1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.506 2007/05/08 13:08:22 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PH/01 Another patch from the Sieve maintainer.
11 PH/02 When an IPv6 address is converted to a string for single-key lookup
12 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
13 dots are used instead of colons so that keys in lsearch files need not
14 contain colons. This was done some time before quoting was made available
15 in lsearch files. However, iplsearch files do require colons in IPv6 keys
16 (notated using the quote facility) so as to distinguish them from IPv4
17 keys. This meant that lookups for IP addresses in host lists did not work
18 for iplsearch lookups.
20 This has been fixed by arranging for IPv6 addresses to be expressed with
21 colons if the lookup type is iplsearch. This is not incompatible, because
22 previously such lookups could never work.
24 The situation is now rather anomolous, since one *can* have colons in
25 ordinary lsearch keys. However, making the change in all cases is
26 incompatible and would probably break a number of configurations.
32 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
33 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
36 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
39 PH/02 In an ACL statement such as
41 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
43 if a client was not listed at all, or was listed with a value other than
44 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
45 the condition was not true (as it should be), so access was not denied.
46 The bug was that the ! inversion was incorrectly passed on to the second
47 item. This has been fixed.
49 PH/03 Added additional dnslists conditions == and =& which are different from
50 = and & when the dns lookup returns more than one IP address.
52 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
53 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
55 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
56 FSYNC, which compiles an option called disable_fsync that allows for
57 bypassing fsync(). The documentation is heavily laced with warnings.
59 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
61 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
62 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
63 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
64 including adding "make clean"; (3) Added -fPIC when compiling the test
65 dynamically loaded module, to get rid of a warning.
67 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
68 message fails, move_frozen_messages = true and ignore_bounce_errors_after
69 = 0s. The bug is otherwise harmless.
71 PH/07 There was a bug in the dovecot authenticator such that the value of
72 $auth1 could be overwritten, and so not correctly preserved, after a
73 successful authentication. This usually meant that the value preserved by
74 the server_setid option was incorrect.
76 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
78 PH/09 Installed PCRE release 7.0.
80 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
81 run for batched SMTP input. It is now run at the start of every message
82 in the batch. While fixing this I discovered that the process information
83 (output by running exiwhat) was not always getting set for -bs and -bS
84 input. This is fixed, and it now also says "batched" for BSMTP.
86 PH/11 Added control=no_pipelining.
88 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
89 patch, slightly modified), and move the expansion of helo_data till after
90 the connection is made in the smtp transport (so it can use these
93 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
95 PH/14 Added log_selector = +pid.
97 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
99 PH/16 Add ${if forany and ${if forall.
101 PH/17 Added dsn_from option to vary the From: line in DSNs.
103 PH/18 Flush SMTP output before performing a callout, unless control =
104 no_callout_flush is set.
106 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
107 was true (the default) a successful delivery failed to delete the retry
108 item, thus causing premature timeout of the address. The bug is now
111 PH/20 Added hosts_avoid_pipelining to the smtp transport.
113 PH/21 Long custom messages for fakedefer and fakereject are now split up
114 into multiline reponses in the same way that messages for "deny" and
115 other ACL rejections are.
117 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
118 with slight modification.
120 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
121 draft, changing the syntax and factoring some duplicate code".
123 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
124 for deliveries of the second and subsequent messages over the same SMTP
127 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
128 ${reduce, with only minor "tidies".
130 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
132 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
133 expansion side effects.
135 PH/27 When a message times out after an over-quota error from an Exim-imposed
136 quota, the bounce message says "mailbox is full". This message was not
137 being given when it was a system quota that was exceeded. It now should
140 MH/03 Made $recipients available in local_scan(). local_scan() already has
141 better access to the recipient list through recipients_list[], but
142 $recipients can be useful in postmaster-provided expansion strings.
144 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
145 in the case of a MAIL command with additional options following the
146 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
147 were accidentally chopped off.
149 PH/29 SMTP synchronization checks are implemented when a command is read -
150 there is a check that no more input is waiting when there shouldn't be
151 any. However, for some commands, a delay in an ACL can mean that it is
152 some time before the response is written. In this time, more input might
153 arrive, invalidly. So now there are extra checks after an ACL has run for
154 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
155 pipelining has not been advertised.
157 PH/30 MH's patch to allow iscntrl() characters to be list separators.
159 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
160 returned in the SMTP response when smtp_return_error_details was false.
163 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
164 instead of the C I/O that was originally supplied, because problems were
167 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
168 Exim which did not show up earlier: it was assuming that a call to
169 SSL_CTX_set_info_callback() might give an error value. In fact, there is
170 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
171 was a macro that became an assignment, so it seemed to work. This has
172 changed to a proper function call with a void return, hence the compile
173 error. Exim's code has been fixed.
175 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
178 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
180 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
182 PH/37 If a message is not accepted after it has had an id assigned (e.g.
183 because it turns out to be too big or there is a timeout) there is no
184 "Completed" line in the log. When some messages of this type were
185 selected by exigrep, they were listed as "not completed". Others were
186 picked up by some special patterns. I have improved the selection
187 criteria to be more general.
189 PH/38 The host_find_failed option in the manualroute router can now be set
190 to "ignore", to completely ignore a host whose IP address cannot be
191 found. If all hosts are ignored, the behaviour is controlled by the new
192 host_all_ignored option.
194 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
195 homing or because of multiple MX records with /mx) generated more than
196 one IP address, and the following item turned out to be the local host,
197 all the secondary addresses of the first item were incorrectly removed
198 from the list, along with the local host and any following hosts (which
199 is what is supposed to happen).
201 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
202 whoever called Exim into the -H file. In the case of the daemon it was
203 behaving confusingly. When first started, it used values for whoever
204 started the daemon, but after a SIGHUP it used the Exim user (because it
205 calls itself on a restart). I have changed the code so that it now always
208 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
209 message are rejected with the same error (e.g. no authentication or bad
210 sender address), and a DATA command is nevertheless sent (as can happen
211 with PIPELINING or a stupid MUA), the error message that was given to the
212 RCPT commands is included in the rejection of the DATA command. This is
213 intended to be helpful for MUAs that show only the final error to their
216 PH/42 Another patch from the Sieve maintainer.
218 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
219 Eximstats - Fixed some broken HTML links and added missing column headers
221 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
222 columns for Rejects, Temp Rejects, Ham, and Spam rows.
224 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
226 PH/43 Yet another patch from the Sieve maintainer.
228 PH/44 I found a way to check for a TCP/IP connection going away before sending
229 the response to the final '.' that terminates a message, but only in the
230 case where the client has not sent further data following the '.'
231 (unfortunately, this is allowed). However, in many cases there won't be
232 any further data because there won't be any more messages to send. A call
233 to select() can be used: if it shows that the input is "ready", there is
234 either input waiting, or the socket has been closed. An attempt to read
235 the next input character can distinguish the two cases. Previously, Exim
236 would have sent an OK response which the client would never have see.
237 This could lead to message repetition. This fix should cure that, at
238 least in a lot of common cases.
240 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
241 advertised in response to EHLO.
247 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
248 fixed by 4.65/MH/01 (is this a record?) are fixed:
250 (i) An empty string was always treated as zero by the numeric comparison
251 operators. This behaviour has been restored.
253 (ii) It is documented that the numeric comparison operators always treat
254 their arguments as decimal numbers. This was broken in that numbers
255 starting with 0 were being interpreted as octal.
257 While fixing these problems I realized that there was another issue that
258 hadn't been noticed. Values of message_size_limit (both the global option
259 and the transport option) were treated as octal if they started with 0.
260 The documentation was vague. These values are now always treated as
261 decimal, and I will make that clear in the documentation.
267 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
268 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
271 MH/01 Don't check that the operands of numeric comparison operators are
272 integers when their expansion is in "skipping" mode (fixes bug
273 introduced by 4.64-PH/07).
275 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
276 child addresses, Exim now panics and dies. Previously, because the count
277 is held in a short int, deliveries were likely to be lost. As such a
278 large number of recipients for a single message is ridiculous
279 (performance will be very, very poor), I have chosen to impose a limit
280 rather than extend the field.
286 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
287 leftover -K file (the existence of which was triggered by #402).
288 While we were at it, introduced process PID as part of the -K
289 filename. This should rule out race conditions when creating
292 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
293 processing considerably. Previous code took too long for large mails,
294 triggering a timeout which in turn triggers #401.
296 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
297 in the DK code in transports.c. sendfile() is not really portable,
298 hence the _LINUX specificness.
300 TF/01 In the add_headers option to the mail command in an Exim filter,
301 there was a bug that Exim would claim a syntax error in any
302 header after the first one which had an odd number of characters
305 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
306 callout verification, Exim cached a "reject" for the entire domain. This
307 is correct for most verifications, but it is not correct for a recipient
308 verification with use_sender or use_postmaster set, because in that case
309 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
310 case of MAIL FROM:<> rejection from other early rejections (e.g.
311 rejection of HELO). When verifying a recipient using a non-null MAIL
312 address, the cache is ignored if it shows MAIL FROM:<> rejection.
313 Whatever the result of the callout, the value of the domain cache is
314 left unchanged (for any other kind of callout, getting as far as trying
315 RCPT means that the domain itself is ok).
317 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
320 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
321 manifest itself as EPIPE rather than ECONNECT. When tidying away a
322 session, the daemon ignores ECONNECT errors and logs others; it now
323 ignores EPIPE as well.
325 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
326 (quoted-printable decoding).
328 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
329 later the small subsequent patch to fix an introduced bug.
331 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
333 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
335 PH/08 An error is now given if message_size_limit is specified negative.
337 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
338 to be given (somewhat) arbitrary names.
340 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
343 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
344 miscellaneous code fixes
346 PH/10 Added the log_reject_target ACL modifier to specify where to log
349 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
350 hostname. This is wrong, because it relates to the incoming message (and
351 probably the interface on which it is arriving) and not to the outgoing
352 callout (which could be using a different interface). This has been
353 changed to use the value of the helo_data option from the smtp transport
354 instead - this is what is used when a message is actually being sent. If
355 there is no remote transport (possible with a router that sets up host
356 addresses), $smtp_active_hostname is used.
358 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
359 tweaks were necessary in order to get it to work (see also 21 below):
360 (a) The code assumed that strncpy() returns a negative number on buffer
361 overflow, which isn't the case. Replaced with Exim's string_format()
363 (b) There were several signed/unsigned issues. I just did the minimum
364 hacking in of casts. There is scope for a larger refactoring.
365 (c) The code used strcasecmp() which is not a standard C function.
366 Replaced with Exim's strcmpic() function.
367 (d) The code set only $1; it now sets $auth1 as well.
368 (e) A simple test gave the error "authentication client didn't specify
369 service in request". It would seem that Dovecot has changed its
370 interface. Fortunately there's a specification; I followed it and
371 changed what the client sends and it appears to be working now.
373 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
376 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
377 address A is aliased to B and C, where B exists and C does not. Without
378 -v the output is "A verified" because verification stops after a
379 successful redirection if more than one address is generated. However,
380 with -v the child addresses are also verified. Exim was outputting "A
381 failed to verify" and then showing the successful verification for C,
382 with its parentage. It now outputs "B failed to verify", showing B's
383 parentage before showing the successful verification of C.
385 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
386 look up a TXT record in a specific list after matching in a combined
389 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
390 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
391 they consult the DNS. I had assumed they would set it the way they
392 wanted; and indeed my experiments on Linux seem to show that in some
393 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
394 To be on the safe side, however, I have now made the interface to
395 host_find_byname() similar to host_find_bydns(), with an argument
396 containing the DNS resolver options. The host_find_byname() function now
397 sets these options at its start, just as host_find_bydns() does. The smtp
398 transport options dns_qualify_single and dns_search_parents are passed to
399 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
400 of host_find_byname() use the default settings of RES_DEFNAMES
401 (qualify_single) but not RES_DNSRCH (search_parents).
403 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
404 spool_read_header() do less string testing, by means of a preliminary
405 switch on the second character of optional "-foo" lines. (This is
406 overdue, caused by the large number of possibilities that now exist.
407 Originally there were few.) While I was there, I also converted the
408 str(n)cmp tests so they don't re-test the leading "-" and the first
409 character, in the hope this might squeeze out yet more improvement.
411 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
412 flag allowing group syntax was set by the header_syntax check but not
413 turned off, possible causing trouble later; (2) The flag was not being
414 set at all for the header_verify test, causing "group"-style headers to
415 be rejected. I have now set it in this case, and also caused header_
416 verify to ignore an empty address taken from a group. While doing this, I
417 came across some other cases where the code for allowing group syntax
418 while scanning a header line wasn't quite right (mostly, not resetting
419 the flag correctly in the right place). These bugs could have caused
420 trouble for malformed header lines. I hope it is now all correct.
422 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
423 with the "reply" argument non-NULL. The code, however (which originally
424 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
425 but it didn't always do it. This confused somebody who was copying the
426 code for some other use. I have removed all the tests.
428 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
429 feature that was used to support insecure browsers during the U.S. crypto
430 embargo. It requires special client support, and Exim is probably the
431 only MTA that supported it -- and would never use it because real RSA is
432 always available. This code has been removed, because it had the bad
433 effect of slowing Exim down by computing (never used) parameters for the
434 RSA_EXPORT functionality.
436 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
437 authenticator to fail if there's a tab character in the incoming data
438 (there should never be unless someone is messing about, as it's supposed
439 to be base64-encoded). Also added, on Timo's advice, the "secured" option
440 if the connection is using TLS or if the remote IP is the same as the
441 local IP, and the "valid-client-cert option" if a client certificate has
444 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
445 authenticators. This can be used for authorization after authentication
446 succeeds. (In the case of plaintext, it servers for both authentication
449 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
450 if any retry times were supplied.
452 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
453 connection, where there is no client IP address to check. In this
454 situation, the verify now always succeeds.
456 PH/25 Applied John Jetmore's -Mset patch.
458 PH/26 Added -bem to be like -Mset, but loading a message from a file.
460 PH/27 In a string expansion for a processed (not raw) header when multiple
461 headers of the same name were present, leading whitespace was being
462 removed from all of them, but trailing whitespace was being removed only
463 from the last one. Now trailing whitespace is removed from each header
464 before concatenation. Completely empty headers in a concatenation (as
467 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
468 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
470 PH/29 [Removed. This was a change that I later backed out, and forgot to
471 correct the ChangeLog entry (that I had efficiently created) before
472 committing the later change.]
474 PH/30 Exim was sometimes attempting to deliver messages that had suffered
475 address errors (4xx response to RCPT) over the same connection as other
476 messages routed to the same hosts. Such deliveries are always "forced",
477 so retry times are not inspected. This resulted in far too many retries
478 for the affected addresses. The effect occurred only when there were more
479 hosts than the hosts_max_try setting in the smtp transport when it had
480 the 4xx errors. Those hosts that it had tried were not added to the list
481 of hosts for which the message was waiting, so if all were tried, there
482 was no problem. Two fixes have been applied:
484 (i) If there are any address or message errors in an SMTP delivery, none
485 of the hosts (tried or untried) are now added to the list of hosts
486 for which the message is waiting, so the message should not be a
487 candidate for sending over the same connection that was used for a
488 successful delivery of some other message. This seems entirely
489 reasonable: after all the message is NOT "waiting for some host".
490 This is so "obvious" that I'm not sure why it wasn't done
491 previously. Hope I haven't missed anything, but it can't do any
492 harm, as the worst effect is to miss an optimization.
494 (ii) If, despite (i), such a delivery is accidentally attempted, the
495 routing retry time is respected, so at least it doesn't keep
496 hammering the server.
498 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
499 in ${readsocket because some servers need this prod.
501 PH/32 Added some extra debug output when updating a wait-xxx database.
503 PH/33 The hint "could be header name not terminated by colon", which has been
504 given for certain expansion errors for a long time, was not being given
505 for the ${if def:h_colon_omitted{... case.
507 PH/34 The spec says: "With one important exception, whenever a domain list is
508 being scanned, $domain contains the subject domain." There was at least
509 one case where this was not true.
511 PH/35 The error "getsockname() failed: connection reset by peer" was being
512 written to the panic log as well as the main log, but it isn't really
513 panic-worthy as it just means the connection died rather early on. I have
514 removed the panic log writing for the ECONNRESET error when getsockname()
517 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
518 runs only) independently of the message's sender address. This meant
519 that, if the 4xx error was in fact related to the sender, a different
520 message to the same recipient with a different sender could confuse
521 things. In particualar, this can happen when sending to a greylisting
522 server, but other circumstances could also provoke similar problems.
523 I have changed the default so that the retry time for these errors is now
524 based a combination of the sender and recipient addresses. This change
525 can be overridden by setting address_retry_include_sender=false in the
528 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
529 remote server are returned as part of bounce messages. This was not
530 happening for LMTP over a pipe (the lmtp transport), but now it is the
531 same for both kinds of LMTP.
533 PH/38 Despite being documented as not happening, Exim was rewriting addresses
534 in header lines that were in fact CNAMEs. This is no longer the case.
536 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
537 and queue runs started by the daemon processed all messages. This has
538 been fixed so that -R and -S can now usefully be given with -q<time>.
540 PH/40 Import PCRE release 6.7 (fixes some bugs).
542 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
544 PH/42 Give an error if -q is specified more than once.
546 PH/43 Renamed the variables $interface_address and $interface_port as
547 $received_ip_address and $received_port, to make it clear that these
548 values apply to message reception, and not to the outgoing interface when
549 a message is delivered. (The old names remain recognized, of course.)
551 PH/44 There was no timeout on the connect() call when using a Unix domain
552 socket in the ${readsocket expansion. There now is.
554 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
555 be meaningful with "accept".
557 SC/01 Eximstats V1.43
558 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
560 SC/02 Eximstats V1.44
561 Use a glob alias rather than an array ref in the generated
562 parser. This improves both readability and performance.
564 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
565 Collect SpamAssassin and rejection statistics.
566 Don't display local sender or destination tables unless
567 there is data to show.
568 Added average volumes into the top table text output.
570 SC/04 Eximstats V1.46
571 Collect data on the number of addresses (recipients)
572 as well as the number of messages.
574 SC/05 Eximstats V1.47
575 Added 'Message too big' to the list of mail rejection
576 reasons (thanks to Marco Gaiarin).
578 SC/06 Eximstats V1.48
579 Mainlog lines which have GMT offsets and are too short to
580 have a flag are now skipped.
582 SC/07 Eximstats V1.49 (Alain Williams)
583 Added the -emptyok flag.
585 SC/08 Eximstats V1.50
586 Fixes for obtaining the IP address from reject messages.
588 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
589 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
590 whitesspace changes from 4.64-PH/27
592 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
595 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
596 are found, allow negative numbers in numeric criteria)
598 JJ/06 exipick.20061117.2, added new $message_body_missing variable
600 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
601 to match changes made in 4.64-PH/43
603 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
605 PH/47 Put in an explicit test for a DNS lookup of an address record where the
606 "domain" is actually an IP address, and force a failure. This locks out
607 those revolvers/nameservers that support "A-for-A" lookups, in
608 contravention of the specifications.
610 PH/48 When a host name was looked up from an IP address, and the subsequent
611 forward lookup of the name timed out, the host name was left in
612 $sender_host_name, contrary to the specification.
614 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
615 restricted to single-key lookups, Exim was not diagnosing an error if
616 * or *@ was used with a query-style lookup.
