4500 bug that caused no ill effects; however, a more serious problem was that
4501 the procedure for building the man page automatically didn't always
4502 operate. Consequently, release 4.20 contains an out-of-date version. This
4503 shouldn't happen again.
4505 5. When building Exim with embedded Perl support, the script that builds the
4506 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4507 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4509 6. The freeze_tell option was not being used for messages that were frozen on
4510 arrival, either by an ACL or by local_scan().
4512 7. Added the smtp_incomplete_transaction log selector.
4514 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4515 was accepting AUTH without a new EHLO.
4517 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4518 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4521 10. Logging of TCP/IP connections (when configured) now happens in the main
4522 daemon process instead of the child process, so that the TCP/IP connection
4523 count is more accurate (but it can never be perfect).
4525 11. The use of "drop" in a nested ACL was not being handled correctly in the
4526 outer ACL. Now, if condition failure induced by the nested "drop" causes
4527 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4528 or "require"), the connection is dropped.
4530 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4531 that yield "discard" can now be used with an "accept" or a "discard" verb,
4532 but an error is generated for any others (because I can't see a useful way
4533 to define what should happen).
4535 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4536 are now processed in the same way as lines in the Exim configuration file.
4537 In particular, continuation lines are supported.
4539 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4541 15. Added -ti meaning -t -i.
4543 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4544 domains, and warn by logging if others are found.
4546 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4547 returned from GIFCONF when Exim is trying to find the list of interfaces on
4548 a host. The code in os.c has been modified to copy each ifreq to an aligned
4549 structure in all cases.
4551 Also, in some cases, the returned ifreq's were being copied to a 'struct
4552 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4553 means the last couple of bytes of an IPv6 address could be chopped if the
4554 ifreq contained only a normal sockaddr (14 bytes storage).
4556 18. Named domain lists were not supported in the hosts_treat_as_local option.
4557 An entry such as +xxxx was not recognized, and was treated as a literal
4560 19. Ensure that header lines added by a DATA ACL are included in the reject log
4561 if the ACL subsequently rejects the message.
4563 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4564 MD5 (which is deprecated).
4566 21. When testing a filter file using -bf, Exim was writing a message when it
4567 took the sender from a "From " line in the message, but it was not doing so
4568 when it took $return_path from a Return-Path: header line. It now does.
4570 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4571 with a valid header line field name (a series of printing characters
4572 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4574 23. Changed "disc" in the source to "disk" to conform to the documentation and
4575 the book and for uniformity.
4577 24. Ignore Sendmail's -Ooption=value command line item.
4579 25. When execve() failed while trying to run a command in a pipe transport,
4580 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4581 could be confused with a return value of 69 from the command itself. This
4582 has been changed to 127, the value the shell returns if it is asked to run
4583 a non-existent command. The wording for the related log line suggests a
4584 non-existent command as the problem.
4586 26. If received_header_text expands to an empty string, do not add a Received:
4587 header line to the message. (Well, it adds a token one on the spool, but
4588 marks it "old" so that it doesn't get used or transmitted.)
4590 27. Installed eximstats 1.28 (addition of -nt option).
4592 28. There was no check for failure on the call to getsockname() in the daemon
4593 code. This can fail if there is a shortage of resources on the system, with
4594 ENOMEM, for example. A temporary error is now given on failure.
4596 29. Contrary to the C standard, it seems that in some environments, the
4597 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4598 program. Exim now does this explicitly; it affects the formatting of
4599 timestamps using strftime().
4601 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4602 complaints. I've now initialized all the variables, to avoid this.
4604 32. Header lines added by a system filter were not being "seen" during
4605 transport-time rewrites.
4607 33. The info_callback() function passed to OpenSSL is set up with type void
4608 (*)(SSL *, int, int), as described somewhere. However, when calling the
4609 function (actually a macro) that sets it up, the type void(*)() is
4610 expected. I've put in a cast to prevent warnings from picky compilers.
4612 34. If a DNS black list lookup found a CNAME record, but there were no A
4613 records associated with the domain it pointed at, Exim crashed.
4615 35. If a DNS black list lookup returned more than one A record, Exim ignored
4616 all but the first. It now scans all returned addresses if a particular IP
4617 value is being sought. In this situation, the contents of the
4618 $dnslist_value variable are a list of all the addresses, separated by a
4621 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4622 to accept a host name and all its aliases if the forward lookup for any of
4623 them yielded the IP address of the incoming connection. Now it accepts only
4624 those names whose forward lookup yields the correct IP address. Any other
4625 names are discarded. This closes a loophole whereby a rogue DNS
4626 administrator could create reverse DNS records to break through a
4627 wildcarded host restriction in an ACL.