618 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
620 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
621 long ago, but noone interested enough thought of it. Let's just say that
622 the "1.1" means that there are some new functions that weren't there at
623 some point in the past.
625 PH/51 Error processing for expansion failure of helo_data from an smtp
626 transport during callout processing was broken.
628 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
629 tested/used via the -bh/-bhc/-bs options.
631 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
632 bug, fixed in subsequent PCRE releases).
634 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
635 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
637 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
643 SC/01 Use a glob alias rather than an array ref in eximstats generated
644 parser. This improves both readability and performance.
646 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
647 Don't display local sender or destination tables in eximstats unless
648 there is data to show.
649 Added average volumes into the eximstats top table text output.
651 SC/03 Collect data on the number of addresses (recipients) as well
652 as the number of messages in eximstats.
654 TF/01 Correct an error in the documentation for the redirect router. Exim
655 does (usually) call initgroups() when daemonizing.
657 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
658 with consistent privilege compared to when running as a daemon.
660 TF/03 Note in the spec that $authenticated_id is not set for local
661 submissions from trusted users.
663 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
664 Thanks to Dean Brooks <dean@iglou.com> for the patch.
666 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
667 by adding some example configuration directives to the default
668 configuration file. A little bit of work is required to uncomment the
669 directives and define how usernames and passwords are checked, but
670 there is now a framework to start from.
672 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
673 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
674 without this. I don't know how relevant this is to other LDAP libraries.
676 PH/02 Add the verb name to the "unknown ACL verb" error.
678 PH/03 Magnus Holmgren's patch for filter_prepend_home.
680 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
682 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
683 directory not expanded when it should be if an expanded home directory
684 was set for the address (which is overridden by the transport).
686 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
689 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
690 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
691 because it is too late at that time, and has no effect.
693 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
694 security issue with \' (bugzilla #107). I could not use the
695 PQescapeStringConn() function, because it needs a PGconn value as one of
698 PH/08 When testing addresses using -bt, indicate those final addresses that
699 are duplicates that would not cause an additional delivery. At least one
700 person was confused, thinking that -bt output corresponded to deliveries.
701 (Suppressing duplicates isn't a good idea as you lose the information
702 about possibly different redirections that led to the duplicates.)
704 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
705 systems where poll() doesn't work, in particular OS X.
707 PH/10 Added more information to debugging output for retry time not reached.
709 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
710 operations in malware.c.
712 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
715 PH/13 If write_rejectlog was set false when logging was sent to syslog with
716 syslog_duplication set false, log lines that would normally be written
717 both the the main log and to the reject log were not written to syslog at
720 PH/14 In the default configuration, change the use of "message" in ACL warn
721 statements to "add_header".
723 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
724 not followed by a command (e.g. "seen endif").
726 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
727 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
730 PH/17 Added extra conditions to the default value of delay_warning_condition
734 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
735 { match{$h_precedence:}{(?i)bulk|list|junk} } \
736 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
739 The Auto-Submitted: and various List- headers are standardised, whereas I
740 don't think Precedence: ever was.
742 PH/18 Refactored debugging code in route_finduser() to show more information,
743 in particular, the error code if getpwnam() issues one.
745 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
746 This is apparently needed in addition to the PH/07 change above to avoid
747 any possible encoding problems.
749 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
750 but not after initializing Perl.
752 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
753 output them only if debugging. By default they are written stderr,
754 apparently, which is not desirable.
756 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
759 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
762 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
764 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
765 authenticated or an ident call has been made. Suppress the default
766 values for $authenticated_id and $authenticated_sender (but permit -oMai
767 and -oMas) when testing with -bh.
769 PH/24 Re-jigged the order of the tests in the default configuration so that the
770 tests for valid domains and recipients precede the DNS black list and CSA
771 tests, on the grounds that those ones are more expensive.
773 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
774 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
775 This bug exists in every version of Exim that I still have, right back to
778 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
779 However, an attempt to turn on case-sensitivity in a regex key by
780 including (?-i) didn't work because the subject string was already
781 lowercased, and the effects were non-intuitive. It turns out that a
782 one-line patch can be used to allow (?-i) to work as expected.
788 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
789 other effects) broke the use of negated acl sub-conditions.
791 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
794 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
795 "Deny" causes Exim to reject the incoming connection with a 554 error.
796 Unfortunately, if there is a major crisis, such as a disk failure,
797 tcp-wrappers gives "deny", whereas what one would like would be some
798 kind of temporary error. A kludge has been added to help with this.
799 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
800 554 error is used if errno is still zero or contains ENOENT (which occurs
801 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
804 PH/03 Add -lutil to the default FreeBSD LIBS setting.
806 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
807 errors. Otherwise a message that provokes a temporary error (when other
808 messages do not) can cause a whole host to time out.
810 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
811 addresses were routed directly to files or pipes from a redirect router.
812 File deliveries just didn't batch; pipe deliveries might have suffered
815 PH/06 A failure to get a lock for a hints database would erroneously always say
816 "Failed to get write lock", even when it was really a read lock.
818 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
819 of 0600. This has been changed to use the value of the lockfile_mode
820 option (which defaults to 0600).
822 PH/08 Applied small patch from the Sieve maintainer.
824 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
825 folder from quota calculations, a direct delivery into this folder messed
826 up the contents of the maildirsize file. This was because the regex was
827 used only to exclude .Trash (or whatever) when the size of the mailbox
828 was calculated. There was no check that a delivery was happening into an
829 excluded directory. This bug has been fixed by ignoring all quota
830 processing for deliveries into excluded directories.
832 PH/10 Added the maildirfolder_create_regex option to appendfile.
838 PH/01 The code for finding all the local interface addresses on a FreeBSD
839 system running IPv6 was broken. This may well have applied to all BSD
840 systems, as well as to others that have similar system calls. The broken
841 code found IPv4 interfaces correctly, but gave incorrect values for the
842 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
843 that it would not match correctly against @[] and not recognize the IPv6
846 PH/02 The ipliteral router was not recognizing addresses of the form user@
847 [ipv6:....] because it didn't know about the "ipv6:" prefix.
849 PH/03 Added disable_ipv6.
851 PH/04 Changed $reply_address to use the raw form of the headers instead of the
852 decoded form, because it is most often used to construct To: headers
853 lines in autoreplies, and the decoded form may well be syntactically
854 invalid. However, $reply_address has leading white space removed, and all
855 newlines turned into spaces so that the autoreply transport does not
858 PH/05 If group was specified without a user on a router, and no group or user
859 was specified on a transport, the group from the router was ignored.
861 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
862 for visible compile-time settings that can be used to change these
863 numbers, for those that want even more. Backwards compatibility with old
864 spool files has been maintained. However, going back to a previous Exim
865 release will lost any variables that are in spool files.
867 PH/07 Two small changes when running in the test harness: increase delay when
868 passing a TCP/IP connection to a new process, in case the original
869 process has to generate a bounce, and remove special handling of
870 127.0.0.2 (sic), which is no longer necessary.
872 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
873 be the same on different OS.
875 PH/09 Moved a debug statement in filter processing to avoid a race problem when
878 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
879 whether --show-vars was specified or not
881 JJ/02 exipick: Added support for new ACL variable spool format introduced
884 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
885 syntactically invalid From: or Reply-to: line, and a filter used this to
886 generate an autoreply, and therefore failed to obtain an address for the
887 autoreply, Exim could try to deliver to a non-existent relative file
888 name, causing unrelated and misleading errors. What now happens is that
889 it logs this as a hard delivery error, but does not attempt to create a
892 PH/11 The exinext utility has a -C option for testing purposes, but although
893 the given file was scanned by exinext itself; it wasn't being passed on
894 when Exim was called.
896 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
897 an end-of-file indication when reading a command response.
899 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
900 compiled. In many other places in Exim, IPv6 addresses are always
901 recognized, so I have changed this. It also means that IPv4 domain
902 literals of the form [IPV4:n.n.n.n] are now always recognized.
904 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
905 used if the router is not running as root, for example, when verifying at
906 ACL time, or when using -bh. The debugging output from this situation was
907 non-existent - all you got was a failure to exec. I have made two
910 (a) Failures to set uid/gid, the current directory, or a process leader
911 in a subprocess such as that created by queryprogram now generate
912 suitable debugging ouput when -d is set.
914 (b) The queryprogram router detects when it is not running as root,
915 outputs suitable debugging information if -d is set, and then runs
916 the subprocess without attempting to change uid/gid.
918 PH/15 Minor change to Makefile for building test_host (undocumented testing
921 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
922 additional section of a DNS packet that returns MX or SRV records.
923 Instead, it always explicitly searches for A/AAAA records. This avoids
924 major problems that occur when a DNS server includes only records of one
925 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
926 fixed another bug: if SRV records were looked up and the corresponding
927 address records were *not* found in the additional section, the port
928 values from the SRV records were lost.
930 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
931 using the correct key (the original address) when searching the retry
932 rules in order to find which one to use for generating the retry hint.
934 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
935 adding the default one. Similarly, if it contains a Reply-To: header, the
936 errors_reply_to option, if set, is not used.
938 PH/19 When calculating a retry time, Exim used to measure the "time since
939 failure" by looking at the "first failed" field in the retry record. Now
940 it does not use this if it is later than than the arrival time of the
941 message. Instead it uses the arrival time. This makes for better
942 behaviour in cases where some deliveries succeed, thus re-setting the
943 "first failed" field. An example is a quota failure for a huge message
944 when small messages continue to be delivered. Without this change, the
945 "time since failure" will always be short, possible causing more frequent
946 delivery attempts for the huge message than are intended.
947 [Note: This change was subsequently modified - see PH/04 for 4.62.]
949 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
950 $1, $2, $3) because the numerical variables can be reset during some
951 expansion items (e.g. "match"), thereby losing the authentication data.
953 PH/21 Make -bV show the size of off_t variables so that the test suite can
954 decide whether to run tests for quotas > 2G.
956 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
957 mailbox_size, and mailbox_filecount in the appendfile transport. If a
958 filecount value is greater than 2G or if a quota value is greater than 2G
959 on a system where the size of off_t is not greater than 4, a panic error
962 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
963 never match. The debug and -bh output now contains an explicit error
964 message indicating a malformed IPv4 address or mask.
966 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
967 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
970 PH/25 Do not write to syslog when running in the test harness. The only
971 occasion when this arises is a failure to open the main or panic logs
972 (for which there is an explicit test).
974 PH/26 Added the /no_tell option to "control=freeze".
976 PH/27 If a host name lookup failed very early in a connection, for example, if
977 the IP address matched host_lookup and the reverse lookup yielded a name
978 that did not have a forward lookup, an error message of the form "no IP
979 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
980 could be logged. Now it outputs the IP address instead of "NULL".
982 PH/28 An enabling patch from MH: add new function child_open_exim2() which
983 allows the sender and the authenticated sender to be set when
984 submitting a message from within Exim. Since child_open_exim() is
985 documented for local_scan(), the new function should be too.
987 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
988 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
989 results in an empty string is now treated as unset.
991 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
993 PH/31 Added sender_verify_fail logging option.
995 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
996 needed by RFC 822 but not by RFC 2822 was commented out. I have now
997 tidied the source and removed it altogether.
999 PH/33 When a queue run was abandoned because the load average was too high, a
1000 log line was always written; now it is written only if the queue_run log
1001 selector is set. In addition, the log line for abandonment now contains
1002 information about the queue run such as the pid. This is always present
1003 in "start" and "stop" lines but was omitted from the "abandon" line.
1005 PH/34 Omit spaces between a header name and the colon in the error message that
1006 is given when verify = headers_syntax fails (if there are lots of them,
1007 the message gets confusing).
1009 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1010 names, as there are now some PTR records that contain slashes. This check
1011 is only to protect against broken name servers that fall over on strange
1012 characters, so the fact that it applies to all lookups doesn't matter.
1014 PH/36 Now that the new test suite is complete, we can remove some of the
1015 special code in Exim that was needed for the old test suite. For example,
1016 sorting DNS records because real resolvers return them in an arbitrary
1017 order. The new test suite's fake resolver always returns records in the
1020 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1021 bounces) except when queue_only is set, to avoid logging races between
1022 the different processes.
1024 PH/38 Panic-die if .include specifies a non-absolute path.
1026 PH/39 A tweak to the "H" retry rule from its user.
1028 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1029 a label. They prevented compilation on older perls.
1031 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1032 a warning to be raised on newish perls.
1034 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1035 on queue. Changes to match documented behaviour of showing count of
1036 messages matching specified criteria.
1038 PH/40 Changed the default ident timeout from 30s to 5s.
1040 PH/41 Added support for the use of login_cap features, on those BSD systems
1041 that have them, for controlling the resources used by pipe deliveries.
1043 PH/42 The content-scanning code uses fopen() to create files in which to put
1044 message data. Previously it was not paying any attention to the mode of
1045 the files. Exim runs with umask(0) because the rest of the code creates
1046 files with open(), and sets the required mode explicitly. Thus, these
1047 files were ending up world-writeable. This was not a big issue, because,
1048 being within the spool directory, they were not world-accessible. I have
1049 created a function called modefopen, which takes an additional mode
1050 argument. It sets umask(777), creates the file, chmods it to the required
1051 mode, then resets the umask. All the relevant calls to fopen() in the
1052 content scanning code have been changed to use this function.
1054 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1055 to 24 hours. This avoids potential overflow problems when processing G
1056 and H retry rules. I suspect nobody ever tinkers with this value.
1058 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1060 PH/45 When the plaintext authenticator is running as a client, the server's
1061 challenges are checked to ensure they are valid base64 strings. By
1062 default, the authentication attempt is cancelled if an invalid string is
1063 received. Setting client_ignore_invalid_base64 true ignores these errors.
1064 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1065 they are received. Thus, the responses can be made to depend on the
1066 challenges. If an invalid string is ignored, an empty string is placed in
1069 PH/46 Messages that are created by the autoreply transport now contains a
1070 References: header, in accordance with RFCs 2822 and 3834.
1072 PH/47 Added authenticated_sender_force to the smtp transport.
1074 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1076 PH/49 Installed latest patch from the Sieve maintainer.
1078 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1079 was also set, the appendfile transport was unnecessarily scanning a
1080 directory of message files (e.g. for maildir delivery) to find the count
1081 of files (along with the size), even though it did not need this
1082 information. It now does the scan only if it needs to find either the
1083 size of the count of files.
1085 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1087 PH/52 Two bugs concerned with error handling when the smtp transport is
1090 (i) Exim was not creating retry information for temporary errors given
1091 for individual recipients after the DATA command when the smtp transport
1092 was used in LMTP mode. This meant that they could be retried too
1093 frequently, and not timed out correctly.
1095 (ii) Exim was setting the flag that allows error details to be returned
1096 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1097 recipients that were returned after the DATA command.
1099 PH/53 This is related to PH/52, but is more general: for any failing address,
1100 when detailed error information was permitted to be returned to the
1101 sender, but the error was temporary, then after the final timeout, only
1102 "retry timeout exceeded" was returned. Now it returns the full error as
1103 well as "retry timeout exceeded".
1105 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1106 do this, and (what is worse) MTAs that accept it.
1108 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1109 will now be deprecated.
1111 PH/56 New os.c-cygwin from the Cygwin maintainer.
1113 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1114 formats (previously only available in exim formats via -bpr, -bpru,
1115 and -bpra. Now also available in native and exiqgrep formats)
1117 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1118 with very large, slow to parse queues
1120 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1122 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1124 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1125 responses to authentication challenges, though it was showing the
1126 challenges; (ii) I've removed the CR characters from the debug output for
1129 PH/58 Allow for the insertion of a newline as well as a space when a string
1130 is turned into more than one encoded-word during RFC 2047 encoding. The
1131 Sieve code now uses this.
1133 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1134 data_4xx, lost_connection, tls_required.
1136 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1137 message was being sent as an SMTP response.
1139 PH/61 Add -l and -k options to exicyclog.
1141 PH/62 When verifying, if an address was redirected to one new address, so that
1142 verification continued, and the new address failed or deferred after
1143 having set something in $address_data, the value of $address_data was not
1144 passed back to the ACL. This was different to the case when no
1145 redirection occurred. The value is now passed back in both cases.
1147 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1148 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1149 use login_cap.h, so on its own it isn't the distinguishing feature. The
1150 new name refers directly to the setclassresources() function.
1152 PH/65 Added configuration files for NetBSD3.
1154 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1156 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1157 is preferred over IPv4.
1159 PH/68 The bounce_return_message and bounce_return_body options were not being
1160 honoured for bounces generated during the reception of non-SMTP messages.
1161 In particular, this applied to messages rejected by the ACL. This bug has
1162 been fixed. However, if bounce_return_message is true and bounce_return_
1163 body is false, the headers that are returned for a non-SMTP message
1164 include only those that have been read before the error was detected.
1165 (In the case of an ACL rejection, they have all been read.)
1167 PH/69 The HTML version of the specification is now built in a directory called
1168 spec_html instead of spec.html, because the latter looks like a path with
1169 a MIME-type, and this confuses some software.
1171 PH/70 Catch two compiler warnings in sieve.c.
1173 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1174 function verify_get_ident() calls ip_connect() to connect a socket, but
1175 if the "connect()" function timed out, ip_connect() used to close the
1176 socket. However, verify_get_ident() also closes the socket later, and in
1177 between Exim writes to the log, which may get opened at this point. When
1178 the socket was closed in ip_connect(), the log could get the same file
1179 descriptor number as the socket. This naturally causes chaos. The fix is
1180 not to close the socket in ip_connect(); the socket should be closed by
1181 the function that creates it. There was only one place in the code where
1182 this was missing, in the iplookup router, which I don't think anybody now
1183 uses, but I've fixed it anyway.
1185 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1186 well as to direct DNS lookups. Otherwise the handling of names in host
1187 lists is inconsistent and therefore confusing.
1193 PH/01 Two changes to the default runtime configuration:
1195 (1) Move the checks for relay_from_hosts and authenticated clients from
1196 after to before the (commented out) DNS black list checks.
1198 (2) Add control=submission to the relay_from_hosts and authenticated
1199 clients checks, on the grounds that messages accepted by these
1200 statements are most likely to be submissions.
1202 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1204 (1) Generate an error if the third argument for the ${prvs expansion is
1207 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1210 (3) Reset the variables that are obtained from the first argument of
1211 ${prvscheck and used in the second argument before leaving the code,
1212 because their memory is reclaimed, so using them afterwards may do
1215 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1216 one (it's much easier than Tom thought :-).
1218 (5) Because of (4), we can now allow for the use of $prvscheck_result
1219 inside the third argument.
1221 PH/03 For some reason, the default setting of PATH when running a command from
1222 a pipe transport was just "/usr/bin". I have changed it to
1225 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1226 anything to be listed in the output from -bV.
1228 PH/05 When a filter generated an autoreply, the entire To: header line was
1229 quoted in the delivery log line, like this:
1231 => >A.N.Other <ano@some.domain> <original@ddress> ...
1233 This has been changed so that it extracts the operative address. There
1234 may be more than one such address. If so, they are comma-separated, like
1237 => >ano@some.domain,ona@other.domain <original@ddress> ...
1239 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1240 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1241 not being looked up in the rDNS to get a host name, Exim was showing the
1242 IP address twice in Received: lines, even though the IP addresses were
1243 identical. For example:
1245 Received: from [1.2.3.4] (helo=[1.2.3.4])
1247 However, if the real host name was known, it was omitting the HELO data
1248 if it matched the actual IP address. This has been tidied up so that it
1249 doesn't show the same IP address twice.
1251 PH/07 When both +timestamp and +memory debugging was on, the value given by
1252 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1253 called by the debug printing, thereby overwriting the timestamp buffer.
1254 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1256 PH/08 When the original message was included in an autoreply transport, it
1257 always said "this is a copy of the message, including all the headers",
1258 even if body_only or headers_only was set. It now gives an appropriate
1261 PH/09 Applied a patch from the Sieve maintainer which:
1263 o fixes some comments
1264 o adds the (disabled) notify extension core
1265 o adds some debug output for the result of if/elsif tests
1266 o points to the current vacation draft in the documentation
1267 and documents the missing references header update
1271 o fixes a bug in processing the envelope test (when testing
1272 multiple envelope elements, the last element determinted the
1275 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1276 Electronic Mail") by including:
1278 Auto-submitted: auto-generated
1280 in the messages that it generates (bounce messages and others, such as
1281 warnings). In the case of bounce messages for non-SMTP mesages, there was
1282 also a typo: it was using "Auto_submitted" (underscore instead of
1283 hyphen). Since every message generated by Exim is necessarily in response
1284 to another message, thes have all been changed to:
1286 Auto-Submitted: auto-replied
1288 in accordance with these statements in the RFC:
1290 The auto-replied keyword:
1292 - SHOULD be used on messages sent in direct response to another
1293 message by an automatic process,
1295 - MUST NOT be used on manually-generated messages,
1297 - MAY be used on Delivery Status Notifications (DSNs) and Message
1298 Disposition Notifications (MDNs),
1300 - MUST NOT be used on messages generated by automatic or periodic
1301 processes, except for messages which are automatic responses to
1304 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1305 to the default Received: header definition.
1307 PH/12 Added log selector acl_warn_skipped (default on).
1309 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1310 variables because (a) they are in the wrong storage pool and (b) even if
1311 they were copied, it wouldn't work properly because of the caching.
1313 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1314 checking when decoding. Apparently there are clients that generate
1315 overlong encoded strings. Why am I not surprised?
1317 PH/15 If the first argument of "${if match_address" was not empty, but did not
1318 contain an "@" character, Exim crashed. Now it writes a panic log message
1319 and treats the condition as false.
1321 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1323 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1324 extension "envelope-auth". The code is finished and in agreement with
1325 other implementations, but there is no documentation so far and in fact,
1326 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1327 not changing the active code.
1329 Print executed "if" and "elsif" statements when debugging is used. This
1330 helps a great deal to understand what a filter does.
1332 Document more things not specified clearly in RFC3028. I had all this
1333 sorted out, when out of a sudden new issues came to my mind. Oops."
1335 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1338 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1339 canonical form (as documented). However, after a host name lookup from
1340 the IP address, check_host() was doing a simple string comparison with
1341 addresses acquired from the DNS when checking that the found name did
1342 have the original IP as one of its addresses. Since any found IPv6
1343 addresses are likely to be in abbreviated form, the comparison could
1344 fail. Luckily, there already exists a function for doing the comparison
1345 by converting both addresses to binary, so now that is used instead of
1346 the text comparison.
1348 PH/20 There was another similar case to PH/19, when a complete host name was
1349 given in a host list; looking up its IP address could give an abbreviated
1350 form, whereas the current host's name might or might not be abbreviated.
1351 The same fix has been applied.
1357 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1358 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1361 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1362 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1364 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1366 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1367 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1368 Also, for "Auto-Submitted", treat anything other than "no" as
1369 non-personal, in accordance with RFC 3834. (Previously it treated
1370 anything starting "auto-" as non-personal.)
1372 TF/01 The control=submission/name=... option had a problem with syntax
1373 errors if the name included a slash character. The /name= option
1374 now slurps the rest of the string, so it can include any characters
1375 but it must come last in the list of options (after /sender_retain
1378 PH/05 Some modifications to the interface to the fake nameserver for the new
1386 TK/01 Added the "success_on_redirect" address verification option. See
1387 NewStuff for rationale and an example.
1389 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1391 PH/02 Patch to exigrep to allow it to work on syslog lines.
1393 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1394 fread() to skip over the body file's header line, because in Cygwin the
1395 header line is locked and is inaccessible.
1397 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1398 co-exist for some time) to make it clear that it is the Exim ID that is
1399 referenced, not the Message-ID: header line.
1401 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1402 string_format() function, because snprintf() does not exist on all
1405 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1406 ${stat: expansion item.
1408 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1409 protocol synchronization error", to keep the pedants happy.
1411 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1412 well as for IRIX systems, when gcc is being used. See the host.c source
1415 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1417 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1420 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1421 transport and to the smtp transport in LMTP mode.
1423 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1425 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1426 run a filter in a subprocess. This could lead to confusion in subsequent
1427 lookups in the parent process. There should also be a search_tidyup() at
1428 the end of the subprocess.
1430 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1431 only if the host matched helo_try_verify_hosts, which caused the
1432 verification to occur when the EHLO/HELO command was issued. The ACL just
1433 tested the remembered result. Now, if a previous verification attempt has
1434 not happened, "verify = helo" does it there and then.
1436 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1438 TK/03 Fix log output including CR from clamd.
1440 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1441 exist provoked a memory error which could cause a segfault.
1443 PH/15 Installed PCRE 6.2
1445 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1447 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1448 of the problem. Specifically, suggested +O2 rather than +O1 for the
1451 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1453 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1454 record, the port was not correctly logged when the outgoing_port log
1455 selector was set (it logged the transort's default port).
1457 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1458 fallback_hosts, and "hosts" in the smtp transport.
1460 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1461 host errors such as "Connection refused".
1463 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1464 authentication with radiusclient 0.4.9:
1466 - Error returned from rc_read_config was caught wrongly
1467 - Username/password not passed on to radius server due to wrong length.
1469 The presumption is that some radiusclient API changes for 4.51/PH/17
1470 were not taken care of correctly. The code is still untested by me (my
1471 Linux distribution still has 0.3.2 of radiusclient), but it was
1472 contributed by a Radius user.
1474 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1475 expanding the "port" option of the smtp transport.
1477 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1478 while reading a MIME header. Thanks to Tom Hughes for a patch.
1480 PH/24 Include config.h inside local_scan.h so that configuration settings are
1483 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1484 that in an ACL for RCPT (for example), you can examine exactly what was
1487 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1488 commands, but it was not correctly comparing the address with the actual
1489 client host address. Thus, it would show the EHLO address in Received:
1490 header lines when this was not necessary.
1492 PH/27 Added the % operator to ${eval:}.
1494 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1495 it should be ignoring failures (because with -C, for example, it has lost
1496 privilege). It wasn't ignoring creation failures other than "already
1499 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1502 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1503 because an input line was too long, either on its own, or by virtue of
1504 too many continuations, the temporary file was not being removed, and the
1505 return code was incorrect.
1507 PH/31 Missing "BOOL" in function definition in filtertest.c.
1509 PH/32 Applied Sieve patches from the maintainer.
1511 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1513 PH/33 Added "verify = not_blind".
1515 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1516 Local/Makefile (with some defaults set). These are used in built scripts
1517 such as exicyclog, but they have never been used in the exim_install
1518 script (though there are many overriding facilities there). I have
1519 arranged that the exim_install script now takes note of these two
1522 PH/35 Installed configuration files for Dragonfly.
1524 PH/36 When a locally submitted message by a trusted user did not contain a
1525 From: header, and the sender address was obtained from -f or from an SMTP
1526 MAIL command, and the trusted user did not use -F to supply a sender
1527 name, $originator_name was incorrectly used when constructing a From:
1528 header. Furthermore, $originator_name was used for submission mode
1529 messages from external hosts without From: headers in a similar way,
1530 which is clearly wrong.
1532 PH/37 Added control=suppress_local_fixups.
1534 PH/38 When log_selector = +received_sender was set, and the addition of the
1535 sender made the log line's construction buffer exactly full, or one byte
1536 less than full, an overflow happened when the terminating "\n" was
1539 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1540 when the result of a list match is failure because a DNS lookup failed.
1542 PH/40 RM_COMMAND is now used in the building process.
1544 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1545 the "build-* directories that it finds.
1547 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1548 address in a domain literal was a prefix of an interface address.
1550 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1551 when verifying a sender address, unless rewrite_headers is false.
1553 PH/44 Wrote a long comment about why errors_to addresses are verified as
1554 recipients, not senders.
1556 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1557 the ratelimit ACL was added.
1559 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1561 PH/46 Added extra information about PostgreSQL errors to the error string.
1563 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1564 avoiding the need to install special zones in a real server. This is
1565 backwards compatible; if it can't find the fake resolver, it drops back.
1566 Thus, both old and new test suites can be run.
1568 TF/02 Added util/ratelimit.pl
1570 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1571 clock is set back in time.
1573 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1574 Candler <B.Candler@pobox.com>.
1576 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1577 OK for addresses that are the result of redirections.
1579 PH/48 A number of further additions for the benefit of the new test suite,
1580 including a fake gethostbyname() that interfaces to the fake DNS resolver
1583 TF/06 The fix for widen_domains has also been applied to qualify_single and
1584 search_parents which are the other dnslookup options that can cause
1587 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1590 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1592 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1593 service" if it encountered an MX record with an empty target hostname.
1594 The message is now "an MX or SRV record indicated no SMTP service".
1596 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1597 if the DNS of the sending site is misconfigured. This is quite a
1598 common situation. This change restores the behaviour of treating a
1599 helo verification defer as a failure.
1601 PH/51 If self=fail was set on a router, the bounce message did not include the
1602 actual error message.
1608 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1610 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1611 command itself ended in error, the underlying message about the transport
1612 filter timeout was being overwritten with the pipe command error. Now the
1613 underlying error message should be appended to the second error message.
1615 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1617 PH/02 Reduce the amount of output that "make" produces by default. Full output
1618 can still be requested.
1620 PH/03 The warning log line about a condition test deferring for a "warn" verb
1621 was being output only once per connection, rather than after each
1622 occurrence (because it was using the same function as for successful
1623 "warn" verbs). This seems wrong, so I have changed it.
1625 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1626 it should not have, which might have caused a crash in the right
1627 circumstances, but probably never did.
1629 PH/04 Installed a modified version of Tony Finch's patch to make submission
1630 mode fix the return path as well as the Sender: header line, and to
1631 add a /name= option so that you can make the user's friendly name appear
1634 TF/03 Added the control = fakedefer ACL modifier.
1636 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1637 Mark Lowes for thorough testing.
1639 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1641 TK/03 Merged latest SRS patch from Miles Wilton.
1643 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1644 with the definition in sysexits.h (which is #included earlier).
1645 Fortunately, Exim does not actually use EX_OK. The code used to try to
1646 preserve the sysexits.h value, by assumimg that macro definitions were
1647 scanned for macro replacements. I have been disabused of this notion,
1648 so now the code just undefines EX_OK before #including unistd.h.
1650 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1651 in the smtp transport. When a block could not be written in a single
1652 write() function, the timeout was being re-applied to each part-write.
1653 This seems wrong - if the receiver was accepting one byte at a time it
1654 would take for ever. The timeout is now adjusted when this happens. It
1655 doesn't have to be particularly precise.
1657 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1658 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1660 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1661 without a domain if the check to <postmaster@domain> fails.
1663 SC/01 Eximstats: added -xls and the ability to specify output files
1664 (patch written by Frank Heydlauf).
1666 SC/02 Eximstats: use FileHandles for outputing results.
1668 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1670 SC/04 Eximstats: fixed display of large numbers with -nvr option
1672 SC/05 Eximstats: fixed merging of reports with empty tables.
1674 SC/06 Eximstats: added the -include_original_destination flag
1676 SC/07 Eximstats: removed tabs and trailing whitespace.
1678 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1680 TK/06 MBOX spool code: Add real "From " MBOX separator line
1681 so the .eml file is really in mbox format (even though
1682 most programs do not really care). Patch from Alex Miller.
1684 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1685 The latter is generated from $received_to and is only set if the
1686 message has one envelope recipient. SA can use these headers,
1687 obviously out-of-the-box. Patch from Alex Miller.
1689 PH/08 The ${def test on a variable was returning false if the variable's
1690 value was "0", contrary to what the specification has always said!
1691 The result should be true unless the variable is empty.
1693 PH/09 The syntax error of a character other than { following "${if
1694 def:variable_name" (after optional whitespace) was not being diagnosed.
1695 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1696 accidental colon was present, for example, could give incorrect results.
1698 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1699 result is used (not including appendfile, where other changes are about
1702 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1703 This involved changing a lot of size variables from int to off_t. It
1704 should work with maildirs and everything.
1706 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1707 spamd dying while we are connected to it.
1709 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1712 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1713 function for BDB 4.3.
1715 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1717 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1718 the sources for PCRE have been re-arranged and more files are now
1721 PH/15 The code I had for printing potentially long long variables in PH/11
1722 above was not the best (it lost precision). The length of off_t variables
1723 is now inspected at build time, and an appropriate printing format (%ld
1724 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1725 to be "long long int" or "long int". This is needed for the internal
1726 formatting function string_vformat().
1728 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1729 the configuration file to be ":syslog", then the script "guesses" where
1730 the logs files are, rather than using the compiled in default. In our
1731 case the guess is not the same as the compiled default, so the script
1732 suddenly stopped working when I started to use syslog. The patch checks
1733 to see if log_file_path is "". If so, it attempts to read it from exim
1734 with no configuration file to get the compiled in version, before it
1735 falls back to the previous guessing code."
1737 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1738 implementing BATV in an Exim configuration. See NewStuff for the gory
1741 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1742 Makefile that are specific to HP-UX.
1744 PH/18 If the "use_postmaster" option was set for a recipient callout together
1745 with the "random" option, the postmaster address was used as the MAIL
1746 FROM address for the random test, but not for the subsequent recipient
1747 test. It is now used for both.
1749 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1750 patch removes a few documentation additions to RFC 3028, because the
1751 latest draft now contains them. It adds the new en;ascii-case comparator
1752 and a new error check for 8bit text in MIME parts. Comparator and
1753 require names are now matched exactly. I enabled the subaddress
1754 extension, but it is not well tested yet (read: it works for me)."
1756 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1757 rework some of the code of TK/09 above to avoid the hardwired use of
1758 "%lld" and "long long". Replaced the call to snprintf() with a call to
1761 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1762 records point to non-existent hosts", "retry timeout exceeded", and
1763 "retry time not reached for any host after a long failure period".
1765 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1766 experimental DomainKeys support:
1768 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1769 (2) On an error such as an illegally used "control", the wrong name for
1770 the control was given.
1772 These problems did NOT occur unless DomainKeys support was compiled.
1774 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1776 PH/24 Added ${if match_ip condition.
1778 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1779 databases so that it will be absolutely obvious if a crash occurs in the
1780 DB library. This is a regular occurrence (often caused by mis-matched
1783 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1784 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1785 picked up on a user's system that detects such things. There doesn't seem
1786 to be a gcc warning option for this - only an attribute that has to be
1787 put on the function's prototype. It seems that in Fedora Core 4 they have
1788 set this on a number of new functions. No doubt there will be more in due
1791 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1792 specify a transport. However, if an address that was verified by such a
1793 router was the subject of a callout, Exim crashed because it tried to
1794 read the rcpt_include_affixes from the non-existent transport. Now it
1795 just assumes that the setting of that option is false. This bug was
1796 introduced by 4.51/PH/31.
1798 PH/28 Changed -d+all to exclude +memory, because that information is very
1799 rarely of interest, but it makes the output a lot bigger. People tend to
1800 do -d+all out of habit.
1802 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1803 code in os-type was giving problems when libc.so lives in lib64, like on
1806 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1807 aren't the modern standard, and it seems that some systems' include files
1808 don't always have them. Exim was already checking for some of the newer
1809 ones like T_AAAA, and defining it itself. I've added checks for all the
1810 record types that Exim uses.
1812 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1813 not automatically generating a new one, as it is supposed to. This
1814 prevented TLS from working. If the file did exist, but contained invalid
1815 data, a new version was generated, as expected. It was only the case of a
1816 non-existent file that was broken.
1818 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1819 with a change in libdomainkeys > 0.64.
1821 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1822 from DNS. If the selector record carries the flag, it now has
1823 precedence over the domain-wide flag.
1825 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1827 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1828 the use of a port name that isn't defined in /etc/services) occurred, the
1829 message was deferred as in a normal delivery, and thus remained on the
1830 spool, instead of being failed because of the mua_wrapper setting. This
1831 is now fixed, and I tidied up some of the mua_wrapper messages at the
1834 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1835 the messages in a hash of arrays rather than using individual hashes.
1836 This is a bit cleaner and results in dramatic memory savings, albeit
1837 at a slight CPU cost.
1839 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1840 as requested by Marc Sherman.
1842 SC/10 Eximstats: added histograms for user specified patterns as requested
1845 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1847 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1848 fopen() in the content-scanning modules that did not already have it.
1854 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1855 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1857 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1859 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1861 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1862 to test Sieve filters that use "vacation".
1864 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1865 that changes the way the GnuTLS parameters are stored in the cache file.
1866 The new format can be generated externally. For backward compatibility,
1867 if the data in the cache doesn't make sense, Exim assumes it has read an
1868 old-format file, and it generates new data and writes a new file. This
1869 means that you can't go back to an older release without removing the
1872 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1873 work if there are any delivery delays because "one_time" forces the
1874 parent to be marked "delivered", so its unseen clone is never tried
1875 again. For this reason, Exim now forbids the simultaneous setting of
1878 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1879 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1880 entry if you want to know the details. The fix, however, neglected to
1881 consider the case when local delivery batching is involved. The test for
1882 "previously delivered" was not happening when checking to see if an
1883 address could be batched with a previous (undelivered) one; under
1884 certain circumstances this could lead to multiple deliveries to the same
1887 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1888 in its include files, and this causes problems building Exim.
1890 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1891 header_syntax) but Exim was just ignoring anything given after a slash.
1892 In particular, this caused confusion with an attempt to use "verify =
1893 reverse_host_lookup/defer_ok". An error is now given when options are
1894 supplied for verify items that do not have them. (Maybe reverse_host_
1895 lookup should have a defer_ok option, but that's a different point.)
1897 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1898 defined by RFC 821) to 2048, because there were problems with some AUTH
1899 commands, and RFC 1869 says the size should be increased for extended
1900 SMTP commands that take arguments.
1902 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1905 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1906 "unknown" error; now it says that the functionality isn't in the binary.
1908 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1909 an address' error message when a string expansion fails (syntax or
1910 whatever). Otherwise the password may appear in the log. Following change
1911 PH/42 below, there is no longer a chance of it appearing in a bounce
1914 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1916 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1917 of its addresses was ever tried. (Bugzilla bug #2.)
1919 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1920 the result incorrectly in the debug output. (It correctly added a newline
1921 to what was transported.)
1923 TF/01 Added $received_time.
1925 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1926 commented out examples of how to interface to a virus scanner and to
1927 SpamAssassin. Also added commented examples of av_scanner and
1928 spamd_address settings.
1930 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1931 and controls are allowed in which ACLs. There were a couple of minor
1932 errors. Some of the entries in the conditions table (which is a table of
1933 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1934 negation of where the condition IS allowed.
1936 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1938 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1939 header file does not have a version number, so I've had to invent a new
1940 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1941 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1942 radiusclient), but it was contributed by a Radius user.
1944 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1945 files or directories, for OpenSSL.
1947 PH/19 When an Exim process that is running as root has to create an Exim log
1948 file, it does so in a subprocess that runs as exim:exim so as to get the
1949 ownership right at creation (otherwise, other Exim processes might see
1950 the file with the wrong ownership). There was no test for failure of this
1951 fork() call, which would lead to the process getting stuck as it waited
1952 for a non-existent subprocess. Forks do occasionally fail when resources
1953 run out. I reviewed all the other calls to fork(); they all seem to check
1956 PH/20 When checking for unexpected SMTP input at connect time (before writing
1957 the banner), Exim was not dealing correctly with a non-positive return
1958 from the read() function. If the client had disconnected by this time,
1959 the result was a log entry for a synchronization error with an empty
1960 string after "input=" when read() returned zero. If read() returned -1
1961 (an event I could not check), uninitialized data bytes were printed.
1962 There were reports of junk text (parts of files, etc) appearing after
1965 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1967 PH/22 Added support for macro redefinition, and (re)definition in between
1968 driver and ACL definitions.
1970 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1971 forgetting to use the resulting value; it was using the unexpanded value.
1973 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1974 hadn't been configured. The fix is from Juergen Kreileder, who
1975 understands it better than I do:
1977 "Here's what I see happening with three configured cyrus_sasl
1978 authenticators configured (plain, login, cram-md5):
1980 On startup auth_cyrus_sasl_init() gets called for each of these.
1981 This means three calls to sasl_listmech() without a specified mech_list.
1982 => SASL tests which mechs of all available mechs actually work
1983 => three warnings about OTP not working
1984 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1986 With the patch, sasl_listmech() also gets called three times. But now
1987 SASL's mech_list option is set to the server_mech specified in the the
1988 authenticator. Or in other words, the answer from sasl_listmech()
1989 gets limited to just the mech you're testing for (which is different
1991 => the return list contains just 'plain' or 'login', 'cram-md5' or
1992 nothing depending on the value of ob->server_mech.
1994 I've just tested the patch: Authentication still works fine,
1995 unavailable mechs specified in the exim configuration are still
1996 caught, and the auth.log warnings about OTP are gone."
1998 PH/25 When debugging is enabled, the contents of the command line are added
1999 to the debugging output, even when log_selector=+arguments is not
2002 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2003 answer is "GNU", and only if the return is "GNU/something" is the answer
2006 PH/27 $acl_verify_message is now set immediately after the failure of a
2007 verification in an ACL, and so is available in subsequent modifiers. In
2008 particular, the message can be preserved by coding like this:
2010 warn !verify = sender
2011 set acl_m0 = $acl_verify_message
2013 Previously, $acl_verify_message was set only while expanding "message"
2014 and "log_message" when a very denied access.
2016 PH/28 Modified OS/os.c-Linux with
2018 -#ifndef OS_LOAD_AVERAGE
2019 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2021 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2022 nomenclature these days.)
2024 PH/29 Installed patch from the Sieve maintainer that adds the options
2025 sieve_useraddress and sieve_subaddress to the redirect router.
2027 PH/30 In these circumstances:
2028 . Two addresses routed to the same list of hosts;
2029 . First host does not offer TLS;
2030 . First host accepts first address;
2031 . First host gives temporary error to second address;
2032 . Second host offers TLS and a TLS session is established;
2033 . Second host accepts second address.
2034 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2035 and peerdn, if requested) that were in fact used only for the second
2038 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2039 attention to any local part prefix or suffix that was matched by the
2040 router that accepted the address. It now behaves in the same way as it
2041 does for delivery: the affixes are removed from the local part unless
2042 rcpt_include_affixes is set on the transport.
2044 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2045 timeout during the DATA phase of an incoming message.
2047 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2048 applied a patch sanctioned by the Sieve maintainer.
2050 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2051 the uid or gid is negative. A case of a negative gid caused this to be
2052 noticed. The fix allows for either to be negative.
2054 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2055 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2058 PH/36 The change PH/12 above was broken. Fixed it.
2060 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2061 the grounds that routing the same address twice would always produce the
2062 same answer. This might have been true once, but it is certainly no
2063 longer true now. Routing a child address may depend on the previous
2064 routing that produced that child. Some complicated redirection strategies
2065 went wrong when messages had multiple recipients, and made Exim's
2066 behaviour dependent on the order in which the addresses were given.
2068 I have moved the duplicate checking until after the routing is complete.
2069 Exim scans the addresses that are assigned to local and remote
2070 transports, and removes any duplicates. This means that more work will be
2071 done, as duplicates will always all be routed, but duplicates are
2072 presumably rare, so I don't expect this is of any significance.
2074 For deliveries to pipes, files, and autoreplies, the duplicate checking
2075 still happens during the routing process, since they are not going to be
2078 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2079 It corrects a timeout issue with spamd. This is Ian's comment: "The
2080 background is that sometimes spamd either never reads data from a
2081 connection it has accepted, or it never writes response data. The exiscan
2082 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2083 blindly assumes that writes won't block so it may never time out."
2085 PH/39 Allow G after quota size as well as K and M.
2087 PH/40 The value set for $authenticated_id in an authenticator may not contain
2088 binary zeroes or newlines because the value is written to log lines and
2089 to spool files. There was no check on this. Now the value is run through
2090 the string_printing() function so that such characters are converted to
2091 printable escape sequences.
2093 PH/41 $message_linecount is a new variable that contains the total number of
2094 lines in the message. Compare $body_linecount, which is the count for the
2097 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2098 bounce and delay warning messages, except in certain special cases, which
2101 (a) An SMTP error message from a remote host;
2102 (b) A message specified in a :fail: redirection;
2103 (c) A message specified in a "fail" command in a system filter;
2104 (d) A message specified in a FAIL return from the queryprogram router;
2105 (e) A message specified by the cannot_route_message router option.
2107 In these cases only, Exim does include the error details in bounce and
2108 warning messages. There are also a few cases where bland messages such
2109 as "unrouteable address" or "local delivery error" are given.
2111 PH/43 $value is now also set for the "else" part of a ${run expansion.
2113 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2114 being worked on, but at least Exim now implements the latest version to
2117 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2118 process to complete was treated as a delivery failure, a timeout while
2119 writing the message to the pipe was logged, but erroneously treated as a
2120 successful delivery. Such timeouts include transport filter timeouts. For
2121 consistency with the overall process timeout, these timeouts are now
2122 treated as errors, giving rise to delivery failures by default. However,
2123 there is now a new Boolean option for the pipe transport called
2124 timeout_defer, which, if set TRUE, converts the failures into defers for
2125 both kinds of timeout. A transport filter timeout is now identified in
2128 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2129 systems where "make" and "gmake" are different, calling "gmake" at top
2130 level broke things. I've arranged for the value of $(MAKE) to be passed
2131 from the Makefile to this script so that it can call the same version of
2135 A note about Exim versions 4.44 and 4.50
2136 ----------------------------------------
2138 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2139 changes of various kinds. As a consequence, a big documentation update was
2140 needed. This delayed the release for rather longer than seemed good, especially
2141 in the light of a couple of (minor) security issues. Therefore, the changes
2142 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2143 release. So 4.44 and 4.50 are in effect two different branches that both start
2146 I have left the 4.50 change log unchanged; it contains all the changes since
2147 4.43. The change log for 4.44 is below; many of its items are identical to
2148 those for 4.50. This seems to be the most sensible way to preserve the
2149 historical information.
2155 1. Minor wording change to the doc/README.SIEVE file.
2157 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2158 computation of the current number of files was incorrect.
2160 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2161 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2164 4. Give more explanation in the error message when the command for a transport
2165 filter fails to execute.
2167 5. There are several places where Exim runs a non-Exim command in a
2168 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2169 was being done only for the command run by the queryprogram router. It is
2170 now done for all such subprocesses. The other cases are: ${run, transport
2171 filters, and the commands run by the lmtp and pipe transports.
2173 6. Added CONFIGURE_GROUP build-time option.
2175 7. Some older OS have a limit of 256 on the maximum number of file
2176 descriptors. Exim was using setrlimit() to set 1000 as a large value
2177 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2178 systems. I've change it so that if it can't get 1000, it tries for 256.
2180 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2181 was an oversight, and furthermore, ever since the addition of extra
2182 controls (e.g. 4.43/32), the checks on when to allow different forms of
2183 "control" were broken. There should now be diagnostics for all cases when a
2184 control that does not make sense is encountered.
2186 9. Added the /retain_sender option to "control=submission".
2188 10. $recipients is now available in the predata ACL (oversight).
2190 11. Tidy the search cache before the fork to do a delivery from a message
2191 received from the command line. Otherwise the child will trigger a lookup
2192 failure and thereby defer the delivery if it tries to use (for example) a
2193 cached ldap connection that the parent has called unbind on.
2195 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2196 of $address_data from the recipient verification was clobbered by the
2197 sender verification.
2199 13. The value of address_data from a sender verification is now available in
2200 $sender_address_data in subsequent conditions in the ACL statement.
2202 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2204 15. Added a new option "connect=<time>" to callout options, to set a different
2207 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2208 was its contents. (It was OK if the option was not defined at all.)
2210 17. A "Completed" log line is now written for messages that are removed from
2211 the spool by the -Mrm option.
2213 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2214 information about exactly what failed.
2216 19. Added -dd to debug only the daemon process.
2218 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2219 handles timeouts, both on the server side and network timeouts. Renamed the
2220 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2222 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2223 It is now set to "smtps".
2225 22. $host_address is now set to the target address during the checking of
2226 ignore_target_hosts.
2228 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2229 being passed; this would have caused $sender_host_name to have been used if
2230 matching the list had actually called for a host name (not very likely,
2231 since this list is usually IP addresses). A host name is now passed as
2234 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2235 code that specifies a non-restarting handler (typically sigaction() in
2236 modern systems) in an attempt to fix a rare and obscure crash bug.
2238 25. Narrowed the window for a race in the daemon that could cause it to ignore
2239 SIGCHLD signals. This is not a major problem, because they are used only to
2240 wake it up if nothing else does.
2242 26. A malformed maildirsize file could cause Exim to calculate negative values
2243 for the mailbox size or file count. Odd effects could occur as a result.
2244 The maildirsize information is now recalculated if the size or filecount
2247 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2248 support for a long time. Removed HAVE_SYS_VFS_H.
2250 28. Installed the latest version of exipick from John Jetmore.
2252 29. In an address list, if the pattern was not a regular expression, an empty
2253 subject address (from a bounce message) matched only if the pattern was an
2254 empty string. Non-empty patterns were not even tested. This was the wrong
2255 because it is perfectly reasonable to use an empty address as part of a
2256 database query. An empty address is now tested by patterns that are
2257 lookups. However, all the other forms of pattern expect the subject to
2258 contain a local part and a domain, and therefore, for them, an empty
2259 address still always fails if the pattern is not itself empty.
2261 30. Exim went into a mad DNS loop when attempting to do a callout where the
2262 host was specified on an smtp transport, and looking it up yielded more
2263 than one IP address.
2265 31. Re-factored the code for checking spool and log partition space into a
2266 function that finds that data and another that does the check. The former
2267 is then used to implement four new variables: $spool_space, $log_space,
2268 $spool_inodes, and $log_inodes.
2270 32. The RFC2047 encoding function was originally intended for short strings
2271 such as real names; it was not keeping to the 75-character limit for
2272 encoded words that the RFC imposes. It now respects the limit, and
2273 generates multiple encoded words if necessary. To be on the safe side, I
2274 have increased the buffer size for the ${rfc2047: expansion operator from
2277 33. It is now permitted to omit both strings after an "if" condition; if the
2278 condition is true, the result is "true". As before, when the second string
2279 is omitted, a false condition yields an empty string. This makes it less
2280 cumbersome to write custom ACL and router conditions.
2282 34. Failure to deliver a bounce message always caused it to be frozen, even if
2283 there was an errors_to setting on the router. The errors_to setting is now
2286 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2287 canonical form (fully expanded) before being placed in
2288 $sender_host_address.
2290 36. The table in the code that translates DNS record types into text (T_A to
2291 "A" for instance) was missing entries for NS and CNAME. It is just possible
2292 that this could have caused confusion if both these types were looked up
2293 for the same domain, because the text type is used as part of Exim's
2294 per-process caching. But the chance of anyone hitting this buglet seems
2297 37. The dnsdb lookup has been extended in a number of ways.
2299 (1) There is a new type, "zns", which walks up the domain tree until it
2300 finds some nameserver records. It should be used with care.
2302 (2) There is a new type, "mxh", which is like "mx" except that it returns
2303 just the host names, not the priorities.
2305 (3) It is now possible to give a list of domains (or IP addresses) to be
2306 looked up. The behaviour when one of the lookups defers can be
2307 controlled by a keyword.
2309 (4) It is now possible to specify the separator character for use when
2310 multiple records are returned.
2312 38. The dnslists ACL condition has been extended: it is now possible to supply
2313 a list of IP addresses and/or domains to be looked up in a particular DNS
2316 39. Added log_selector=+queue_time_overall.
2318 40. When running the queue in the test harness, wait just a tad after forking a
2319 delivery process, to get repeatability of debugging output.
2321 41. Include certificate and key file names in error message when GnuTLS fails
2322 to set them up, because the GnuTLS error message doesn't include the name
2323 of the failing file when there is a problem reading it.
2325 42. Allow both -bf and -bF in the same test run.
2327 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2329 44. The "Exiscan patch" is now merged into the mainline Exim source.
2331 45. Sometimes the final signoff response after QUIT could fail to get
2332 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2333 < 0 before doing a fflush(). This bug looks as though it goes back to the
2334 introduction of TLS in release 3.20, but "sometimes" must have been rare
2335 because the tests only now provoked it.
2337 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2338 (this can affect the format of dates).
2340 47. exim_tidydb, when checking for the continued existence of a message for
2341 which it has found a message-specific retry record, was not finding
2342 messages that were in split spool directories. Consequently, it was
2343 deleting retry records that should have stayed in existence.
2345 48. Steve fixed some bugs in eximstats.
2347 49. The SPA authentication driver was not abandoning authentication and moving
2348 on to the next authenticator when an expansion was forced to fail,
2349 contradicting the general specification for all authenticators. Instead it
2350 was generating a temporary error. It now behaves as specified.
2352 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2353 (the order specifies the preference for clients). The order is now AES256,
2354 AES128, 3DES, ARCFOUR128.
2356 51. Small patch to Sieve code - explicitly set From: when generating an
2359 52. Exim crashed if a remote delivery caused a very long error message to be
2360 recorded - for instance if somebody sent an entire SpamAssassin report back
2361 as a large number of 550 error lines. This bug was coincidentally fixed by
2362 increasing the size of one of Exim's internal buffers (big_buffer) that
2363 happened as part of the Exiscan merge. However, to be on the safe side, I
2364 have made the code more robust (and fixed the comments that describe what
2367 53. Now that there can be additional text after "Completed" in log lines (if
2368 the queue_time_overall log selector is set), a one-byte patch to exigrep
2369 was needed to allow it to recognize "Completed" as not the last thing in
2372 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2373 patch that reportedly fixes this has been added. I am not expert enough to
2374 create a test for it. This is what the patch creator wrote:
2376 "I found a little strange behaviour of ldap code when working with
2377 Windows 2003 AD Domain, where users was placed in more than one
2378 Organization Units. When I tried to give exim partial DN, the exit code
2379 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2380 But simultaneously result of request was absolutely normal ldap result,
2381 so I produce this patch..."
2383 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2384 REFERENCE, so I have modified the code to exclude the patch when that macro
2387 55. Some experimental protocols are using DNS PTR records for new purposes. The
2388 keys for these records are domain names, not reversed IP addresses. The
2389 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2390 leaves it alone. Component reversal etc. now happens only for IP addresses.
2393 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2395 57. Double the size of the debug message buffer (to 2048) so that more of very
2396 long debug lines gets shown.
2398 58. The exicyclog utility now does better if the number of log files to keep
2399 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2401 59. Two changes related to the smtp_active_hostname option:
2403 (1) $smtp_active_hostname is now available as a variable.
2404 (2) The default for smtp_banner uses $smtp_active_hostname instead
2405 of $primary_hostname.
2407 60. The host_aton() function is supposed to be passed a string that is known
2408 to be a valid IP address. However, in the case of IPv6 addresses, it was
2409 not checking this. This is a hostage to fortune. Exim now panics and dies
2410 if the condition is not met. A case was found where this could be provoked
2411 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2412 components; fortuitously, this particular loophole had already been fixed
2413 by change 4.50/55 above.
2415 If there are any other similar loopholes, the new check in host_aton()
2416 itself should stop them being exploited. The report I received stated that
2417 data on the command line could provoke the exploit when Exim was running as
2418 exim, but did not say which command line option was involved. All I could
2419 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2420 running as the user.
2423 61. There was a buffer overflow vulnerability in the SPA authentication code
2424 (which came originally from the Samba project). I have added a test to the
2425 spa_base64_to_bits() function which I hope fixes it.
2428 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2429 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2431 63. The daemon start-up calls getloadavg() while still root for those OS that
2432 need the first call to be done as root, but it missed one case: when
2433 deliver_queue_load_max is set with deliver_drop_privilege. This is
2434 necessary for the benefit of the queue runner, because there is no re-exec
2435 when deliver_drop_privilege is set.
2437 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2438 This has been fixed.
2440 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2441 was in use, was not putting the data itself into the right store pool;
2442 consequently, it could be overwritten for a subsequent message in the same
2443 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2446 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2448 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2449 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2450 the function were treating the return as a boolean value, which happened to
2451 work because 0=false and not-0=true, but is not correct code.
2453 68. The host_aton() function was not handling scoped IPv6 addresses (those
2454 with, for example, "%eth0" on the end) correctly.
2456 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2457 negated items (that is, ~something) in unsigned ints. Some compilers
2458 apparently mutter when there is no cast.
2460 70. If an address verification called from an ACL failed, and did not produce a
2461 user-specific message (i.e. there was only a "system" message), nothing was
2462 put in $acl_verify_message. In this situation, it now puts the system
2465 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2466 session; change 4.31/43 added the unwanted input to the log line - except
2467 that it did not do this in the start of session case. It now does.
2469 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2470 This is wrong and can cause the other end to generate a synchronization
2471 error if it is another Exim or anything else that does the synchronization
2472 check. A QUIT command is no longer sent after a timeout.
2474 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2475 during host lookups.
2477 74. The defer_ok option of callout verification was not working if it was used
2478 when verifying addresses in header lines, that is, for this case:
2480 verify = header_sender/callout=defer_ok
2482 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2483 those file descriptors could be used for SMTP connections. If anything
2484 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2485 could be sent to the SMTP client, causing chaos. The daemon now opens
2486 stdin, stdout, and stderr to /dev/null when it puts itself into the
2489 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2490 log by default. The user can override this with suitable Perl magic.
2492 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2493 the log message when discard triggers, was not working for the DATA ACL or
2494 for the non-SMTP ACL.
2496 78. Error message wording change in sieve.c.
2498 79. If smtp_accept_max_per_host was set, the number of connections could be
2499 restricted to fewer than expected, because the daemon was trying to set up
2500 a new connection before checking whether the processes handling previous
2501 connections had finished. The check for completed processes is now done
2502 earlier. On busy systems, this bug wouldn't be noticed because something
2503 else would have woken the daemon, and it would have reaped the completed
2506 80. If a message was submitted locally by a user whose login name contained one
2507 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2508 It caused a spool format error. I have fixed the spool reading code. A
2509 related problem was that the "from" clause in the Received: line became
2510 illegal because of the space(s). It is now covered by ${quote_local_part.
2512 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2515 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2517 83. Patch from Sieve maintainer to fix unterminated string problem in
2518 "vacation" handling.
2520 84. Some minor changes to the Linux configuration files to help with other
2521 OS variants using glibc.
2523 85. One more patch for Sieve to update vacation handling to latest spec.
2526 ----------------------------------------------------
2527 See the note above about the 4.44 and 4.50 releases.
2528 ----------------------------------------------------
2534 1. Change 4.43/35 introduced a bug that caused file counts to be
2535 incorrectly computed when quota_filecount was set in an appendfile
2538 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2539 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2542 3. Give more explanation in the error message when the command for a transport
2543 filter fails to execute.
2545 4. There are several places where Exim runs a non-Exim command in a
2546 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2547 was being done only for the command run by the queryprogram router. It is
2548 now done for all such subprocesses. The other cases are: ${run, transport
2549 filters, and the commands run by the lmtp and pipe transports.
2551 5. Some older OS have a limit of 256 on the maximum number of file
2552 descriptors. Exim was using setrlimit() to set 1000 as a large value
2553 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2554 systems. I've change it so that if it can't get 1000, it tries for 256.
2556 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2557 was an oversight, and furthermore, ever since the addition of extra
2558 controls (e.g. 4.43/32), the checks on when to allow different forms of
2559 "control" were broken. There should now be diagnostics for all cases when a
2560 control that does not make sense is encountered.
2562 7. $recipients is now available in the predata ACL (oversight).
2564 8. Tidy the search cache before the fork to do a delivery from a message
2565 received from the command line. Otherwise the child will trigger a lookup
2566 failure and thereby defer the delivery if it tries to use (for example) a
2567 cached ldap connection that the parent has called unbind on.
2569 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2570 of $address_data from the recipient verification was clobbered by the
2571 sender verification.
2573 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2574 was its contents. (It was OK if the option was not defined at all.)
2576 11. A "Completed" log line is now written for messages that are removed from
2577 the spool by the -Mrm option.
2579 12. $host_address is now set to the target address during the checking of
2580 ignore_target_hosts.
2582 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2583 being passed; this would have caused $sender_host_name to have been used if
2584 matching the list had actually called for a host name (not very likely,
2585 since this list is usually IP addresses). A host name is now passed as
2588 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2589 code that specifies a non-restarting handler (typically sigaction() in
2590 modern systems) in an attempt to fix a rare and obscure crash bug.
2592 15. Narrowed the window for a race in the daemon that could cause it to ignore
2593 SIGCHLD signals. This is not a major problem, because they are used only to
2594 wake it up if nothing else does.
2596 16. A malformed maildirsize file could cause Exim to calculate negative values
2597 for the mailbox size or file count. Odd effects could occur as a result.
2598 The maildirsize information is now recalculated if the size or filecount
2601 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2602 support for a long time. Removed HAVE_SYS_VFS_H.
2604 18. Updated exipick to current release from John Jetmore.
2606 19. Allow an empty sender to be matched against a lookup in an address list.
2607 Previously the only cases considered were a regular expression, or an
2610 20. Exim went into a mad DNS lookup loop when doing a callout where the
2611 host was specified on the transport, if the DNS lookup yielded more than
2614 21. The RFC2047 encoding function was originally intended for short strings
2615 such as real names; it was not keeping to the 75-character limit for
2616 encoded words that the RFC imposes. It now respects the limit, and
2617 generates multiple encoded words if necessary. To be on the safe side, I
2618 have increased the buffer size for the ${rfc2047: expansion operator from
2621 22. Failure to deliver a bounce message always caused it to be frozen, even if
2622 there was an errors_to setting on the router. The errors_to setting is now
2625 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2626 canonical form (fully expanded) before being placed in
2627 $sender_host_address.
2629 24. Updated eximstats to version 1.33
2631 25. Include certificate and key file names in error message when GnuTLS fails
2632 to set them up, because the GnuTLS error message doesn't include the name
2633 of the failing file when there is a problem reading it.
2635 26. Expand error message when OpenSSL has problems setting up cert/key files.
2638 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2639 (this can affect the format of dates).
2641 28. exim_tidydb, when checking for the continued existence of a message for
2642 which it has found a message-specific retry record, was not finding
2643 messages that were in split spool directories. Consequently, it was
2644 deleting retry records that should have stayed in existence.
2646 29. eximstats updated to version 1.35
2647 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2648 1.35 - bugfix such that pie charts by volume are generated correctly
2650 30. The SPA authentication driver was not abandoning authentication and moving
2651 on to the next authenticator when an expansion was forced to fail,
2652 contradicting the general specification for all authenticators. Instead it
2653 was generating a temporary error. It now behaves as specified.
2655 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2656 (the order specifies the preference for clients). The order is now AES256,
2657 AES128, 3DES, ARCFOUR128.
2659 31. Small patch to Sieve code - explicitly set From: when generating an
2662 32. Exim crashed if a remote delivery caused a very long error message to be
2663 recorded - for instance if somebody sent an entire SpamAssassin report back
2664 as a large number of 550 error lines. This bug was coincidentally fixed by
2665 increasing the size of one of Exim's internal buffers (big_buffer) that
2666 happened as part of the Exiscan merge. However, to be on the safe side, I
2667 have made the code more robust (and fixed the comments that describe what
2670 33. Some experimental protocols are using DNS PTR records for new purposes. The
2671 keys for these records are domain names, not reversed IP addresses. The
2672 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2673 leaves it alone. Component reversal etc. now happens only for IP addresses.
2676 34. The host_aton() function is supposed to be passed a string that is known
2677 to be a valid IP address. However, in the case of IPv6 addresses, it was
2678 not checking this. This is a hostage to fortune. Exim now panics and dies
2679 if the condition is not met. A case was found where this could be provoked
2680 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2681 components; fortuitously, this particular loophole had already been fixed
2682 by change 4.50/55 or 4.44/33 above.
2684 If there are any other similar loopholes, the new check in host_aton()
2685 itself should stop them being exploited. The report I received stated that
2686 data on the command line could provoke the exploit when Exim was running as
2687 exim, but did not say which command line option was involved. All I could
2688 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2689 running as the user.
2692 35. There was a buffer overflow vulnerability in the SPA authentication code
2693 (which came originally from the Samba project). I have added a test to the
2694 spa_base64_to_bits() function which I hope fixes it.
2697 36. The daemon start-up calls getloadavg() while still root for those OS that
2698 need the first call to be done as root, but it missed one case: when
2699 deliver_queue_load_max is set with deliver_drop_privilege. This is
2700 necessary for the benefit of the queue runner, because there is no re-exec
2701 when deliver_drop_privilege is set.
2703 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2704 was in use, was not putting the data itself into the right store pool;
2705 consequently, it could be overwritten for a subsequent message in the same
2706 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2709 38. Sometimes the final signoff response after QUIT could fail to get
2710 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2711 < 0 before doing a fflush(). This bug looks as though it goes back to the
2712 introduction of TLS in release 3.20, but "sometimes" must have been rare
2713 because the tests only now provoked it.
2719 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2720 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2721 Now it returns an integer. A number of calls were still expecting a T/F
2722 return. Fortuitously, in all cases, the tests worked in OK situations,
2723 which is the norm. However, things would have gone wrong on any write
2724 failures on the smtp file descriptor. This function is used when sending
2725 messages over SMTP and also when doing verify callouts.
2727 2. When Exim is called to do synchronous delivery of a locally submitted
2728 message (the -odf or -odi options), it no longer closes stderr before doing
2731 3. Implemented the mua_wrapper option.
2733 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2735 5. Implemented the functions header_remove(), header_testname(),
2736 header_add_at_position(), and receive_remove_recipient(), and exported them
2739 6. If an ACL "warn" statement specified the addition of headers, Exim already
2740 inserted X-ACL-Warn: at the start if there was no header name. However, it
2741 was not making this test for the second and subsequent header lines if
2742 there were newlines in the string. This meant that an invalid header could
2743 be inserted if Exim was badly configured.
2745 7. Allow an ACL "warn" statement to add header lines at the start or after all
2746 the Received: headers, as well as at the end.
2748 8. Added the rcpt_4xx retry error code.
2750 9. Added postmaster_mailfrom=xxx to callout verification option.
2752 10. Added mailfrom=xxxx to the callout verification option, for verify=
2755 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2756 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2758 12. Inserted some casts to stop certain compilers warning when using pointer
2759 differences as field lengths or precisions in printf-type calls (mostly
2760 affecting debugging statements).
2762 13. Added optional readline() support for -be (dynamically loaded).
2764 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2765 same clock tick as a message's arrival, so that its received time was the
2766 same as the "first fail" time on the retry record, and that message
2767 remained on the queue past the ultimate address timeout, every queue runner
2768 would try a delivery (because it was past the ultimate address timeout) but
2769 after another failure, the ultimate address timeout, which should have then
2770 bounced the address, did not kick in. This was a "< instead of <=" error;
2771 in most cases the first failure would have been in the next clock tick
2772 after the received time, and all would be well.
2774 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2775 being recognized when the domain list was tested by the match_domain
2776 condition in an expansion string.
2778 16. Added the ${str2b64: operator.
2780 17. Exim was always calling setrlimit() to set a large limit for the number of
2781 processes, without checking whether the existing limit was already
2782 adequate. (It did check for the limit on file descriptors.) Furthermore,
2783 errors from getrlimit() and setrlimit() were being ignored. Now they are
2784 logged to the main and panic logs, but Exim does carry on, to try to do its
2785 job under whatever limits there are.
2787 18. Imported PCRE 5.0.
2789 19. Trivial typo in log message " temporarily refused connection" (the leading
2792 20. If the log selector return_path_on_delivery was set and an address was
2793 redirected to /dev/null, the delivery process crashed because it assumed
2794 that a return path would always be set for a "successful" delivery. In this
2795 case, the whole delivery is bypassed as an optimization, and therefore no
2798 21. Internal re-arrangement: the function for sending a challenge and reading
2799 a response while authentication was assuming a zero-terminated challenge
2800 string. It's now changed to take a pointer and a length, to allow for
2801 binary data in such strings.
2803 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2805 23. Exim was not respecting finduser_retries when seeking the login of the
2806 uid under which it was called; it was always trying 10 times. (The default
2807 setting of finduser_retries is zero.) Also, it was sleeping after the final
2808 failure, which is pointless.
2810 24. Implemented tls_on_connect_ports.
2812 25. Implemented acl_smtp_predata.
2814 26. If the domain in control=submission is set empty, Exim assumes that the
2815 authenticated id is a complete email address when it generates From: or
2816 Sender: header lines.
2818 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2819 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2820 chown and chgrp in /bin and hostname in /usr/bin.
2822 28. Exim was keeping the "process log" file open after each use, just as it
2823 does for the main log. This opens the possibility of it remaining open for
2824 long periods when the USR1 signal hits a daemon. Occasional processlog
2825 errors were reported, that could have been caused by this. Anyway, it seems
2826 much more sensible not to leave this file open at all, so that is what now
2829 29. The long-running daemon process does not normally write to the log once it
2830 has entered its main loop, and it closes the log before doing so. This is
2831 so that log files can straightforwardly be renamed and moved. However,
2832 there are a couple of unusual error situations where the daemon does write
2833 log entries, and I had neglected to close the log afterwards.
2835 30. The text of an SMTP error response that was received during a remote
2836 delivery was being truncated at 512 bytes. This is too short for some of
2837 the long messages that one sometimes sees. I've increased the limit to
2840 31. It is now possible to make retry rules that apply only when a message has a
2841 specific sender, in particular, an empty sender.
2843 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2844 it possible to be selective about when SMTP synchronization is enforced.
2846 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2848 32. Implemented hosts_connection_nolog.
2850 33. Added an ACL for QUIT.
2852 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2855 35. Added mailbox_size and mailbox_filecount to appendfile.
2857 36. Added control = no_multiline_responses to ACLs.
2859 37. There was a bug in the logic of the code that waits for the clock to tick
2860 in the case where the clock went backwards by a substantial amount such
2861 that the microsecond fraction of "now" was more than the microsecond
2862 fraction of "then" (but the whole seconds number was less).
2864 38. Added support for the libradius Radius client library this is found on
2865 FreeBSD (previously only the radiusclient library was supported).
2871 1. When certain lookups returned multiple values in the form name=value, the
2872 quoting of the values was not always being done properly. Specifically:
2873 (a) If the value started with a double quote, but contained no whitespace,
2875 (b) If the value contained whitespace other than a space character (i.e.
2876 tabs or newlines or carriage returns) it was not quoted.
2877 This fix has been applied to the mysql and pgsql lookups by writing a
2878 separate quoting function and calling it from the lookup code. The fix
2879 should probably also be applied to nisplus, ibase and oracle lookups, but
2880 since I cannot test any of those, I have not disturbed their existing code.
2882 2. A hit in the callout cache for a specific address caused a log line with no
2883 reason for rejecting RCPT. Now it says "Previous (cached) callout
2884 verification failure".
2886 3. There was an off-by-one bug in the queryprogram router. An over-long
2887 return line was truncated at 256 instead of 255 characters, thereby
2888 overflowing its buffer with the terminating zero. As well as fixing this, I
2889 have increased the buffer size to 1024 (and made a note to document this).
2891 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2892 when Exim is waiting for an SMTP response from a remote server, Exim
2893 restarts its select() call on the socket, thereby resetting its timeout.
2894 This is not a problem when such interrupts are rare. Somebody set up a cron
2895 job to run exiwhat every 2 minutes, which is less than the normal select()
2896 timeout (5 or 10 minutes). This meant that the select() timeout never
2897 kicked in because it was always reset. I have fixed this by comparing the
2898 time when an interrupt arrives with the time at the start of the first call
2899 to select(). If more time than the timeout has elapsed, the interrupt is
2900 treated as a timeout.
2902 5. Some internal re-factoring in preparation for the addition of Sieve
2903 extensions (by MH). In particular, the "personal" test is moved to a
2904 separate function, and given an option for scanning Cc: and Bcc: (which is
2905 not set for Exim filters).
2907 6. When Exim created an email address using the login of the caller as the
2908 local part (e.g. when creating a From: or Sender: header line), it was not
2909 quoting the local part when it contained special characters such as @.
2911 7. Installed new OpenBSD configuration files.
2913 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2914 try to make them clearer.
2916 9. Callout options, other than the timeout value, were being ignored when
2917 verifying sender addresses in header lines. For example, when using
2919 verify = header_sender/callout=no_cache
2921 the cache was (incorrectly) being used.
2923 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2924 only the Cygwin environment.
2926 11. When return_path_on_delivery was set as a log selector, if different remote
2927 addresses in the same message used different return paths and parallel
2928 remote delivery occurred, the wrong values would sometimes be logged.
2929 (Whenever a remote delivery process finished, the return path value from
2930 the most recently started remote delivery process was logged.)
2932 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2933 lines when AUTH and/or TLS are in use. This is the "received protocol"
2934 field. Exim used to use "asmtp" for authenticated SMTP, without any
2935 indication (in the protocol name) for TLS use. Now it follows the RFC and
2936 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2937 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2938 names appear in log lines as well as in Received: header lines.
2940 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2941 extensions, and comparison tests, and to fix some bugs.
2943 14. Changes to the "personal" filter test:
2945 (1) The test was buggy in that it was just doing the equivalent of
2946 "contains" tests on header lines. For example, if a user's address was
2947 anne@some.where, the "personal" test would incorrectly be true for
2949 To: susanne@some.where
2951 This test is now done by extracting each address from the header in turn,
2952 and checking the entire address. Other tests that are part of "personal"
2953 are now done using regular expressions (for example, to check local parts
2954 of addresses in From: header lines).
2956 (2) The list of non-personal local parts in From: addresses has been
2957 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2958 taken from the Sieve specification recommendations.
2960 (3) If the message contains any header line starting with "List-" it is
2961 treated as non-personal.
2963 (4) The test for "circular" in the Subject: header line has been removed
2964 because it now seems ill-conceived.
2966 15. Minor typos in src/EDITME comments corrected.
2968 16. Installed latest exipick from John Jetmore.
2970 17. If headers_add on a router specified a text string that was too long for
2971 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2972 of string_sprintf() is now avoided.
2974 18. $message_body_size was not set (it was always zero) when running the DATA
2975 ACL and the local_scan() function.
2977 19. For the "mail" command in an Exim filter, no default was being set for
2978 the once_repeat time, causing a random time value to be used if "once" was
2979 specified. (If the value happened to be <= 0, no repeat happened.) The
2980 default is now 0s, meaning "never repeat". The "vacation" command was OK
2981 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2982 (I found it when inspecting the code).
2984 20. There is now an overall timeout for performing a callout verification. It
2985 defaults to 4 times the callout timeout, which applies to individual SMTP
2986 commands during the callout. The overall timeout applies when there is more
2987 than one host that can be tried. The timeout is checked before trying the
2988 next host. This prevents very long delays if there are a large number of
2989 hosts and all are timing out (e.g. when the network connections are timing
2990 out). The value of the overall timeout can be changed by specifying an
2991 additional sub-option for "callout", called "maxwait". For example:
2993 verify = sender/callout=5s,maxwait=20s
2995 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2996 to the end before writing, but this should make it even safer).
2998 22. Exim was forgetting that it had advertised PIPELINING for the second and
2999 subsequent messages on an SMTP connection. It was also not resetting its
3000 memory on STARTTLS and an internal HELO.
3002 23. When Exim logs an SMTP synchronization error within a session, it now
3003 records whether PIPELINING has been advertised or not.
3005 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3006 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3007 rather than long int.
3009 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3011 26. Added the never_mail option to autoreply.
3017 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3018 crash if the getsockname() call failed; this can happen if a connection is
3019 closed very soon after it is established. The problem was simply in the
3020 order in which certain operations were done, causing Exim to try to write
3021 to the SMTP stream before it had set up the file descriptor. The bug has
3022 been fixed by making things happen in the correct order.
3028 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3029 before the connection was closed, thus losing the rejection response.
3031 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3032 some early Solaris releases, but causes trouble in current releases where
3033 socklen_t is defined.
3035 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3038 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3041 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3042 that releasing the top of it at the end releases what was used for sub-
3043 expansions (unless the block got too big). However, discard this block if
3044 the first thing is a variable or header, so that we can use its block when
3045 it is dynamic (useful for very large $message_headers, for example).
3047 6. Lookups now cache *every* query, not just the most recent. A new, separate
3048 store pool is used for this. It can be recovered when all lookup caches are
3049 flushed. Lookups now release memory at the end of their result strings.
3050 This has involved some general refactoring of the lookup sources.
3052 7. Some code has been added to the store_xxx() functions to reduce the amount
3053 of flapping under certain conditions.
3055 8. log_incoming_interface used to affect only the <= reception log lines. Now
3056 it causes the local interface and port to be added to several more SMTP log
3057 lines, for example "SMTP connection from", and rejection lines.
3059 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3061 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3063 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3064 could be overwritten at the end of the current message (or the start of a
3065 new message if it was set in a HELO ACL). The value is now preserved for
3066 the duration of the SMTP connection.
3068 12. If a transport had a headers_rewrite setting, and a matching header line
3069 contained an unqualified address, that address was qualified, even if it
3070 did not match any rewriting rules. The underlying bug was that the values
3071 of the flags that permit the existence of unqualified sender and recipient
3072 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3073 non-local messages, and by -bnq for local messages) were not being
3074 preserved with the message after it was received.
3076 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3077 "next input=" as part of the text comprising the host identity instead of
3078 the correct text. The code was using the same buffer for two different
3079 strings. However, depending on which order the printing function evaluated
3080 its arguments, the bug did not always show up. Under Linux, for example, my
3081 test suite worked just fine.
3083 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3084 doesn't work with some older versions of Perl. It has been changed to "my",
3085 which in any case is probably the better facility to use.
3087 15. A really picky compiler found some instances of statements for creating
3088 error messages that either had too many or two few arguments for the format
3091 16. The size of the buffer for calls to the DNS resolver has been increased
3092 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3093 for addresses that have a lot of PTR records. This alleviates a problem; it
3094 does not fully solve it.
3096 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3097 buffer now truncates the list and logs the incident, which is the same
3098 action as happens when Exim is looking up a host name and its aliases.
3099 Previously in this situation something unpredictable would happen;
3100 sometimes it was "internal error: store_reset failed".
3102 18. If a server dropped the connection unexpectedly when an Exim client was
3103 using GnuTLS and trying to read a response, the client delivery process
3104 crashed while trying to generate an error log message.
3106 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3107 string, for example:
3109 warn message = H1: something\nH2: something
3111 the text was added as a single header line from Exim's point of view
3112 though it ended up OK in the delivered message. However, searching for the
3113 second and subsequent header lines using $h_h2: did not work. This has been
3114 fixed. Similarly, if a system filter added multiple headers in this way,
3115 the routers could not see them.
3117 20. Expanded the error message when iplsearch is called with an invalid key to
3118 suggest using net-iplsearch in a host list.
3120 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3121 ACLs are no longer actually imposed (and a message to that effect is
3124 22. If a "gecos" field in a passwd entry contained escaped characters, in
3125 particular, if it contained a \" sequence, Exim got it wrong when building
3126 a From: or a Sender: header from that name. A second bug also caused
3127 incorrect handling when an unquoted " was present following a character
3128 that needed quoting.
3130 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3131 was not being matched caselessly.
3133 24. Arranged for all hyphens in the exim.8 source to be escaped with
3136 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3137 properly. Recipient callout cache records were still being keyed to include
3138 the sender, even when use_sender was set false. This led to far more
3139 callouts that were necessary. The sender is no longer included in the key
3140 when use_sender is false.
3142 26. Added "control = submission" modifier to ACLs.
3144 27. Added the ${base62d: operator to decode base 62 numbers.
3146 28. dnsdb lookups can now access SRV records.
3148 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3149 the configuration file.
3151 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3152 (-v) mode. This makes the output for a verbose queue run more intelligible.
3154 31. Added a use_postmaster feature to recipient callouts.
3156 32. Added the $body_zerocount variable, containing the number of binary zero
3157 bytes in the message body.
3159 33. The time of last modification of the "new" subdirectory is now used as the
3160 "mailbox time last read" when there is a quota error for a maildir
3163 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3165 35. Added +ignore_unknown as a special item in host lists.
3167 36. Code for decoding IPv6 addresses in host lists is now included, even if
3168 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3169 address was recognized as an IP address, but was then not correctly decoded
3170 into binary, causing unexpected and incorrect effects when compared with
3177 1. Very minor rewording of debugging text in manualroute to say "list of
3178 hosts" instead of "hostlist".
3180 2. If verify=header_syntax was set, and a header line with an unqualified
3181 address (no domain) and a large number of spaces between the end of the
3182 name and the colon was received, the reception process suffered a buffer
3183 overflow, and (when I tested it) crashed. This was caused by some obsolete
3184 code that should have been removed. The fix is to remove it!
3186 3. When running in the test harness, delay a bit after writing a bounce
3187 message to get a bit more predictability in the log output.
3189 4. Added a call to search_tidyup() just before forking a reception process. In
3190 theory, someone could use a lookup in the expansion of smtp_accept_max_
3191 per_host which, without the tidyup, could leave open a database connection.
3193 5. Added the variables $recipient_data and $sender_data which get set from a
3194 lookup success in an ACL "recipients" or "senders" condition, or a router
3195 "senders" option, similar to $domain_data and $local_part_data.
3197 6. Moved the writing of debug_print from before to after the "senders" test
3200 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3201 problems for message scanning, either using a data ACL, or using
3202 local_scan() because the Received: header was not generated till after they
3203 were called (in order to set the time as the time of reception completion).
3204 I have revised the way this works. The header is now generated after the
3205 body is received, but before the ACL or local_scan() are called. After they
3206 are run, the timestamp in the header is updated.
3212 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3213 before starting a queue runner without re-exec. This happened only when
3214 deliver_drop_privilege was set or when the Exim user was set to root. The
3215 effect of the bug was that timeouts during subsequent deliveries caused
3216 crashes instead of being properly handled. The handler is now left at its
3217 default (and expected) setting.
3219 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3220 message, again when deliver_drop_privilege is set or Exim is run as root.
3221 The bug described in (1) was not present in this case, but the tidying up
3222 of the other signals was missing. I have made the two cases consistent.
3224 3. The ignore_target_hosts setting on a manualroute router was being ignored
3225 for hosts that were looked up using the /MX notation.
3227 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3230 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3231 operated on the sender address. After changing the $sender_address to <>
3232 for the sender address verify, Exim was re-instated it as the original
3233 (before rewriting) address, but remembering that it had rewritten it, so it
3234 wasn't rewriting it again. This bug also had the effect of breaking the
3235 sender address verification caching when the sender address was rewritten.
3237 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3238 This has been changed so that if the ip literal address matches
3239 ignore_target_hosts, the router declines.
3241 7. Added expansion conditions match_domain, match_address, and match_local_
3242 part (NOT match_host).
3244 8. The placeholder for the Received: header didn't have a length field set.
3246 9. Added code to Exim itself and to exim_lock to test for a specific race
3247 condition that could lead to file corruption when using MBX delivery. The
3248 issue is with the lockfile that is created in /tmp. If this file is removed
3249 after a process has opened it but before that process has acquired a lock,
3250 there is the potential for a second process to recreate the file and also
3251 acquire a lock. This could lead to two Exim processes writing to the file
3252 at the same time. The added code performs the same test as UW imapd; it
3253 checks after acquiring the lock that its file descriptor still refers to
3254 the same named file.
3256 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3257 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3260 11. Added the smtp_active_hostname option. If used, this will typically be made
3261 to depend on the incoming interface address. Because $interface_address is
3262 not set up until the daemon has forked a reception process, error responses
3263 that can happen earlier (such as "too many connections") no longer contain
3266 12. If an expansion in a condition on a "warn" statement fails because a lookup
3267 defers, the "warn" statement is abandoned, and the next ACL statement is
3268 processed. Previously this caused the whole ACL to be aborted.
3270 13. Added the iplsearch lookup type.
3272 14. Added ident_timeout as a log selector.
3274 15. Added tls_certificate_verified as a log selector.
3276 16. Added a global option tls_require_ciphers (compare the smtp transport
3277 option of the same name). This controls incoming TLS connections.
3279 17. I finally figured out how to make tls_require_ciphers do a similar thing
3280 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3281 before starting the TLS session.
3283 18. Tabs are now shown as \t in -bP output.
3285 19. If the log selector return_path_on_delivery was set, Exim crashed when
3286 bouncing a message because it had too many Received: header lines.
3288 20. If two routers both had headers_remove settings, and the first one included
3289 a superfluous trailing colon, the final name in the first list and the
3290 first name in the second list were incorrectly joined into one item (with a
3291 colon in the middle).
3297 1. Added -C and -D options to the exinext utility, mainly to make it easier
3298 to include in the automated testing, but these could be helpful when
3299 multiple configurations are in use.
3301 2. The exinext utility was not formatting the output nicely when there was
3302 an alternate port involved in the retry record key, nor when there was a
3303 message id as well (for retries that were specific to a specific message
3304 and a specific host). It was also confused by IPv6 addresses, because of
3305 the additional colons they contain. I have fixed the IPv4 problem, and
3306 patched it up to do a reasonable job for IPv6.
3308 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3309 delivery, the log line now contains "pipelined" if PIPELINING was used.
3311 4. An SMTP transport process used to panic and die if the bind() call to set
3312 an explicit outgoing interface failed. This has been changed; it is now
3313 treated in the same way as a connect() failure.
3315 5. A reference to $sender_host_name in the part of a conditional expansion
3316 that was being skipped was still causing a DNS lookup. This no longer
3319 6. The def: expansion condition was not recognizing references to header lines
3320 that used bh_ and bheader_.
3322 7. Added the _cache feature to named lists.
3324 8. The code for checking quota_filecount in the appendfile transport was
3325 allowing one more file than it should have been.
3327 9. For compatibility with Sendmail, the command line option
3335 and sets the incoming protocol and host name (for trusted callers). The
3336 host name and its colon can be omitted when only the protocol is to be set.
3337 Note the Exim already has two private options, -pd and -ps, that refer to
3338 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3339 "s", but I don't think that's a major issue.
3341 10. A number of refactoring changes to the code, none of which should affect
3344 (a) The number of logging options was getting close to filling up the
3345 32-bit word that was used as a bit map. I have split them into two classes:
3346 those that are passed in the argument to log_write(), and those that are
3347 only ever tested independently outside of that function. These are now in
3348 separate 32-bit words, so there is plenty of room for expansion again.
3349 There is no change in the user interface or the logging behaviour.
3351 (b) When building, for example, log lines, the code previously used a
3352 macro that called string_cat() twice, in order to add two strings. This is
3353 not really sufficiently general. Furthermore, there was one instance where
3354 it was actually wrong because one of the argument was used twice, and in
3355 one call a function was used. (As it happened, calling the function twice
3356 did not affect the overall behaviour.) The macro has been replaced by a
3357 function that can join an arbitrary number of extra strings onto a growing
3360 (c) The code for expansion conditions now uses a table and a binary chop
3361 instead of a serial search (which was left over from when there were very
3362 few conditions). Also, it now recognizes conditions like "pam" even when
3363 the relevant support is not compiled in: a suitably worded error message is
3364 given if an attempt is made to use such a condition.
3366 11. Added ${time_interval:xxxxx}.
3368 12. A bug was causing one of the ddress fields not to be passed back correctly
3369 from remote delivery subprocesses. The field in question was not being
3370 subsequently used, so this caused to problems in practice.
3372 13. Added new log selectors queue_time and deliver_time.
3374 14. Might have fixed a bug in maildirsizefile handling that threw up
3375 "unexpected character" debug warnings, and recalculated the data
3376 unnecessarily. In any case, I expanded the warning message to give more
3379 15. Added the message "Restricted characters in address" to the statements in
3380 the default ACL that block characters like @ and % in local parts.
3382 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3383 Three changes have been made:
3385 (a) There was a serious bug; a negative response to MAIL caused the whole
3386 recipient domain to be cached as invalid, thereby blocking all messages
3387 to all local parts at the same domain, from all senders. This bug has
3388 been fixed. The domain is no longer cached after a negative response to
3389 MAIL if the sender used is not empty.
3391 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3394 (c) A new callout option, "use_sender" has been added for people who want
3395 the modified behaviour.
3401 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3404 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3405 indeed breaks things for older releases.
3407 3. Added additional logging to the case where there is a problem reading data
3408 from a filter that is running in a subprocess using a pipe, in order to
3409 try to track down a specific problem.
3411 4. Testing facility fudge: when running in the test harness and attempting
3412 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3413 getting "No route to host". Convert this to a timeout.
3415 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3418 6. Some OS don't have socklen_t but use size_t instead. This affects the
3419 fifth argument of getsockopt() amongst other things. This is now
3420 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3421 can be set for individual OS. I have set it for SunOS5, OSF1, and
3422 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3423 some earlier ones do not.
3425 7. Change 4.30/15 was not doing the test caselessly.
3427 8. The standard form for an IPv6 address literal was being rejected by address
3428 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3429 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3430 this, as well as the form without the "IPv6" on the front (but only when
3431 address literals are enabled, of course).
3433 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3435 10. Exim crashed if a message with an empty sender address specified by -f
3436 encountered a router with an errors_to setting. This could be provoked only
3437 by a command such as
3441 where an empty string was supplied; "<>" did not hit this bug.
3443 11. Installed PCRE release 4.5.
3445 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3446 remained set. It is now erased.
3448 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3449 times from message ids (which are base 36 rather than the normal 62).
3451 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3452 were being counted as actual protocol errors, and logged if the log
3453 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3454 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3455 and DATA following a set of rejected RCPTs do not count as protocol errors.
3456 In other words, Exim assumes they were pipelined, though this may not
3457 actually be the case. Of course, in all cases the client gets an
3458 appropriate error code.
3460 15. If a lookup fails in an ACL condition, a message about the failure may
3461 be available; it is used if testing the ACL cannot continue, because most
3462 such messages specify what the cause of the deferral is. However, some
3463 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3464 that caused an old message to be retained and used if a later statement
3465 caused a defer, replacing the real cause of the deferral.
3467 16. If an IP address had so many PTR records that the DNS lookup buffer
3468 was not large enough to hold them, Exim could crash while trying to process
3469 the truncated data. It now detects and logs this case.
3471 17. Further to 4.21/58, another change has been made: if (and only if) the
3472 first line of a message (the first header line) ends with CRLF, a bare LF
3473 in a subsequent header line has a space inserted after it, so as not to
3474 terminate the header.
3476 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3477 unnecessarily, used atoi() instead of strtol(), and didn't check the
3478 termination when getting file sizes from file names by regex.
3480 19. Completely re-implemented the support for maildirsize files, in the light
3481 of a number of problems with the previous contributed implementation
3482 (4.30/29). In particular:
3484 . If the quota is zero, the maildirsize file is maintained, but no quota is
3487 . If the maildir directory does not exist, it is created before any attempt
3488 to write a maildirsize file.
3490 . The quota value in the file is just a cache; if the quota is changed in
3491 the transport, the new value overrides.
3493 . A regular expression is available for excluding directories from the
3496 20. The autoreply transport checks the characters in options that define the
3497 message's headers; it allows continued headers, but it was checking with
3498 isspace() after an embedded newline instead of explicitly looking for a
3501 21. If all the "regular" hosts to which an address was routed had passed their
3502 expiry times, and had not reached their retry times, the address was
3503 bounced, even if fallback hosts were defined. Now Exim should go on to try
3506 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3507 equivalent code in the SMTP transport. Some hosts send humungous responses
3508 to HELO/EHLO, more than 1024 it seems.
3510 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3511 gives compiler warnings in some environments. I've now done it "properly",
3514 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3515 (because of problems with the built-in one) was declared to return uschar *
3516 instead of char *, causing compiler failure.
3518 25. Fixed a file descriptor leak when processing alias/forward files.
3520 26. Fixed a minor format string issue in dbfn.c.
3522 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3524 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3525 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3528 29. When a sender address is verified, it is cached, to save repeating the test
3529 when there is more than one recipient in a message. However, when the
3530 verification involves a callout, it is possible for different callout
3531 options to be set for different recipients. It is too complicated to keep
3532 track of this in the cache, so now Exim always runs a verification when a
3533 callout is required, relying on the callout cache for the optimization.
3534 The overhead is duplication of the address routing, but this should not be
3537 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3538 to be verified with callout=postmaster, and the main callout worked but the
3539 postmaster check failed, the verification correctly failed. However, if a
3540 subsequent RCPT command asked for sender verification *without* the
3541 postmaster check, incorrect caching caused this verification also to fail,
3544 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3545 it was not caching the DNS options (qualify_single, search_parents) that
3546 were used when the lookup failed. A subsequent lookup with different
3547 options therefore always gave the same answer, though there were cases
3548 where it should not have. (Example: a "domains = !$mx_any" option on a
3549 dnslookup router: the "domains" option is always processed without any
3550 widening, but the router might have qualify_single set.) Now Exim uses the
3551 cached value only when the same options are set.
3553 32. Added John Jetmore's "exipick" utility to the distribution.
3555 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3556 than a timeout (e.g. a certificate is required, and is not provided), an
3557 Exim server now closes the connection immediately. Previously it waited for
3558 the client to close - but if the client is SSL, it seems that they each
3559 wait for each other, leading to a delay before one of them times out.
3561 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3562 maintained 0.8.x compatibility because I don't think many are using it, and
3563 it is clearly obsolete.
3565 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3568 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3569 client certificate was expired. A simple patch fixes this, though I don't
3570 understand the full logic of why the verify callback is called multiple
3573 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3574 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3575 which causes problems with some clients (such as the Certicom SSL Plus
3576 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3577 disables the coutermeasure allowing Eudora to connect."
3579 38. Exim was not checking that a write() to a log file succeeded. This could
3580 lead to Bad Things if a log got too big, in particular if it hit a file
3581 size limit. Exim now panics and dies if it cannot write to a log file, just
3582 as it does if it cannot open a log file.
3584 39. Modified OS/Makefile-Linux so that it now contains
3586 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3588 The two -D definitions ensure that Exim is compiled with large file
3589 support, which makes it possible to handle log files that are bigger than
3592 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3593 instance) a domain was checked against a named list that involved a lookup,
3594 causing $domain_data to be set, then another domain was checked against the
3595 same list, then the first domain was re-checked, the value of $domain_data
3596 after the final check could be wrong. In particular, if the second check
3597 failed, it could be set empty. This bug probably also applied to
3600 41. The strip_trailing_dot option was not being applied to the address given
3601 with the -f command-line option.
3603 42. The code for reading a message's header from the spool was incrementing
3604 $received_count, but never initializing it. This meant that the value was
3605 incorrect (doubled) while delivering a message in the same process in which
3606 it was received. In the most common configuration of Exim, this never
3607 happens - a fresh exec is done - but it can happen when
3608 deliver_drop_privilege is set.
3610 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3611 it now includes up to 150 characters of the unexpected data in the log
3614 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3615 and building data strings. The size of both of these buffers was 10 000
3616 bytes - far larger than anybody would *ever* want, thought I. Needless to
3617 say, somebody hit the limit. I have increased the maximum line length to
3618 20 000 and the maximum data length of concatenated lines to 100 000. I have
3619 also fixed two bugs, because there was no checking on these buffers. Tsk,
3620 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3621 buffer is too small.
3623 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3624 lsearch lookups. Now it does.
3626 46. When parsing a route_list item in a manualroute router, a fixed-length
3627 buffer was used for the list of hosts. I made this 1024 bytes long,
3628 thinking that nobody would ever have a list of hosts that long. Wrong.
3629 Somebody had a whole pile of complicated expansion conditions, and the
3630 string was silently truncated, leading to an expansion error. It turns out
3631 that it is easier to change to an unlimited length (owing to other changes
3632 that have happened since this code was originally written) than to build
3633 structure for giving a limitation error. The length of the item that
3634 expands into the list of hosts is now unlimited.
3636 47. The lsearch lookup could not handle data where the length of text line was
3637 more than 4095 characters. Such lines were truncated, leading to shortened
3638 data being returned. It should now handle lines of any length.
3640 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3641 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3644 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3645 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3646 no longer happens in comment lines. A list of replacements is now placed
3647 at the head of all of the source files, except those whose only change is
3648 to replace PERL_COMMAND in the very first #! line.
3650 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3651 messages on the queue, with a bottom-up merge sort, using code contributed
3652 by Michael Haardt. This should make operations like -bp somewhat faster on
3653 large queues. It won't affect queue runners, except when queue_run_in_order
3656 51. Installed eximstats 1.31 in the distribution.