4629 37. If a user filter or a system filter that ran in a subprocess used any of
4630 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4631 the wrong values were passed to the pipe command ($thisaddress had the
4632 value of $0, $0 had the value of $1, etc). This bug was introduced by
4633 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4635 38. Improved the line breaking for long SMTP error messages from ACLs.
4636 Previously, if there was no break point between 40 and 75 characters, Exim
4637 left the rest of the message alone. Two changes have been made: (a) I've
4638 reduced the minimum length to 35 characters; (b) if it can't find a break
4639 point between 35 and 75 characters, it looks ahead and uses the first one
4640 that it finds. This may give the occasional overlong line, but at least the
4641 remaining text gets split now.
4643 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4644 file descriptors might be low, and that setting 1000 would always raise it.
4645 It turns out that in some environments, the limit is already over 1000 and
4646 that lowering it causes trouble. So now Exim takes care not to decrease it.
4648 40. When delivering a message, the value of $return_path is set to $sender_
4649 address at the start of routing (routers may change the value). By an
4650 oversight, this default was not being set up when an address was tested by
4651 -bt or -bv, which affected the outcome if any router or filter referred to
4654 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4655 log only when "message" or "log_message" are set. However, if one of the
4656 conditions was an address verification, or a call to a nested ACL, the
4657 messages generated by the underlying test were being passed through. This
4658 no longer happens. The underlying message is available in $acl_verify_
4659 message for both "message" and "log_message" expansions, so it can be
4660 passed through if needed.
4662 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4663 new expansion $bh_ to give the encoded byte string without charset
4664 translation. Translation happens only if iconv() is available; HAVE_ICONV
4665 indicates this at build time. HEADERS_CHARSET gives the charset to
4666 translate to; headers_charset can change it in the configuration, and
4667 "headers charset" can change it in an individual filter file.
4669 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4670 that creates RFC 2047 encoded "words" labels them as that charset instead
4671 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4672 expansion operator; (ii) when Exim creates a From: line for a local
4673 message; (iii) when a header line is rewritten to include a "phrase" part.
4675 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4676 buggy, causing it to skip the first lines of messages whose message ID
4677 ended in 'D'. This would not have bitten before Exim release 4.14, because
4678 message IDs were unlikely to end in 'D' before then. The effect was to have
4679 incorrect size information for certain domains.
4681 45. #include "config.h" was missing at the start of the crypt16.c module. This
4682 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4685 46. If there was a timeout during a "random" callout check, Exim treated it as
4686 a failure of the random address, and carried on sending RSET and the real
4687 address. If the delay was just some slowness somewhere, the response to the
4688 original RCPT would be taken as a response to RSET and so on, causing
4689 mayhem of various kinds.
4691 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4692 when I implemented it. It didn't allow for the fact that some option values
4693 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4694 the right test for positive values.
4696 48. Domain names in DNS records are case-independent. Exim always looks them up
4697 in lower case. Some resolvers return domain names in exactly the case they
4698 appear in the zone file, that is, they may contain uppercase letters. Not
4699 all resolvers do this - some return always lower case. Exim was treating a
4700 change of case by a resolver as a change of domain, similar to a widening
4701 of a domain abbreviation. This triggered its re-routing code and so it was
4702 trying to route what was effectively the same domain again. This normally
4703 caused routing to fail (because the router wouldn't handle the domain
4704 twice). Now Exim checks for this case specially, and just changes the
4705 casing of the domain that it ultimately uses when it transmits the message
4708 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4711 50. If a filter generated a file delivery with a non-absolute name (possible if
4712 no home directory exists for the router), the forbid_file option was not
4715 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4716 the existing equality matching.
4718 52. Exim was using ints instead of ino_t variables in some places where it was
4719 dealing with inode numbers.
4721 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4722 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4723 called TMPDIR, and if it finds it is different, it changes its value.
4725 54. The smtp_printf() function is now made available to local_scan() so
4726 additional output lines can be written before returning. There is also an
4727 smtp_fflush() function to enable the detection of a dropped connection.
4728 The variables smtp_input and smtp_batched_input are exported to
4731 55. Changed the default runtime configuration: the message "Unknown user"
4732 has been removed from the ACL, and instead placed on the localuser router,
4733 using the cannot_route_message feature. This means that any verification
4734 failures that generate their own messages won't get overridden. Similarly,
4735 the "Unrouteable address" message that was in the ACL for unverifiable
4736 relay addresses has also been removed.
4738 56. Added hosts_avoid_esmtp to the smtp transport.
4740 57. The exicyclog script was not checking for the esoteric option
4741 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4742 will work only if exicyclog is run under the appropriate euid.
4744 58. Following a discussion on the list, the rules by which Exim recognises line
4745 endings on incoming messages have been changed. The -dropcr and drop_cr
4746 options are now no-ops, retained only for backwards compatibility. The
4747 following line terminators are recognized: LF CRLF CR. However, special
4748 processing applies to CR:
4750 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4751 nor a local message in the state where . is a terminator.