3658 52. Added support for SRV lookups to the dnslookup router.
3660 53. If an ACL referred to $message_body or $message_body_end, the value was not
3661 reset for any messages that followed in the same SMTP session.
3663 54. The store-handling optimization for building very long strings was not
3664 differentiating between the different store pools. I don't think this
3665 actually made any difference in practice, but I've tidied it.
3667 55. While running the routers to verify a sender address, $sender_address
3668 was still set to the sender address. This is wrong, because when routing to
3669 send a bounce to the sender, it would be empty. Therefore, I have changed
3670 it so that, while verifying a sender address, $sender_address is set to <>.
3671 (There is no change to what happens when verifying a recipient address.)
3673 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3674 target A or AAAA records (if not already returned) without resetting the
3675 qualify_single or search_parents options of the DNS resolver. These are
3676 inappropriate in this case because the targets of MX and SRV records must
3677 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3678 target that, when qualified, matched something in the local domain. These
3679 two options are now turned off when doing these lookups.
3681 57. It seems that at least some releases of Reiserfs (which does not have the
3682 concept of a fixed number of inodes) returns zero and not -1 for the
3683 number of available inodes. This interacted badly with check_spool_inodes,
3684 which assumed that -1 was the "no such thing" setting. What I have done is
3685 to check that the total number of inodes is greater than zero before doing
3686 the test of how many are available.
3688 58. When a "warn" ACL statement has a log_message modifier, the message is
3689 remembered, and not repeated. This is to avoid a lot of repetition when a
3690 message has many recipients that cause the same warning to be written.
3691 Howewer, Exim was preserving the list of already written lines for an
3692 entire SMTP session, which doesn't seem right. The memory is now reset if a
3693 new message is started.
3695 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3696 debugging output unless log_rewrite was also set.
3698 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3699 of (int)(handle) when we know that handle contains (void *)(-1).
3701 61. The Exim daemon panic-logs an error return when it closes the incoming
3702 connection. However "connection reset by peer" seems to be common, and
3703 isn't really an error worthy of noting specially, so that particular error
3706 62. When Exim is trying to find all the local interfaces, it used to panic and
3707 die if the ioctl to get the interface flags failed. However, it seems that
3708 on at least one OS (Solaris 9) it is possible to have an interface that is
3709 included in the list of interfaces, but for which you get a failure error
3710 for this call. This happens when the interface is not "plumbed" into a
3711 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3712 failure of the "get flags" call assumes that the interface is down.
3714 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3715 makes life easier for people who are doing arithmetic on fields extracted
3716 from dates, where you often get leading zeros that should not be
3717 interpreted as octal.
3719 64. Added qualify_domain to the redirect router, to override the global
3722 65. If a pathologically long header line contained very many addresses (the
3723 report of this problem mentioned 10 000) and each of them was rewritten,
3724 Exim could use up a very large amount of memory. (It kept on making new
3725 copies of the header line as it rewrote, and never released the old ones.)
3726 At the expense of a bit more processing, the header rewriting function has
3727 been changed so that it no longer eats memory in this way.
3729 66. The generation of the Received: header has been moved from the time that a
3730 message starts to be received, to the time that it finishes. The timestamp
3731 in the Received: header should now be very close to that of the <= log
3732 line. There are two side-effects of this change:
3734 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3735 logged header lines no longer include the local Received: line, because
3736 it has not yet been created. The same applies to a copy of the message
3737 that is returned to a non-SMTP sender when a message is rejected.
3739 (b) When a filter file is tested using -bf, no additional Received: header
3740 is added to the test message. After some thought, I decided that this
3743 This change does not affect the value of $received_for. It is still set
3744 after address rewriting, but before local_scan() is called.
3746 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3748 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3749 gave an unhelpful panic error message, and a defer error. I have managed to
3750 change this behaviour so that it now rejects any supplied certificate,
3751 which seems right, as the list of acceptable certificates is empty.
3753 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3754 gave an unhelpful defer error. I have not managed to make this reject any
3755 supplied certificates, but the error message it gives is "no certificate
3756 supplied", which is not helpful.
3758 70. exigrep's output now also includes lines that are not associated with any
3759 message, but which match the given pattern. Implemented by a patch from
3760 Martin Sluka, which also tidied up the Perl a bit.
3762 71. Recipient callout verification, like sender verification, was using <> in
3763 the MAIL FROM command. This isn't really the right thing, since the actual
3764 sender may affect whether the remote host accepts the recipient or not. I
3765 have changed it to use the actual sender in the callout; this means that
3766 the cache record is now keyed on a recipient/sender pair, not just the
3767 recipient address. There doesn't seem to be a real danger of callout loops,
3768 since a callout by the remote host to check the sender would use <>.
3769 [SEE ABOVE: changed after hitting problems.]
3771 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3772 temporary errors. However, in the case of such a code being given after
3773 the end of a data transmission (i.e. after ".") Exim was failing to write
3774 a retry record for the message. (Yes, there was some broken host that was
3775 actually sending 8xx at this point.)
3777 73. An unknown lookup type in a host list could cause Exim to panic-die when
3778 the list was checked. (An example that provoked this was putting <; in the
3779 middle of a list instead of at the start.) If this happened during a DATA
3780 ACL check, a -D file could be left lying around. This kind of configuration
3781 error no longer causes Exim to die; instead it causes a defer errror. The
3782 incident is still logged to the main and panic logs.
3784 74. Buglet left over from Exim 3 conversion. The message "too many messages
3785 in one connection" was written to the rejectlog but not the mainlog, except
3786 when address rewriting (yes!) was being logged.
3788 75. Added write_rejectlog option.
3790 76. When a system filter was run not as root (that is, when system_filter_user
3791 was set), the values of the $n variables were not being returned to the
3792 main process; thus, they were not subsequently available in the $sn
3795 77. Added +return_path_on_delivery log selector.
3797 78. A connection timeout was being treated differently from recipients deferred
3798 when testing hosts_max_try with a message that was older than the host's
3799 retry timeout. (The host should not be counted, thus allowing all hosts to
3800 be tried at least once before bouncing.) This may have been the cause of an
3801 occasionally reported bug whereby a message would remain on the queue
3802 longer than the retry timeout, but would be bounced if a delivery was
3803 forced. I say "may" because I never totally pinned down the problem;
3804 setting up timeout/retry tests is difficult. See also the next item.
3806 79. The ultimate address timeout was not being applied to errors that involved
3807 a combination of host plus message (for example, a timeout on a MAIL
3808 command). When an address resolved to a number of possible hosts, and they
3809 were not all tried for each delivery (e.g. because of hosts_max_try), a
3810 message could remain on the queue longer than the retry timeout.
3812 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3815 81. Fixed an obscure SMTP outgoing bug which required at least the following
3816 conditions: (a) there was another message waiting for the same server;
3817 (b) the server returned 5xx to all RCPT commands in the first message so
3818 that the message was not completed; (c) the server dropped the connection
3819 or gave a negative response to the RSET that Exim sends to abort the
3820 transaction. The observed case was a dropped connection after DATA that had
3821 been sent in pipelining mode. That is, the server had advertised PIPELINING
3822 but was not implementing it correctly. The effect of the bug was incorrect
3823 behaviour, such as trying another host, and this could lead to a crash.
3829 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3830 and daemon.c were passed as pointers to ints; they should have been
3831 pointers to socklen_t variables (which are typically unsigned ints).
3833 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3836 3. Fixed a really odd bug that affected only the testing scheme; patching a
3837 certain fixed string in the binary changed the value of another string that
3838 happened to be identical to the end of the original first string.
3840 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3841 name", it returns that address as the IP address. On some operating
3842 systems (e.g. Solaris), it also passes back the IP address string as the
3843 "host name". However, on others (e.g. Linux), it passes back an empty
3844 string. Exim wasn't checking for this, and was changing the host name to an
3845 empty string, assuming it had been canonicized.
3847 5. Although rare, it is permitted to have more than one PTR record for a given
3848 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3849 all the names associated with an address, because they do in Solaris.
3850 However, it seems that they do not in Linux for data that comes from the
3851 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3852 I found this out when I moved to a new Linux workstation and tried to run
3853 the Exim test suite.
3855 To get round this problem I have changed the code so that it now does its
3856 own call to the DNS to look up PTR records when searching for a host name.
3857 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3858 addresses that are only in /etc/hosts are still found.
3860 This behaviour is, however, controlled by an option called host_lookup_
3861 order, which defaults to "bydns:byaddr". If people want to use the other
3862 order, or indeed, just use one or the other means of lookup, they can
3863 specify it in this variable.
3865 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3866 some operating systems, this comes back from gethostbyaddr() as an empty
3867 string, and this is what Exim used to test for. However, it seems that in
3868 other systems, "." is yielded. Exim now tests for this case too.
3870 7. The values of check_spool_space and check_log_space are now held internally
3871 as a number of kilobytes instead of an absolute number of bytes. If a
3872 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3873 kilobyte. This means that much larger values can be stored.
3875 8. Exim monitor: an attempt to get the action menu when not actually pointing
3876 at a message produces an empty menu entitled "No message selected". This
3877 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3878 no entries in it ("Shell widget menu has zero width and/or height"). So I
3879 have added a single, blank menu entry in this case.
3881 9. Added ${quote_local_part.
3883 10. MIME decoding is now applied to the contents of Subject: header lines when
3886 11. Now that a reference to $sender_host_address automatically causes a reverse
3887 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3888 host lookup before query-style lookups in lists that might use this
3889 variable. This has therefore been abolished, and the "net-" prefix is no
3890 longer necessary for query-style lookups.
3892 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3893 have been a typo for LFLAGS, so it has been changed.
3895 13. The install script calls Exim with "-C /dev/null" in order to find the
3896 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3897 to be output. Howeve, since Exim outputs its version number before the
3898 error, it didn't break the script. It just looked ugly. I fixed this by
3899 always allowing "-C /dev/null" if the caller is root.
3901 14. Ignore overlarge ACL variable number when reading spool file - insurance
3902 against a later release with more variables having written the file.
3904 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3905 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3906 this, as well as the form without the "IPv6" on the front.
3908 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3909 OS/Makefile-Darwin file.
3911 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3912 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3914 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3915 variable. However, some OS use ints for inodes. I've added cast to long int
3916 to get rid of the compiler warning.
3918 19. I had forgotten to lock out "/../" in configuration file names when
3919 ALT_CONFIG_PREFIX was set.
3921 20. Routers used for verification do not need to specify transports. However,
3922 if such a router generated a host list, and callout was configured, Exim
3923 crashed, because it could not find a port number from the (non-existent)
3924 transport. It now assumes port 25 in this circumstance.
3926 21. Added the -t option to exigrep.
3928 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3929 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3930 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3931 them is set, LOOKUP_LSEARCH is forced.
3933 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3935 24. Added sender and host information to the "rejected by local_scan()" log
3936 line; previously there was no indication of these.
3938 25. Added .include_if_exists.
3940 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3941 Linux. It turns out that not all file systems support this. Apparently some
3942 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3943 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3944 not supported on the file descriptor, is now ignored when Exim is trying to
3945 sync a directory. This applies only to Linux.
3947 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3949 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3950 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3953 29. Added support for maildirsize files from supplied patch (modified a bit).
3955 30. The use of :fail: followed by an empty string could lead Exim to respond to
3956 sender verification failures with (e.g.):
3958 550 Verification failed for <xxx>
3959 550 Sender verify failed
3961 where the first response line was missing the '-' that indicates it is not
3962 the final line of the response.
3964 31. The loop for finding the name of the user that called Exim had a hardwired
3965 limit of 10; it now uses the value of finduser_retries, which is used for
3966 all other user lookups.
3968 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3971 33. Exim was neglecting to zero errno before one call of strtol() when
3972 expanding a string and expecting an integer value. On some systems this
3973 resulted in spurious "integer overflow" errors. Also, it was casting the
3974 result into an int without checking.
3976 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3977 did not work. The code looks as if it has *never* worked, though it appears
3978 to have been documented since at least releast 1.62. I have made it work.
3980 35. The "timeout_DNS" error in retry rules, also documented since at least
3981 1.62, also never worked. As it isn't clear exactly what this means, and
3982 clearly it isn't a major issue, I have abolished the feature by treating it
3983 as "timeout", and writing a warning to the main and panic logs.
3985 36. The display of retry rules for -brt wasn't always showing the error code
3988 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3989 timeout_connect_A, timeout_connect_MX.
3991 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3992 to the empty sender.
3994 39. The daemon was not analysing the content of -oX till after it had closed
3995 stderr and disconnected from the controlling terminal. This meant that any
3996 syntax errors were only noted on the panic log, and the return code from
3997 the command was 0. By re-arranging the code a little, I've made the
3998 decoding happen first, so such errors now appear on stderr, and the return
3999 code is 1. However, the actual setting up of the sockets still happens in
4000 the disconnected process, so errors there are still only recorded on the
4003 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4004 connections (as happens on some IP stacks) was logged at start up time as
4005 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4006 it from "IPv6 and IPv4", which means that two separate sockets are being
4009 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4010 says whether AF_INET or AF_INET6 was passed as an argument.
4012 42. Exiwhat output was messed up when time zones were included in log
4015 43. Exiwhat now gives more information about the daemon's listening ports,
4016 and whether -tls-on-connect was used.
4018 44. The "port" option of the smtp transport is now expanded.
4020 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4021 silently ignored. Now an error message is written to the main and panic
4024 46. There's a new ACL modifier called "logwrite" which writes to a log file
4025 as soon as it is encountered.
4027 47. Added $local_user_uid and $local_user_gid at routing time.
4029 48. Exim crashed when trying to verify a sender address that was being
4032 49. Exim was recognizing only a space character after ".include". It now also
4033 recognizes a tab character.
4035 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4036 extracting the relevant information from the specification. The man page no
4037 longer contains scrambled data for the -d option, and I've added a section
4038 at the front about calling Exim under different names.
4040 51. Added "extra_headers" argument to the "mail" command in filter files.
4042 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4045 53. Installed eximstats 1.29.
4047 54. Added transport_filter_timeout as a generic transport option.
4049 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4050 Cc: header lines. This was required by RFC 822, but it not required by RFC
4053 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4054 incoming messages that did not have them. Now it does so only if the
4055 message originates locally, that is, if there is no associated remote host
4056 address. When Resent- header lines are present, this applies to the Resent-
4057 lines rather than the non-Resent- lines.
4059 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4060 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4062 58. Messages for configuration errors now include the name of the main
4063 configuration file - useful now that there may be more than one file in a
4064 list (.included file names were always shown).
4066 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4067 for those rare installations that do not start the daemon as root or run it
4068 setuid root. I've cut out the call to initgroups() if the daemon is not
4071 60. The Exim user and group can now be bound into the binary as text strings
4072 that are looked up at the start of Exim's processing.
4074 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4076 62. Added $mailstore_basename variable.
4078 63. Installed patch to sieve.c from Michael Haardt.
4080 64. When Exim failed to open the panic log after failing to open the main log,
4081 the original message it was trying to log was written to stderr and debug
4082 output, but if they were not available (the usual case in production), it
4083 was lost. Now it is written to syslog before the two lines that record the
4084 failures to open the logs.
4086 65. Users' Exim filters run in subprocesses under the user's uid. It is
4087 possible for a "deliver" command or an alias in a "personal" command to
4088 provoke an address rewrite. If logging of address rewriting is configured,
4089 this fails because the process is not running as root or exim. There may be
4090 a better way of dealing with this, but for the moment (because 4.30 needs
4091 to be released), I have disabled address rewrite logging when running a
4092 filter in a non-root, non-exim process.
4098 1. The buildconfig auxiliary program wasn't quoting the value set for
4099 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4100 not defined. This bug was masked in 4.22 by the effect that was fixed in
4103 2. Some messages that were rejected after a message id was allocated were
4104 shown as "incomplete" by exigrep. It no longer does this for messages that
4105 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4107 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4108 have allow_domain_literals set, the ID did not get logged in the <= line.
4109 Domain literals are now always recognized in Message-ID: header lines.
4111 4. The first argument for a ${extract expansion item is the key name or field
4112 number. Leading and trailing spaces in this item were not being ignored,
4113 causing some misleading effects.
4115 5. When deliver_drop_privilege was set, single queue runner processes started
4116 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4117 same command in the process it spins off) were not dropping privilege.
4119 6. When the daemon running as "exim" started a queue runner, it always
4120 re-executed Exim in the spun-off process. This is a waste of effort when
4121 deliver_drop_privilege is set. The new process now just calls the
4122 queue-runner function directly.
4128 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4131 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4132 ignored. Though the use of -oP was forcing the writing of a pid file, it
4133 was always written to the default place.
4135 3. If the message "no IP address found for host xxxx" is generated during
4136 incoming verification, it is now followed by identification of the incoming
4137 connection (so you can more easily find what provoked it).
4139 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4141 5. Added some features to "harden" Exim a bit more against certain attacks:
4143 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4144 be put in Local/Makefile. This is like the never_users runtime option,
4145 but it cannot be overridden. The default setting is "root".
4147 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4148 prefix string with which any file named in a -C command line option
4151 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4152 is retained for -C and -D only if the caller of Exim is root. Without
4153 it, the exim user may also use -C and -D and retain privilege.
4155 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4156 command line option is disabled.
4158 6. Macro names set by the -D option must start with an upper case letter, just
4159 like macro names defined in the configuration file.
4161 7. Added "dereference=" facility to LDAP.
4163 8. Two instances of the typo "uknown" in the source files are fixed.
4165 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4166 the Configure-Makefile script screwed up while processing it.
4168 10. Incorporated PCRE 4.4.
4170 11. The SMTP synchronization check was not operating right at the start of an
4171 SMTP session. For example, it could not catch a HELO sent before the client
4172 waited for the greeting. There is now a check for outstanding input at the
4173 point when the greeting is written. Because of the duplex, asynchronous
4174 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4175 way, but not yet received, when the check is performed.
4177 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4178 on TCP/IP sockets, because this apparently causes some broken clients to
4181 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4182 unchanged) from the Cygwin maintainer.
4184 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4185 was supported instead of testing for mbx. Effectively a typo.
4187 15. The spa authenticator server code was not checking that the input it
4188 received was valid base64.
4190 16. The debug output line for the "set" modifier in ACLs was not showing the
4191 name of the variable that was being set.
4193 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4195 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4196 lookup on its own. Something else was needed to trigger the lookup. For
4197 example, a match in host_lookup or the need for a host name in a host list.
4198 Now, if $sender_host_name is referenced and the host name has not yet been
4199 looked up, a lookup is performed. If the lookup fails, the variable remains
4200 empty, and $host_lookup_failed is set to "1".
4202 19. Added "eqi" as a case-independent comparison operator.
4204 20. The saslauthd authentication condition could segfault if neither service
4205 nor realm was specified.
4207 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4208 error message that was logged was misleading, and incoming SMTP
4209 connections were dropped. The message is now more accurate, and temporary
4210 errors are given to SMTP connections.
4212 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4213 (except RSET) with a 421 error, until QUIT is received. However, it was
4214 failing to send a response to QUIT.
4216 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4217 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4218 ACL is now run after the helo_try_verify_hosts code.
4220 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4221 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4222 case-independent, so other case variants are also recognized). Apparently
4223 some systems use these upper case variants.
4225 25. If more than two messages were waiting for the same host, and a transport
4226 filter was specified for the transport, Exim sent two messages over the
4227 same TCP/IP connection, and then failed with "socket operation on non-
4228 socket" when it tried to send the third.
4230 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4232 27. The extern definition of crypt16() in expand.c was not being excluded when
4233 the OS had its own crypt16() function.
4235 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4236 as a preferred synonym for return_size_limit, both as an option and as an
4239 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4241 30. Changed the default configuration ACL to relax the local part checking rule
4242 for addresses that are not in any local domains. For these addresses,
4243 slashes and pipe symbols are allowed within local parts, but the sequence
4244 /../ is explicitly forbidden.
4246 31. SPA server authentication was not clearing the challenge buffer before
4249 32. log_message in a "warn" ACL statement was writing to the reject log as
4250 well as to the main log, which contradicts the documentation and doesn't
4251 seem right (because no rejection is happening). So I have stopped it.
4253 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4254 However, I am unable to do any testing of this.
4256 34. Fixed an infelicity in the appendfile transport. When checking directories
4257 for a mailbox, to see if any needed to be created, it was accidentally
4258 using path names with one or more superfluous leading slashes; tracing
4259 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4261 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4262 contents are added to the log line that is written for every discarded
4263 recipient. (Previously a log_message setting was ignored.)
4265 36. The ${quote: operator now quotes the string if it is empty.
4267 37. The install script runs exim in order to find its version number. If for
4268 some reason other than non-existence or emptiness, which it checks, it
4269 could not run './exim', it was installing it with an empty version number,
4270 i.e. as "exim-". This error state is now caught, and the installation is
4273 38. An argument was missing from the function that creates an error message
4274 when Exim fails to connect to the socket for saslauthd authentication.
4275 This could cause Exim to crash, or give a corrupted message.
4277 39. Added isip, isip4, and isip6 to ${if conditions.
4279 40. The ACL variables $acl_xx are now saved with the message, and can be
4280 accessed later in routers, transports, and filters.
4282 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4283 strings in the file are not string-expanded.
4285 42. If a MAIL command specified a SIZE value that was too large to fit into an
4286 int variable, the check against message_size_limit failed. Such values are
4287 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4288 day this will have to be increased, but I don't think I want to be around
4289 when emails are that large.
4296 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4297 iconv() is not standard in FreeBSD.
4299 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4300 IPv6 enabled. The observed symptom was a segmentation fault on return from
4301 the function os_common_find_running_interfaces() in src/os.c.
4303 3. In the check_special_case() function in daemon.c I had used "errno" as an
4304 argument name, which causes warnings on some systems. This was basically a
4305 typo, since it was named "eno" in the comments!
4307 4. The code that waits for the clock to tick (at a resolution of some fraction
4308 of a second) so as to ensure message-id uniqueness was always waiting for
4309 at least one whole tick, when it could have waited for less. [This is
4310 almost certainly not relevant at current processor speeds, where it is
4311 unlikely to ever wait at all. But we try to future-proof.]
4313 5. The function that sleeps for a time interval that includes fractions of a
4314 second contained a race. It did not block SIGALRM between setting the
4315 timer, and suspending (a couple of lines later). If the interval was short
4316 and the sigsuspend() was delayed until after it had expired, the suspension
4317 never ended. On busy systems this could lead to processes getting stuck for
4320 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4321 process, before it forks any delivery processes. The open lookup caching
4322 mechanism meant that the open file or database connection was passed into
4323 the delivery process. The problem was that delivery processes always tidy
4324 up cached lookup data. This could cause a problem for the next delivery
4325 process started by the queue runner, because the external queue runner
4326 process does not know about the closure. So the next delivery process
4327 still has data in the lookup cache. In the case of a file lookup, there was
4328 no problem because closing a file descriptor in a subprocess doesn't affect
4329 the parent. However, if the lookup was caching a connection to a database,
4330 the connection was closed, and the second delivery process was likely to
4331 see errors such as "PGSQL: query failed: server closed the connection
4332 unexpectedly". The problem has been fixed by closing all cached lookups
4333 in a queue runner before running a delivery process.
4335 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4336 seem to have the "const" qualifier which it has on other OS. I've
4339 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4340 *of the same type* (client or server) with the same public name that an
4341 error should be diagnosed.
4343 9. When Exim looked up a host name for an IP address, but failed to find the
4344 original IP address when looking up the host name (a safety check), it
4345 output the message "<ip address> does not match any IP for NULL", which was
4346 confusing, to say the least. The bug was that the host name should have
4347 appeared instead of "NULL".
4349 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4350 user that is built into the binary, and the -C or -D options is used, root
4351 privilege is dropped before the configuration file is read. In addition,
4352 logging is switched to stderr instead of the normal log files. If the
4353 configuration then re-defines the Exim user, the unprivileged environment
4354 is probably not what is expected, so Exim logs a panic warning message (but
4357 However, if deliver_drop_privilege is set, the unprivileged state may well
4358 be exactly what is intended, so the warning has been cut out in that case,
4359 and Exim is allowed to try to write to its normal log files.
4365 1. smtp_return_error_details was not giving details for temporary sender
4366 or receiver verification errors.
4368 2. Diagnose a configuration error if two authenticators have the same public
4371 3. Exim used not to create the message log file for a message until the first
4372 delivery attempt. This could be confusing when incoming messages were held
4373 for policy or load reasons. The message log file is now created at the time
4374 the message is received, and an initial "Received" line is written to it.
4376 4. The automatically generated man page for command line options had a minor
4377 bug that caused no ill effects; however, a more serious problem was that
4378 the procedure for building the man page automatically didn't always
4379 operate. Consequently, release 4.20 contains an out-of-date version. This
4380 shouldn't happen again.
4382 5. When building Exim with embedded Perl support, the script that builds the
4383 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4384 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4386 6. The freeze_tell option was not being used for messages that were frozen on
4387 arrival, either by an ACL or by local_scan().
4389 7. Added the smtp_incomplete_transaction log selector.
4391 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4392 was accepting AUTH without a new EHLO.
4394 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4395 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4398 10. Logging of TCP/IP connections (when configured) now happens in the main
4399 daemon process instead of the child process, so that the TCP/IP connection
4400 count is more accurate (but it can never be perfect).
4402 11. The use of "drop" in a nested ACL was not being handled correctly in the
4403 outer ACL. Now, if condition failure induced by the nested "drop" causes
4404 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4405 or "require"), the connection is dropped.
4407 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4408 that yield "discard" can now be used with an "accept" or a "discard" verb,
4409 but an error is generated for any others (because I can't see a useful way
4410 to define what should happen).
4412 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4413 are now processed in the same way as lines in the Exim configuration file.
4414 In particular, continuation lines are supported.
4416 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4418 15. Added -ti meaning -t -i.
4420 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4421 domains, and warn by logging if others are found.
4423 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4424 returned from GIFCONF when Exim is trying to find the list of interfaces on
4425 a host. The code in os.c has been modified to copy each ifreq to an aligned
4426 structure in all cases.
4428 Also, in some cases, the returned ifreq's were being copied to a 'struct
4429 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4430 means the last couple of bytes of an IPv6 address could be chopped if the
4431 ifreq contained only a normal sockaddr (14 bytes storage).
4433 18. Named domain lists were not supported in the hosts_treat_as_local option.
4434 An entry such as +xxxx was not recognized, and was treated as a literal
4437 19. Ensure that header lines added by a DATA ACL are included in the reject log
4438 if the ACL subsequently rejects the message.
4440 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4441 MD5 (which is deprecated).
4443 21. When testing a filter file using -bf, Exim was writing a message when it
4444 took the sender from a "From " line in the message, but it was not doing so
4445 when it took $return_path from a Return-Path: header line. It now does.
4447 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4448 with a valid header line field name (a series of printing characters
4449 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4451 23. Changed "disc" in the source to "disk" to conform to the documentation and
4452 the book and for uniformity.
4454 24. Ignore Sendmail's -Ooption=value command line item.
4456 25. When execve() failed while trying to run a command in a pipe transport,
4457 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4458 could be confused with a return value of 69 from the command itself. This
4459 has been changed to 127, the value the shell returns if it is asked to run
4460 a non-existent command. The wording for the related log line suggests a
4461 non-existent command as the problem.
4463 26. If received_header_text expands to an empty string, do not add a Received:
4464 header line to the message. (Well, it adds a token one on the spool, but
4465 marks it "old" so that it doesn't get used or transmitted.)
4467 27. Installed eximstats 1.28 (addition of -nt option).
4469 28. There was no check for failure on the call to getsockname() in the daemon
4470 code. This can fail if there is a shortage of resources on the system, with
4471 ENOMEM, for example. A temporary error is now given on failure.
4473 29. Contrary to the C standard, it seems that in some environments, the
4474 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4475 program. Exim now does this explicitly; it affects the formatting of
4476 timestamps using strftime().
4478 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4479 complaints. I've now initialized all the variables, to avoid this.
4481 32. Header lines added by a system filter were not being "seen" during
4482 transport-time rewrites.
4484 33. The info_callback() function passed to OpenSSL is set up with type void
4485 (*)(SSL *, int, int), as described somewhere. However, when calling the
4486 function (actually a macro) that sets it up, the type void(*)() is
4487 expected. I've put in a cast to prevent warnings from picky compilers.
4489 34. If a DNS black list lookup found a CNAME record, but there were no A
4490 records associated with the domain it pointed at, Exim crashed.
4492 35. If a DNS black list lookup returned more than one A record, Exim ignored
4493 all but the first. It now scans all returned addresses if a particular IP
4494 value is being sought. In this situation, the contents of the
4495 $dnslist_value variable are a list of all the addresses, separated by a
4498 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4499 to accept a host name and all its aliases if the forward lookup for any of
4500 them yielded the IP address of the incoming connection. Now it accepts only
4501 those names whose forward lookup yields the correct IP address. Any other
4502 names are discarded. This closes a loophole whereby a rogue DNS
4503 administrator could create reverse DNS records to break through a
4504 wildcarded host restriction in an ACL.
4506 37. If a user filter or a system filter that ran in a subprocess used any of
4507 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4508 the wrong values were passed to the pipe command ($thisaddress had the
4509 value of $0, $0 had the value of $1, etc). This bug was introduced by
4510 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4512 38. Improved the line breaking for long SMTP error messages from ACLs.
4513 Previously, if there was no break point between 40 and 75 characters, Exim
4514 left the rest of the message alone. Two changes have been made: (a) I've
4515 reduced the minimum length to 35 characters; (b) if it can't find a break
4516 point between 35 and 75 characters, it looks ahead and uses the first one
4517 that it finds. This may give the occasional overlong line, but at least the
4518 remaining text gets split now.
4520 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4521 file descriptors might be low, and that setting 1000 would always raise it.
4522 It turns out that in some environments, the limit is already over 1000 and
4523 that lowering it causes trouble. So now Exim takes care not to decrease it.
4525 40. When delivering a message, the value of $return_path is set to $sender_
4526 address at the start of routing (routers may change the value). By an
4527 oversight, this default was not being set up when an address was tested by
4528 -bt or -bv, which affected the outcome if any router or filter referred to
4531 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4532 log only when "message" or "log_message" are set. However, if one of the
4533 conditions was an address verification, or a call to a nested ACL, the
4534 messages generated by the underlying test were being passed through. This
4535 no longer happens. The underlying message is available in $acl_verify_
4536 message for both "message" and "log_message" expansions, so it can be
4537 passed through if needed.
4539 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4540 new expansion $bh_ to give the encoded byte string without charset
4541 translation. Translation happens only if iconv() is available; HAVE_ICONV
4542 indicates this at build time. HEADERS_CHARSET gives the charset to
4543 translate to; headers_charset can change it in the configuration, and
4544 "headers charset" can change it in an individual filter file.
4546 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4547 that creates RFC 2047 encoded "words" labels them as that charset instead
4548 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4549 expansion operator; (ii) when Exim creates a From: line for a local
4550 message; (iii) when a header line is rewritten to include a "phrase" part.
4552 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4553 buggy, causing it to skip the first lines of messages whose message ID
4554 ended in 'D'. This would not have bitten before Exim release 4.14, because
4555 message IDs were unlikely to end in 'D' before then. The effect was to have
4556 incorrect size information for certain domains.
4558 45. #include "config.h" was missing at the start of the crypt16.c module. This
4559 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4562 46. If there was a timeout during a "random" callout check, Exim treated it as
4563 a failure of the random address, and carried on sending RSET and the real
4564 address. If the delay was just some slowness somewhere, the response to the
4565 original RCPT would be taken as a response to RSET and so on, causing
4566 mayhem of various kinds.
4568 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4569 when I implemented it. It didn't allow for the fact that some option values
4570 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4571 the right test for positive values.
4573 48. Domain names in DNS records are case-independent. Exim always looks them up
4574 in lower case. Some resolvers return domain names in exactly the case they
4575 appear in the zone file, that is, they may contain uppercase letters. Not
4576 all resolvers do this - some return always lower case. Exim was treating a
4577 change of case by a resolver as a change of domain, similar to a widening
4578 of a domain abbreviation. This triggered its re-routing code and so it was
4579 trying to route what was effectively the same domain again. This normally
4580 caused routing to fail (because the router wouldn't handle the domain
4581 twice). Now Exim checks for this case specially, and just changes the
4582 casing of the domain that it ultimately uses when it transmits the message
4585 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4588 50. If a filter generated a file delivery with a non-absolute name (possible if
4589 no home directory exists for the router), the forbid_file option was not
4592 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4593 the existing equality matching.
4595 52. Exim was using ints instead of ino_t variables in some places where it was
4596 dealing with inode numbers.
4598 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4599 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4600 called TMPDIR, and if it finds it is different, it changes its value.
4602 54. The smtp_printf() function is now made available to local_scan() so
4603 additional output lines can be written before returning. There is also an
4604 smtp_fflush() function to enable the detection of a dropped connection.
4605 The variables smtp_input and smtp_batched_input are exported to
4608 55. Changed the default runtime configuration: the message "Unknown user"
4609 has been removed from the ACL, and instead placed on the localuser router,
4610 using the cannot_route_message feature. This means that any verification
4611 failures that generate their own messages won't get overridden. Similarly,
4612 the "Unrouteable address" message that was in the ACL for unverifiable
4613 relay addresses has also been removed.
4615 56. Added hosts_avoid_esmtp to the smtp transport.
4617 57. The exicyclog script was not checking for the esoteric option
4618 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4619 will work only if exicyclog is run under the appropriate euid.
4621 58. Following a discussion on the list, the rules by which Exim recognises line
4622 endings on incoming messages have been changed. The -dropcr and drop_cr
4623 options are now no-ops, retained only for backwards compatibility. The
4624 following line terminators are recognized: LF CRLF CR. However, special
4625 processing applies to CR:
4627 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4628 nor a local message in the state where . is a terminator.
4630 (ii) If a bare CR is encountered in a header line, an extra space is added
4631 after the line terminator so as not to end the header. The reasoning
4632 behind this is that bare CRs in header lines are most likely either
4633 to be mistakes, or people trying to play silly games.
4635 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4636 was being incorrectly given as 18 bytes larger than it should have been.
4637 This is a VOB (very old bug).
4639 60. This may never have affected anything current, but just in case it has:
4640 When the local host is found other than at the start of a list of hosts,
4641 the local host, those with the same MX, and any that follow, are discarded.
4642 When the list in question was part of a longer list of hosts, the following
4643 hosts (not currently being processed) were also being discarded. This no
4644 longer happens. I'm not sure if this situation could ever has previously
4647 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4650 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4651 header. This is something that is recommended in a new Internet Draft, and
4652 is something that is documented as being done by Sendmail. There are two
4653 possible values. For messages generated by the autoreply transport, Exim
4656 Auto-Submitted: auto-replied
4658 whereas for all other generated messages (e.g. bounces) it adds
4660 Auto-Submitted: auto-generated
4662 63. The "personal" condition in filters now includes a test for the
4663 Auto-Submitted: header. If it contains the string "auto-" the message it
4664 not considered personal.
4666 64. Added rcpt_include_affixes as a generic transport option.
4668 65. Added queue_only_override (default true).
4670 66. Added the syslog_duplication option.
4672 67. If what should have been the first header line of a message consisted of
4673 a space followed by a colon, Exim was mis-interpreting it as a header line.
4674 It isn't of course - it is syntactically invalid and should therefore be
4675 treated as the start of the message body. The misbehaviour could have
4676 caused a number of strange effects, including loss of data in subsequent
4677 header lines, and spool format errors.
4679 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4680 client host had authenticated. This control can now be exercised by an ACL
4681 for more flexibility.
4683 69. By default, callouts do not happen when testing with -bh. There is now a
4684 variant, -bhc, which does actually run the callout code, including
4685 consulting and updating the callout cache.
4687 70. Added support for saslauthd authentication, courtesy of Alexander
4690 71. If statvfs() failed on the spool or log directories while checking their
4691 size for availability, Exim confusingly gave the error "space shortage".
4692 Furthermore, in debugging mode it crashed with a floating point exception.
4693 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4694 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4695 serious problem, Exim now writes to the main and panic logs when this
4696 happens, with details of the failure. It then refuses to accept the
4697 incoming message, giving the message "spool directory problem" or "log
4698 directory problem" with a 421 code for SMTP messages.
4700 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4701 0, 1, and 2 exist, because some OS complain for execs without them (see
4702 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4703 descriptors. However, the code omitted to check that the open succeeded,
4704 causing mysterious errors if for some reason the permissions on /dev/null
4705 got screwed. Now Exim writes a message to the main and panic logs, and
4706 bombs out if it can't open /dev/null.
4708 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4709 interact so that it is all more flexible. It is supposed to remain
4710 backwards compatible. Also added extra_local_interfaces.
4712 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4713 to bomb out with an assertion failure - to the client this appears as a
4714 connection drop. This problem occurs in the part of the code that was taken
4715 from the Samba project. Fortunately, the assertion is in a very simple
4716 function, so I have fixed this by reproducing the function inline in the
4717 one place where it is called, and arranging for authentication to fail
4718 instead of killing the process with assert().
4720 75. The SPA client code was not working when the server requested OEM rather
4721 than Unicode encoding.
4723 76. Added code to make require_files with a specific uid setting more usable in
4724 the case where statting the file as root fails - usually a non-root-mounted
4725 NFS file system. When this happens and the failure is EACCES, Exim now
4726 forks a subprocess and does the per-uid checking as the relevant uid.
4728 77. Added process_log_path.
4730 78. If log_file_path was not explicitly set, a setting of check_log_space or
4731 check_log_inodes was ignored.
4733 79. If a space check for the spool or log partitions fails, the incident is now
4734 logged. Of course, in the latter case the data may get lost...
4736 80. Added the %p formatting code to string_format() so that it can be used to
4737 print addresses in debug_print(). Adjusted all the address printing in the
4738 debugging in store.c to use %p rather than %d.
4740 81. There was a concern that a line of code in smtp_in.c could overflow a
4741 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4742 initially expressed, the concern was not well-founded, because trailing
4743 spaces are removed early. However, if the trailing spaces were followed by
4744 a NULL, they did not get removed, so the overflow was possible. Two fixes
4747 (a) I re-wrote the offending code in a cleaner fashion.
4748 (b) If an incoming SMTP command contains a NULL character, it is rejected
4751 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4752 runs initgroups(), so that if the Exim user is in any additional groups,
4753 they will be used during message reception.
4759 The change log for 4.20 and earlier releases has been archived.