4753 (ii) If a bare CR is encountered in a header line, an extra space is added
4754 after the line terminator so as not to end the header. The reasoning
4755 behind this is that bare CRs in header lines are most likely either
4756 to be mistakes, or people trying to play silly games.
4758 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4759 was being incorrectly given as 18 bytes larger than it should have been.
4760 This is a VOB (very old bug).
4762 60. This may never have affected anything current, but just in case it has:
4763 When the local host is found other than at the start of a list of hosts,
4764 the local host, those with the same MX, and any that follow, are discarded.
4765 When the list in question was part of a longer list of hosts, the following
4766 hosts (not currently being processed) were also being discarded. This no
4767 longer happens. I'm not sure if this situation could ever has previously
4770 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4773 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4774 header. This is something that is recommended in a new Internet Draft, and
4775 is something that is documented as being done by Sendmail. There are two
4776 possible values. For messages generated by the autoreply transport, Exim
4779 Auto-Submitted: auto-replied
4781 whereas for all other generated messages (e.g. bounces) it adds
4783 Auto-Submitted: auto-generated
4785 63. The "personal" condition in filters now includes a test for the
4786 Auto-Submitted: header. If it contains the string "auto-" the message it
4787 not considered personal.
4789 64. Added rcpt_include_affixes as a generic transport option.
4791 65. Added queue_only_override (default true).
4793 66. Added the syslog_duplication option.
4795 67. If what should have been the first header line of a message consisted of
4796 a space followed by a colon, Exim was mis-interpreting it as a header line.
4797 It isn't of course - it is syntactically invalid and should therefore be
4798 treated as the start of the message body. The misbehaviour could have
4799 caused a number of strange effects, including loss of data in subsequent
4800 header lines, and spool format errors.
4802 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4803 client host had authenticated. This control can now be exercised by an ACL
4804 for more flexibility.
4806 69. By default, callouts do not happen when testing with -bh. There is now a
4807 variant, -bhc, which does actually run the callout code, including
4808 consulting and updating the callout cache.
4810 70. Added support for saslauthd authentication, courtesy of Alexander
4813 71. If statvfs() failed on the spool or log directories while checking their
4814 size for availability, Exim confusingly gave the error "space shortage".
4815 Furthermore, in debugging mode it crashed with a floating point exception.
4816 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4817 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4818 serious problem, Exim now writes to the main and panic logs when this
4819 happens, with details of the failure. It then refuses to accept the
4820 incoming message, giving the message "spool directory problem" or "log
4821 directory problem" with a 421 code for SMTP messages.
4823 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4824 0, 1, and 2 exist, because some OS complain for execs without them (see
4825 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4826 descriptors. However, the code omitted to check that the open succeeded,
4827 causing mysterious errors if for some reason the permissions on /dev/null
4828 got screwed. Now Exim writes a message to the main and panic logs, and
4829 bombs out if it can't open /dev/null.
4831 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4832 interact so that it is all more flexible. It is supposed to remain
4833 backwards compatible. Also added extra_local_interfaces.
4835 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4836 to bomb out with an assertion failure - to the client this appears as a
4837 connection drop. This problem occurs in the part of the code that was taken
4838 from the Samba project. Fortunately, the assertion is in a very simple
4839 function, so I have fixed this by reproducing the function inline in the
4840 one place where it is called, and arranging for authentication to fail
4841 instead of killing the process with assert().
4843 75. The SPA client code was not working when the server requested OEM rather
4844 than Unicode encoding.
4846 76. Added code to make require_files with a specific uid setting more usable in
4847 the case where statting the file as root fails - usually a non-root-mounted
4848 NFS file system. When this happens and the failure is EACCES, Exim now
4849 forks a subprocess and does the per-uid checking as the relevant uid.
4851 77. Added process_log_path.
4853 78. If log_file_path was not explicitly set, a setting of check_log_space or
4854 check_log_inodes was ignored.
4856 79. If a space check for the spool or log partitions fails, the incident is now
4857 logged. Of course, in the latter case the data may get lost...
4859 80. Added the %p formatting code to string_format() so that it can be used to
4860 print addresses in debug_print(). Adjusted all the address printing in the
4861 debugging in store.c to use %p rather than %d.
4863 81. There was a concern that a line of code in smtp_in.c could overflow a
4864 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4865 initially expressed, the concern was not well-founded, because trailing
4866 spaces are removed early. However, if the trailing spaces were followed by
4867 a NULL, they did not get removed, so the overflow was possible. Two fixes
4870 (a) I re-wrote the offending code in a cleaner fashion.
4871 (b) If an incoming SMTP command contains a NULL character, it is rejected
4874 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4875 runs initgroups(), so that if the Exim user is in any additional groups,
4876 they will be used during message reception.
4882 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob