1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.496 2007/03/14 12:15:56 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
10 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
13 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
16 PH/02 In an ACL statement such as
18 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
20 if a client was not listed at all, or was listed with a value other than
21 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
22 the condition was not true (as it should be), so access was not denied.
23 The bug was that the ! inversion was incorrectly passed on to the second
24 item. This has been fixed.
26 PH/03 Added additional dnslists conditions == and =& which are different from
27 = and & when the dns lookup returns more than one IP address.
29 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
30 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
32 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
33 FSYNC, which compiles an option called disable_fsync that allows for
34 bypassing fsync(). The documentation is heavily laced with warnings.
36 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
38 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
39 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
40 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
41 including adding "make clean"; (3) Added -fPIC when compiling the test
42 dynamically loaded module, to get rid of a warning.
44 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
45 message fails, move_frozen_messages = true and ignore_bounce_errors_after
46 = 0s. The bug is otherwise harmless.
48 PH/07 There was a bug in the dovecot authenticator such that the value of
49 $auth1 could be overwritten, and so not correctly preserved, after a
50 successful authentication. This usually meant that the value preserved by
51 the server_setid option was incorrect.
53 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
55 PH/09 Installed PCRE release 7.0.
57 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
58 run for batched SMTP input. It is now run at the start of every message
59 in the batch. While fixing this I discovered that the process information
60 (output by running exiwhat) was not always getting set for -bs and -bS
61 input. This is fixed, and it now also says "batched" for BSMTP.
63 PH/11 Added control=no_pipelining.
65 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
66 patch, slightly modified), and move the expansion of helo_data till after
67 the connection is made in the smtp transport (so it can use these
70 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
72 PH/14 Added log_selector = +pid.
74 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
76 PH/16 Add ${if forany and ${if forall.
78 PH/17 Added dsn_from option to vary the From: line in DSNs.
80 PH/18 Flush SMTP output before performing a callout, unless control =
81 no_callout_flush is set.
83 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
84 was true (the default) a successful delivery failed to delete the retry
85 item, thus causing premature timeout of the address. The bug is now
88 PH/20 Added hosts_avoid_pipelining to the smtp transport.
90 PH/21 Long custom messages for fakedefer and fakereject are now split up
91 into multiline reponses in the same way that messages for "deny" and
92 other ACL rejections are.
94 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
95 with slight modification.
97 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
98 draft, changing the syntax and factoring some duplicate code".
100 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
101 for deliveries of the second and subsequent messages over the same SMTP
104 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
105 ${reduce, with only minor "tidies".
107 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
109 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
110 expansion side effects.
112 PH/27 When a message times out after an over-quota error from an Exim-imposed
113 quota, the bounce message says "mailbox is full". This message was not
114 being given when it was a system quota that was exceeded. It now should
117 MH/03 Made $recipients available in local_scan(). local_scan() already has
118 better access to the recipient list through recipients_list[], but
119 $recipients can be useful in postmaster-provided expansion strings.
121 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
122 in the case of a MAIL command with additional options following the
123 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
124 were accidentally chopped off.
126 PH/29 SMTP synchronization checks are implemented when a command is read -
127 there is a check that no more input is waiting when there shouldn't be
128 any. However, for some commands, a delay in an ACL can mean that it is
129 some time before the response is written. In this time, more input might
130 arrive, invalidly. So now there are extra checks after an ACL has run for
131 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
132 pipelining has not been advertised.
134 PH/30 MH's patch to allow iscntrl() characters to be list separators.
136 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
137 returned in the SMTP response when smtp_return_error_details was false.
140 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
141 instead of the C I/O that was originally supplied, because problems were
144 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
145 Exim which did not show up earlier: it was assuming that a call to
146 SSL_CTX_set_info_callback() might give an error value. In fact, there is
147 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
148 was a macro that became an assignment, so it seemed to work. This has
149 changed to a proper function call with a void return, hence the compile
150 error. Exim's code has been fixed.
152 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
155 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
157 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
159 PH/37 If a message is not accepted after it has had an id assigned (e.g.
160 because it turns out to be too big or there is a timeout) there is no
161 "Completed" line in the log. When some messages of this type were
162 selected by exigrep, they were listed as "not completed". Others were
163 picked up by some special patterns. I have improved the selection
164 criteria to be more general.
166 PH/38 The host_find_failed option in the manualroute router can now be set
167 to "ignore", to completely ignore a host whose IP address cannot be
168 found. If all hosts are ignored, the behaviour is controlled by the new
169 host_all_ignored option.
171 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
172 homing or because of multiple MX records with /mx) generated more than
173 one IP address, and the following item turned out to be the local host,
174 all the secondary addresses of the first item were incorrectly removed
175 from the list, along with the local host and any following hosts (which
176 is what is supposed to happen).
178 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
179 whoever called Exim into the -H file. In the case of the daemon it was
180 behaving confusingly. When first started, it used values for whoever
181 started the daemon, but after a SIGHUP it used the Exim user (because it
182 calls itself on a restart). I have changed the code so that it now always
189 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
190 fixed by 4.65/MH/01 (is this a record?) are fixed:
192 (i) An empty string was always treated as zero by the numeric comparison
193 operators. This behaviour has been restored.
195 (ii) It is documented that the numeric comparison operators always treat
196 their arguments as decimal numbers. This was broken in that numbers
197 starting with 0 were being interpreted as octal.
199 While fixing these problems I realized that there was another issue that
200 hadn't been noticed. Values of message_size_limit (both the global option
201 and the transport option) were treated as octal if they started with 0.
202 The documentation was vague. These values are now always treated as
203 decimal, and I will make that clear in the documentation.
209 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
210 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
213 MH/01 Don't check that the operands of numeric comparison operators are
214 integers when their expansion is in "skipping" mode (fixes bug
215 introduced by 4.64-PH/07).
217 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
218 child addresses, Exim now panics and dies. Previously, because the count
219 is held in a short int, deliveries were likely to be lost. As such a
220 large number of recipients for a single message is ridiculous
221 (performance will be very, very poor), I have chosen to impose a limit
222 rather than extend the field.
228 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
229 leftover -K file (the existence of which was triggered by #402).
230 While we were at it, introduced process PID as part of the -K
231 filename. This should rule out race conditions when creating
234 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
235 processing considerably. Previous code took too long for large mails,
236 triggering a timeout which in turn triggers #401.
238 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
239 in the DK code in transports.c. sendfile() is not really portable,
240 hence the _LINUX specificness.
242 TF/01 In the add_headers option to the mail command in an Exim filter,
243 there was a bug that Exim would claim a syntax error in any
244 header after the first one which had an odd number of characters
247 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
248 callout verification, Exim cached a "reject" for the entire domain. This
249 is correct for most verifications, but it is not correct for a recipient
250 verification with use_sender or use_postmaster set, because in that case
251 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
252 case of MAIL FROM:<> rejection from other early rejections (e.g.
253 rejection of HELO). When verifying a recipient using a non-null MAIL
254 address, the cache is ignored if it shows MAIL FROM:<> rejection.
255 Whatever the result of the callout, the value of the domain cache is
256 left unchanged (for any other kind of callout, getting as far as trying
257 RCPT means that the domain itself is ok).
259 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
262 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
263 manifest itself as EPIPE rather than ECONNECT. When tidying away a
264 session, the daemon ignores ECONNECT errors and logs others; it now
265 ignores EPIPE as well.
267 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
268 (quoted-printable decoding).
270 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
271 later the small subsequent patch to fix an introduced bug.
273 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
275 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
277 PH/08 An error is now given if message_size_limit is specified negative.
279 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
280 to be given (somewhat) arbitrary names.
282 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
285 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
286 miscellaneous code fixes
288 PH/10 Added the log_reject_target ACL modifier to specify where to log
291 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
292 hostname. This is wrong, because it relates to the incoming message (and
293 probably the interface on which it is arriving) and not to the outgoing
294 callout (which could be using a different interface). This has been
295 changed to use the value of the helo_data option from the smtp transport
296 instead - this is what is used when a message is actually being sent. If
297 there is no remote transport (possible with a router that sets up host
298 addresses), $smtp_active_hostname is used.
300 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
301 tweaks were necessary in order to get it to work (see also 21 below):
302 (a) The code assumed that strncpy() returns a negative number on buffer
303 overflow, which isn't the case. Replaced with Exim's string_format()
305 (b) There were several signed/unsigned issues. I just did the minimum
306 hacking in of casts. There is scope for a larger refactoring.
307 (c) The code used strcasecmp() which is not a standard C function.
308 Replaced with Exim's strcmpic() function.
309 (d) The code set only $1; it now sets $auth1 as well.
310 (e) A simple test gave the error "authentication client didn't specify
311 service in request". It would seem that Dovecot has changed its
312 interface. Fortunately there's a specification; I followed it and
313 changed what the client sends and it appears to be working now.
315 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
318 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
319 address A is aliased to B and C, where B exists and C does not. Without
320 -v the output is "A verified" because verification stops after a
321 successful redirection if more than one address is generated. However,
322 with -v the child addresses are also verified. Exim was outputting "A
323 failed to verify" and then showing the successful verification for C,
324 with its parentage. It now outputs "B failed to verify", showing B's
325 parentage before showing the successful verification of C.
327 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
328 look up a TXT record in a specific list after matching in a combined
331 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
332 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
333 they consult the DNS. I had assumed they would set it the way they
334 wanted; and indeed my experiments on Linux seem to show that in some
335 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
336 To be on the safe side, however, I have now made the interface to
337 host_find_byname() similar to host_find_bydns(), with an argument
338 containing the DNS resolver options. The host_find_byname() function now
339 sets these options at its start, just as host_find_bydns() does. The smtp
340 transport options dns_qualify_single and dns_search_parents are passed to
341 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
342 of host_find_byname() use the default settings of RES_DEFNAMES
343 (qualify_single) but not RES_DNSRCH (search_parents).
345 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
346 spool_read_header() do less string testing, by means of a preliminary
347 switch on the second character of optional "-foo" lines. (This is
348 overdue, caused by the large number of possibilities that now exist.
349 Originally there were few.) While I was there, I also converted the
350 str(n)cmp tests so they don't re-test the leading "-" and the first
351 character, in the hope this might squeeze out yet more improvement.
353 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
354 flag allowing group syntax was set by the header_syntax check but not
355 turned off, possible causing trouble later; (2) The flag was not being
356 set at all for the header_verify test, causing "group"-style headers to
357 be rejected. I have now set it in this case, and also caused header_
358 verify to ignore an empty address taken from a group. While doing this, I
359 came across some other cases where the code for allowing group syntax
360 while scanning a header line wasn't quite right (mostly, not resetting
361 the flag correctly in the right place). These bugs could have caused
362 trouble for malformed header lines. I hope it is now all correct.
364 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
365 with the "reply" argument non-NULL. The code, however (which originally
366 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
367 but it didn't always do it. This confused somebody who was copying the
368 code for some other use. I have removed all the tests.
370 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
371 feature that was used to support insecure browsers during the U.S. crypto
372 embargo. It requires special client support, and Exim is probably the
373 only MTA that supported it -- and would never use it because real RSA is
374 always available. This code has been removed, because it had the bad
375 effect of slowing Exim down by computing (never used) parameters for the
376 RSA_EXPORT functionality.
378 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
379 authenticator to fail if there's a tab character in the incoming data
380 (there should never be unless someone is messing about, as it's supposed
381 to be base64-encoded). Also added, on Timo's advice, the "secured" option
382 if the connection is using TLS or if the remote IP is the same as the
383 local IP, and the "valid-client-cert option" if a client certificate has
386 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
387 authenticators. This can be used for authorization after authentication
388 succeeds. (In the case of plaintext, it servers for both authentication
391 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
392 if any retry times were supplied.
394 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
395 connection, where there is no client IP address to check. In this
396 situation, the verify now always succeeds.
398 PH/25 Applied John Jetmore's -Mset patch.
400 PH/26 Added -bem to be like -Mset, but loading a message from a file.
402 PH/27 In a string expansion for a processed (not raw) header when multiple
403 headers of the same name were present, leading whitespace was being
404 removed from all of them, but trailing whitespace was being removed only
405 from the last one. Now trailing whitespace is removed from each header
406 before concatenation. Completely empty headers in a concatenation (as
409 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
410 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
412 PH/29 [Removed. This was a change that I later backed out, and forgot to
413 correct the ChangeLog entry (that I had efficiently created) before
414 committing the later change.]
416 PH/30 Exim was sometimes attempting to deliver messages that had suffered
417 address errors (4xx response to RCPT) over the same connection as other
418 messages routed to the same hosts. Such deliveries are always "forced",
419 so retry times are not inspected. This resulted in far too many retries
420 for the affected addresses. The effect occurred only when there were more
421 hosts than the hosts_max_try setting in the smtp transport when it had
422 the 4xx errors. Those hosts that it had tried were not added to the list
423 of hosts for which the message was waiting, so if all were tried, there
424 was no problem. Two fixes have been applied:
426 (i) If there are any address or message errors in an SMTP delivery, none
427 of the hosts (tried or untried) are now added to the list of hosts
428 for which the message is waiting, so the message should not be a
429 candidate for sending over the same connection that was used for a
430 successful delivery of some other message. This seems entirely
431 reasonable: after all the message is NOT "waiting for some host".
432 This is so "obvious" that I'm not sure why it wasn't done
433 previously. Hope I haven't missed anything, but it can't do any
434 harm, as the worst effect is to miss an optimization.
436 (ii) If, despite (i), such a delivery is accidentally attempted, the
437 routing retry time is respected, so at least it doesn't keep
438 hammering the server.
440 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
441 in ${readsocket because some servers need this prod.
443 PH/32 Added some extra debug output when updating a wait-xxx database.
445 PH/33 The hint "could be header name not terminated by colon", which has been
446 given for certain expansion errors for a long time, was not being given
447 for the ${if def:h_colon_omitted{... case.
449 PH/34 The spec says: "With one important exception, whenever a domain list is
450 being scanned, $domain contains the subject domain." There was at least
451 one case where this was not true.
453 PH/35 The error "getsockname() failed: connection reset by peer" was being
454 written to the panic log as well as the main log, but it isn't really
455 panic-worthy as it just means the connection died rather early on. I have
456 removed the panic log writing for the ECONNRESET error when getsockname()
459 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
460 runs only) independently of the message's sender address. This meant
461 that, if the 4xx error was in fact related to the sender, a different
462 message to the same recipient with a different sender could confuse
463 things. In particualar, this can happen when sending to a greylisting
464 server, but other circumstances could also provoke similar problems.
465 I have changed the default so that the retry time for these errors is now
466 based a combination of the sender and recipient addresses. This change
467 can be overridden by setting address_retry_include_sender=false in the
470 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
471 remote server are returned as part of bounce messages. This was not
472 happening for LMTP over a pipe (the lmtp transport), but now it is the
473 same for both kinds of LMTP.
475 PH/38 Despite being documented as not happening, Exim was rewriting addresses
476 in header lines that were in fact CNAMEs. This is no longer the case.
478 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
479 and queue runs started by the daemon processed all messages. This has
480 been fixed so that -R and -S can now usefully be given with -q<time>.
482 PH/40 Import PCRE release 6.7 (fixes some bugs).
484 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
486 PH/42 Give an error if -q is specified more than once.
488 PH/43 Renamed the variables $interface_address and $interface_port as
489 $received_ip_address and $received_port, to make it clear that these
490 values apply to message reception, and not to the outgoing interface when
491 a message is delivered. (The old names remain recognized, of course.)
493 PH/44 There was no timeout on the connect() call when using a Unix domain
494 socket in the ${readsocket expansion. There now is.
496 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
497 be meaningful with "accept".
499 SC/01 Eximstats V1.43
500 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
502 SC/02 Eximstats V1.44
503 Use a glob alias rather than an array ref in the generated
504 parser. This improves both readability and performance.
506 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
507 Collect SpamAssassin and rejection statistics.
508 Don't display local sender or destination tables unless
509 there is data to show.
510 Added average volumes into the top table text output.
512 SC/04 Eximstats V1.46
513 Collect data on the number of addresses (recipients)
514 as well as the number of messages.
516 SC/05 Eximstats V1.47
517 Added 'Message too big' to the list of mail rejection
518 reasons (thanks to Marco Gaiarin).
520 SC/06 Eximstats V1.48
521 Mainlog lines which have GMT offsets and are too short to
522 have a flag are now skipped.
524 SC/07 Eximstats V1.49 (Alain Williams)
525 Added the -emptyok flag.
527 SC/08 Eximstats V1.50
528 Fixes for obtaining the IP address from reject messages.
530 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
531 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
532 whitesspace changes from 4.64-PH/27
534 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
537 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
538 are found, allow negative numbers in numeric criteria)
540 JJ/06 exipick.20061117.2, added new $message_body_missing variable
542 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
543 to match changes made in 4.64-PH/43
545 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
547 PH/47 Put in an explicit test for a DNS lookup of an address record where the
548 "domain" is actually an IP address, and force a failure. This locks out
549 those revolvers/nameservers that support "A-for-A" lookups, in
550 contravention of the specifications.
552 PH/48 When a host name was looked up from an IP address, and the subsequent
553 forward lookup of the name timed out, the host name was left in
554 $sender_host_name, contrary to the specification.
556 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
557 restricted to single-key lookups, Exim was not diagnosing an error if
558 * or *@ was used with a query-style lookup.
560 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
562 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
563 long ago, but noone interested enough thought of it. Let's just say that
564 the "1.1" means that there are some new functions that weren't there at
565 some point in the past.
567 PH/51 Error processing for expansion failure of helo_data from an smtp
568 transport during callout processing was broken.
570 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
571 tested/used via the -bh/-bhc/-bs options.
573 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
574 bug, fixed in subsequent PCRE releases).
576 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
577 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
579 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
585 SC/01 Use a glob alias rather than an array ref in eximstats generated
586 parser. This improves both readability and performance.
588 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
589 Don't display local sender or destination tables in eximstats unless
590 there is data to show.
591 Added average volumes into the eximstats top table text output.
593 SC/03 Collect data on the number of addresses (recipients) as well
594 as the number of messages in eximstats.
596 TF/01 Correct an error in the documentation for the redirect router. Exim
597 does (usually) call initgroups() when daemonizing.
599 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
600 with consistent privilege compared to when running as a daemon.
602 TF/03 Note in the spec that $authenticated_id is not set for local
603 submissions from trusted users.
605 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
606 Thanks to Dean Brooks <dean@iglou.com> for the patch.
608 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
609 by adding some example configuration directives to the default
610 configuration file. A little bit of work is required to uncomment the
611 directives and define how usernames and passwords are checked, but
612 there is now a framework to start from.
614 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
615 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
616 without this. I don't know how relevant this is to other LDAP libraries.
618 PH/02 Add the verb name to the "unknown ACL verb" error.
620 PH/03 Magnus Holmgren's patch for filter_prepend_home.
622 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
624 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
625 directory not expanded when it should be if an expanded home directory
626 was set for the address (which is overridden by the transport).
628 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
631 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
632 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
633 because it is too late at that time, and has no effect.
635 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
636 security issue with \' (bugzilla #107). I could not use the
637 PQescapeStringConn() function, because it needs a PGconn value as one of
640 PH/08 When testing addresses using -bt, indicate those final addresses that
641 are duplicates that would not cause an additional delivery. At least one
642 person was confused, thinking that -bt output corresponded to deliveries.
643 (Suppressing duplicates isn't a good idea as you lose the information
644 about possibly different redirections that led to the duplicates.)
646 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
647 systems where poll() doesn't work, in particular OS X.
649 PH/10 Added more information to debugging output for retry time not reached.
651 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
652 operations in malware.c.
654 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
657 PH/13 If write_rejectlog was set false when logging was sent to syslog with
658 syslog_duplication set false, log lines that would normally be written
659 both the the main log and to the reject log were not written to syslog at
662 PH/14 In the default configuration, change the use of "message" in ACL warn
663 statements to "add_header".
665 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
666 not followed by a command (e.g. "seen endif").
668 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
669 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
672 PH/17 Added extra conditions to the default value of delay_warning_condition
676 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
677 { match{$h_precedence:}{(?i)bulk|list|junk} } \
678 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
681 The Auto-Submitted: and various List- headers are standardised, whereas I
682 don't think Precedence: ever was.
684 PH/18 Refactored debugging code in route_finduser() to show more information,
685 in particular, the error code if getpwnam() issues one.
687 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
688 This is apparently needed in addition to the PH/07 change above to avoid
689 any possible encoding problems.
691 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
692 but not after initializing Perl.
694 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
695 output them only if debugging. By default they are written stderr,
696 apparently, which is not desirable.
698 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
701 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
704 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
706 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
707 authenticated or an ident call has been made. Suppress the default
708 values for $authenticated_id and $authenticated_sender (but permit -oMai
709 and -oMas) when testing with -bh.
711 PH/24 Re-jigged the order of the tests in the default configuration so that the
712 tests for valid domains and recipients precede the DNS black list and CSA
713 tests, on the grounds that those ones are more expensive.
715 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
716 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
717 This bug exists in every version of Exim that I still have, right back to
720 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
721 However, an attempt to turn on case-sensitivity in a regex key by
722 including (?-i) didn't work because the subject string was already
723 lowercased, and the effects were non-intuitive. It turns out that a
724 one-line patch can be used to allow (?-i) to work as expected.
730 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
731 other effects) broke the use of negated acl sub-conditions.
733 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
736 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
737 "Deny" causes Exim to reject the incoming connection with a 554 error.
738 Unfortunately, if there is a major crisis, such as a disk failure,
739 tcp-wrappers gives "deny", whereas what one would like would be some
740 kind of temporary error. A kludge has been added to help with this.
741 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
742 554 error is used if errno is still zero or contains ENOENT (which occurs
743 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
746 PH/03 Add -lutil to the default FreeBSD LIBS setting.
748 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
749 errors. Otherwise a message that provokes a temporary error (when other
750 messages do not) can cause a whole host to time out.
752 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
753 addresses were routed directly to files or pipes from a redirect router.
754 File deliveries just didn't batch; pipe deliveries might have suffered
757 PH/06 A failure to get a lock for a hints database would erroneously always say
758 "Failed to get write lock", even when it was really a read lock.
760 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
761 of 0600. This has been changed to use the value of the lockfile_mode
762 option (which defaults to 0600).
764 PH/08 Applied small patch from the Sieve maintainer.
766 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
767 folder from quota calculations, a direct delivery into this folder messed
768 up the contents of the maildirsize file. This was because the regex was
769 used only to exclude .Trash (or whatever) when the size of the mailbox
770 was calculated. There was no check that a delivery was happening into an
771 excluded directory. This bug has been fixed by ignoring all quota
772 processing for deliveries into excluded directories.
774 PH/10 Added the maildirfolder_create_regex option to appendfile.
780 PH/01 The code for finding all the local interface addresses on a FreeBSD
781 system running IPv6 was broken. This may well have applied to all BSD
782 systems, as well as to others that have similar system calls. The broken
783 code found IPv4 interfaces correctly, but gave incorrect values for the
784 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
785 that it would not match correctly against @[] and not recognize the IPv6
788 PH/02 The ipliteral router was not recognizing addresses of the form user@
789 [ipv6:....] because it didn't know about the "ipv6:" prefix.
791 PH/03 Added disable_ipv6.
793 PH/04 Changed $reply_address to use the raw form of the headers instead of the
794 decoded form, because it is most often used to construct To: headers
795 lines in autoreplies, and the decoded form may well be syntactically
796 invalid. However, $reply_address has leading white space removed, and all
797 newlines turned into spaces so that the autoreply transport does not
800 PH/05 If group was specified without a user on a router, and no group or user
801 was specified on a transport, the group from the router was ignored.
803 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
804 for visible compile-time settings that can be used to change these
805 numbers, for those that want even more. Backwards compatibility with old
806 spool files has been maintained. However, going back to a previous Exim
807 release will lost any variables that are in spool files.
809 PH/07 Two small changes when running in the test harness: increase delay when
810 passing a TCP/IP connection to a new process, in case the original
811 process has to generate a bounce, and remove special handling of
812 127.0.0.2 (sic), which is no longer necessary.
814 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
815 be the same on different OS.
817 PH/09 Moved a debug statement in filter processing to avoid a race problem when
820 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
821 whether --show-vars was specified or not
823 JJ/02 exipick: Added support for new ACL variable spool format introduced
826 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
827 syntactically invalid From: or Reply-to: line, and a filter used this to
828 generate an autoreply, and therefore failed to obtain an address for the
829 autoreply, Exim could try to deliver to a non-existent relative file
830 name, causing unrelated and misleading errors. What now happens is that
831 it logs this as a hard delivery error, but does not attempt to create a
834 PH/11 The exinext utility has a -C option for testing purposes, but although
835 the given file was scanned by exinext itself; it wasn't being passed on
836 when Exim was called.
838 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
839 an end-of-file indication when reading a command response.
841 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
842 compiled. In many other places in Exim, IPv6 addresses are always
843 recognized, so I have changed this. It also means that IPv4 domain
844 literals of the form [IPV4:n.n.n.n] are now always recognized.
846 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
847 used if the router is not running as root, for example, when verifying at
848 ACL time, or when using -bh. The debugging output from this situation was
849 non-existent - all you got was a failure to exec. I have made two
852 (a) Failures to set uid/gid, the current directory, or a process leader
853 in a subprocess such as that created by queryprogram now generate
854 suitable debugging ouput when -d is set.
856 (b) The queryprogram router detects when it is not running as root,
857 outputs suitable debugging information if -d is set, and then runs
858 the subprocess without attempting to change uid/gid.
860 PH/15 Minor change to Makefile for building test_host (undocumented testing
863 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
864 additional section of a DNS packet that returns MX or SRV records.
865 Instead, it always explicitly searches for A/AAAA records. This avoids
866 major problems that occur when a DNS server includes only records of one
867 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
868 fixed another bug: if SRV records were looked up and the corresponding
869 address records were *not* found in the additional section, the port
870 values from the SRV records were lost.
872 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
873 using the correct key (the original address) when searching the retry
874 rules in order to find which one to use for generating the retry hint.
876 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
877 adding the default one. Similarly, if it contains a Reply-To: header, the
878 errors_reply_to option, if set, is not used.
880 PH/19 When calculating a retry time, Exim used to measure the "time since
881 failure" by looking at the "first failed" field in the retry record. Now
882 it does not use this if it is later than than the arrival time of the
883 message. Instead it uses the arrival time. This makes for better
884 behaviour in cases where some deliveries succeed, thus re-setting the
885 "first failed" field. An example is a quota failure for a huge message
886 when small messages continue to be delivered. Without this change, the
887 "time since failure" will always be short, possible causing more frequent
888 delivery attempts for the huge message than are intended.
889 [Note: This change was subsequently modified - see PH/04 for 4.62.]
891 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
892 $1, $2, $3) because the numerical variables can be reset during some
893 expansion items (e.g. "match"), thereby losing the authentication data.
895 PH/21 Make -bV show the size of off_t variables so that the test suite can
896 decide whether to run tests for quotas > 2G.
898 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
899 mailbox_size, and mailbox_filecount in the appendfile transport. If a
900 filecount value is greater than 2G or if a quota value is greater than 2G
901 on a system where the size of off_t is not greater than 4, a panic error
904 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
905 never match. The debug and -bh output now contains an explicit error
906 message indicating a malformed IPv4 address or mask.
908 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
909 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
912 PH/25 Do not write to syslog when running in the test harness. The only
913 occasion when this arises is a failure to open the main or panic logs
914 (for which there is an explicit test).
916 PH/26 Added the /no_tell option to "control=freeze".
918 PH/27 If a host name lookup failed very early in a connection, for example, if
919 the IP address matched host_lookup and the reverse lookup yielded a name
920 that did not have a forward lookup, an error message of the form "no IP
921 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
922 could be logged. Now it outputs the IP address instead of "NULL".
924 PH/28 An enabling patch from MH: add new function child_open_exim2() which
925 allows the sender and the authenticated sender to be set when
926 submitting a message from within Exim. Since child_open_exim() is
927 documented for local_scan(), the new function should be too.
929 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
930 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
931 results in an empty string is now treated as unset.
933 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
935 PH/31 Added sender_verify_fail logging option.
937 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
938 needed by RFC 822 but not by RFC 2822 was commented out. I have now
939 tidied the source and removed it altogether.
941 PH/33 When a queue run was abandoned because the load average was too high, a
942 log line was always written; now it is written only if the queue_run log
943 selector is set. In addition, the log line for abandonment now contains
944 information about the queue run such as the pid. This is always present
945 in "start" and "stop" lines but was omitted from the "abandon" line.
947 PH/34 Omit spaces between a header name and the colon in the error message that
948 is given when verify = headers_syntax fails (if there are lots of them,
949 the message gets confusing).
951 PH/35 Change the default for dns_check_names_pattern to allow slashes within
952 names, as there are now some PTR records that contain slashes. This check
953 is only to protect against broken name servers that fall over on strange
954 characters, so the fact that it applies to all lookups doesn't matter.
956 PH/36 Now that the new test suite is complete, we can remove some of the
957 special code in Exim that was needed for the old test suite. For example,
958 sorting DNS records because real resolvers return them in an arbitrary
959 order. The new test suite's fake resolver always returns records in the
962 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
963 bounces) except when queue_only is set, to avoid logging races between
964 the different processes.
966 PH/38 Panic-die if .include specifies a non-absolute path.
968 PH/39 A tweak to the "H" retry rule from its user.
970 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
971 a label. They prevented compilation on older perls.
973 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
974 a warning to be raised on newish perls.
976 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
977 on queue. Changes to match documented behaviour of showing count of
978 messages matching specified criteria.
980 PH/40 Changed the default ident timeout from 30s to 5s.
982 PH/41 Added support for the use of login_cap features, on those BSD systems
983 that have them, for controlling the resources used by pipe deliveries.
985 PH/42 The content-scanning code uses fopen() to create files in which to put
986 message data. Previously it was not paying any attention to the mode of
987 the files. Exim runs with umask(0) because the rest of the code creates
988 files with open(), and sets the required mode explicitly. Thus, these
989 files were ending up world-writeable. This was not a big issue, because,
990 being within the spool directory, they were not world-accessible. I have
991 created a function called modefopen, which takes an additional mode
992 argument. It sets umask(777), creates the file, chmods it to the required
993 mode, then resets the umask. All the relevant calls to fopen() in the
994 content scanning code have been changed to use this function.
996 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
997 to 24 hours. This avoids potential overflow problems when processing G
998 and H retry rules. I suspect nobody ever tinkers with this value.
1000 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1002 PH/45 When the plaintext authenticator is running as a client, the server's
1003 challenges are checked to ensure they are valid base64 strings. By
1004 default, the authentication attempt is cancelled if an invalid string is
1005 received. Setting client_ignore_invalid_base64 true ignores these errors.
1006 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1007 they are received. Thus, the responses can be made to depend on the
1008 challenges. If an invalid string is ignored, an empty string is placed in
1011 PH/46 Messages that are created by the autoreply transport now contains a
1012 References: header, in accordance with RFCs 2822 and 3834.
1014 PH/47 Added authenticated_sender_force to the smtp transport.
1016 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1018 PH/49 Installed latest patch from the Sieve maintainer.
1020 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1021 was also set, the appendfile transport was unnecessarily scanning a
1022 directory of message files (e.g. for maildir delivery) to find the count
1023 of files (along with the size), even though it did not need this
1024 information. It now does the scan only if it needs to find either the
1025 size of the count of files.
1027 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1029 PH/52 Two bugs concerned with error handling when the smtp transport is
1032 (i) Exim was not creating retry information for temporary errors given
1033 for individual recipients after the DATA command when the smtp transport
1034 was used in LMTP mode. This meant that they could be retried too
1035 frequently, and not timed out correctly.
1037 (ii) Exim was setting the flag that allows error details to be returned
1038 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1039 recipients that were returned after the DATA command.
1041 PH/53 This is related to PH/52, but is more general: for any failing address,
1042 when detailed error information was permitted to be returned to the
1043 sender, but the error was temporary, then after the final timeout, only
1044 "retry timeout exceeded" was returned. Now it returns the full error as
1045 well as "retry timeout exceeded".
1047 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1048 do this, and (what is worse) MTAs that accept it.
1050 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1051 will now be deprecated.
1053 PH/56 New os.c-cygwin from the Cygwin maintainer.
1055 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1056 formats (previously only available in exim formats via -bpr, -bpru,
1057 and -bpra. Now also available in native and exiqgrep formats)
1059 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1060 with very large, slow to parse queues
1062 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1064 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1066 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1067 responses to authentication challenges, though it was showing the
1068 challenges; (ii) I've removed the CR characters from the debug output for
1071 PH/58 Allow for the insertion of a newline as well as a space when a string
1072 is turned into more than one encoded-word during RFC 2047 encoding. The
1073 Sieve code now uses this.
1075 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1076 data_4xx, lost_connection, tls_required.
1078 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1079 message was being sent as an SMTP response.
1081 PH/61 Add -l and -k options to exicyclog.
1083 PH/62 When verifying, if an address was redirected to one new address, so that
1084 verification continued, and the new address failed or deferred after
1085 having set something in $address_data, the value of $address_data was not
1086 passed back to the ACL. This was different to the case when no
1087 redirection occurred. The value is now passed back in both cases.
1089 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1090 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1091 use login_cap.h, so on its own it isn't the distinguishing feature. The
1092 new name refers directly to the setclassresources() function.
1094 PH/65 Added configuration files for NetBSD3.
1096 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1098 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1099 is preferred over IPv4.
1101 PH/68 The bounce_return_message and bounce_return_body options were not being
1102 honoured for bounces generated during the reception of non-SMTP messages.
1103 In particular, this applied to messages rejected by the ACL. This bug has
1104 been fixed. However, if bounce_return_message is true and bounce_return_
1105 body is false, the headers that are returned for a non-SMTP message
1106 include only those that have been read before the error was detected.
1107 (In the case of an ACL rejection, they have all been read.)
1109 PH/69 The HTML version of the specification is now built in a directory called
1110 spec_html instead of spec.html, because the latter looks like a path with
1111 a MIME-type, and this confuses some software.
1113 PH/70 Catch two compiler warnings in sieve.c.
1115 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1116 function verify_get_ident() calls ip_connect() to connect a socket, but
1117 if the "connect()" function timed out, ip_connect() used to close the
1118 socket. However, verify_get_ident() also closes the socket later, and in
1119 between Exim writes to the log, which may get opened at this point. When
1120 the socket was closed in ip_connect(), the log could get the same file
1121 descriptor number as the socket. This naturally causes chaos. The fix is
1122 not to close the socket in ip_connect(); the socket should be closed by
1123 the function that creates it. There was only one place in the code where
1124 this was missing, in the iplookup router, which I don't think anybody now
1125 uses, but I've fixed it anyway.
1127 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1128 well as to direct DNS lookups. Otherwise the handling of names in host
1129 lists is inconsistent and therefore confusing.
1135 PH/01 Two changes to the default runtime configuration:
1137 (1) Move the checks for relay_from_hosts and authenticated clients from
1138 after to before the (commented out) DNS black list checks.
1140 (2) Add control=submission to the relay_from_hosts and authenticated
1141 clients checks, on the grounds that messages accepted by these
1142 statements are most likely to be submissions.
1144 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1146 (1) Generate an error if the third argument for the ${prvs expansion is
1149 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1152 (3) Reset the variables that are obtained from the first argument of
1153 ${prvscheck and used in the second argument before leaving the code,
1154 because their memory is reclaimed, so using them afterwards may do
1157 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1158 one (it's much easier than Tom thought :-).
1160 (5) Because of (4), we can now allow for the use of $prvscheck_result
1161 inside the third argument.
1163 PH/03 For some reason, the default setting of PATH when running a command from
1164 a pipe transport was just "/usr/bin". I have changed it to
1167 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1168 anything to be listed in the output from -bV.
1170 PH/05 When a filter generated an autoreply, the entire To: header line was
1171 quoted in the delivery log line, like this:
1173 => >A.N.Other <ano@some.domain> <original@ddress> ...
1175 This has been changed so that it extracts the operative address. There
1176 may be more than one such address. If so, they are comma-separated, like
1179 => >ano@some.domain,ona@other.domain <original@ddress> ...
1181 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1182 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1183 not being looked up in the rDNS to get a host name, Exim was showing the
1184 IP address twice in Received: lines, even though the IP addresses were
1185 identical. For example:
1187 Received: from [1.2.3.4] (helo=[1.2.3.4])
1189 However, if the real host name was known, it was omitting the HELO data
1190 if it matched the actual IP address. This has been tidied up so that it
1191 doesn't show the same IP address twice.
1193 PH/07 When both +timestamp and +memory debugging was on, the value given by
1194 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1195 called by the debug printing, thereby overwriting the timestamp buffer.
1196 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1198 PH/08 When the original message was included in an autoreply transport, it
1199 always said "this is a copy of the message, including all the headers",
1200 even if body_only or headers_only was set. It now gives an appropriate
1203 PH/09 Applied a patch from the Sieve maintainer which:
1205 o fixes some comments
1206 o adds the (disabled) notify extension core
1207 o adds some debug output for the result of if/elsif tests
1208 o points to the current vacation draft in the documentation
1209 and documents the missing references header update
1213 o fixes a bug in processing the envelope test (when testing
1214 multiple envelope elements, the last element determinted the
1217 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1218 Electronic Mail") by including:
1220 Auto-submitted: auto-generated
1222 in the messages that it generates (bounce messages and others, such as
1223 warnings). In the case of bounce messages for non-SMTP mesages, there was
1224 also a typo: it was using "Auto_submitted" (underscore instead of
1225 hyphen). Since every message generated by Exim is necessarily in response
1226 to another message, thes have all been changed to:
1228 Auto-Submitted: auto-replied
1230 in accordance with these statements in the RFC:
1232 The auto-replied keyword:
1234 - SHOULD be used on messages sent in direct response to another
1235 message by an automatic process,
1237 - MUST NOT be used on manually-generated messages,
1239 - MAY be used on Delivery Status Notifications (DSNs) and Message
1240 Disposition Notifications (MDNs),
1242 - MUST NOT be used on messages generated by automatic or periodic
1243 processes, except for messages which are automatic responses to
1246 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1247 to the default Received: header definition.
1249 PH/12 Added log selector acl_warn_skipped (default on).
1251 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1252 variables because (a) they are in the wrong storage pool and (b) even if
1253 they were copied, it wouldn't work properly because of the caching.
1255 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1256 checking when decoding. Apparently there are clients that generate
1257 overlong encoded strings. Why am I not surprised?
1259 PH/15 If the first argument of "${if match_address" was not empty, but did not
1260 contain an "@" character, Exim crashed. Now it writes a panic log message
1261 and treats the condition as false.
1263 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1265 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1266 extension "envelope-auth". The code is finished and in agreement with
1267 other implementations, but there is no documentation so far and in fact,
1268 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1269 not changing the active code.
1271 Print executed "if" and "elsif" statements when debugging is used. This
1272 helps a great deal to understand what a filter does.
1274 Document more things not specified clearly in RFC3028. I had all this
1275 sorted out, when out of a sudden new issues came to my mind. Oops."
1277 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1280 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1281 canonical form (as documented). However, after a host name lookup from
1282 the IP address, check_host() was doing a simple string comparison with
1283 addresses acquired from the DNS when checking that the found name did
1284 have the original IP as one of its addresses. Since any found IPv6
1285 addresses are likely to be in abbreviated form, the comparison could
1286 fail. Luckily, there already exists a function for doing the comparison
1287 by converting both addresses to binary, so now that is used instead of
1288 the text comparison.
1290 PH/20 There was another similar case to PH/19, when a complete host name was
1291 given in a host list; looking up its IP address could give an abbreviated
1292 form, whereas the current host's name might or might not be abbreviated.
1293 The same fix has been applied.
1299 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1300 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1303 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1304 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1306 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1308 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1309 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1310 Also, for "Auto-Submitted", treat anything other than "no" as
1311 non-personal, in accordance with RFC 3834. (Previously it treated
1312 anything starting "auto-" as non-personal.)
1314 TF/01 The control=submission/name=... option had a problem with syntax
1315 errors if the name included a slash character. The /name= option
1316 now slurps the rest of the string, so it can include any characters
1317 but it must come last in the list of options (after /sender_retain
1320 PH/05 Some modifications to the interface to the fake nameserver for the new
1328 TK/01 Added the "success_on_redirect" address verification option. See
1329 NewStuff for rationale and an example.
1331 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1333 PH/02 Patch to exigrep to allow it to work on syslog lines.
1335 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1336 fread() to skip over the body file's header line, because in Cygwin the
1337 header line is locked and is inaccessible.
1339 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1340 co-exist for some time) to make it clear that it is the Exim ID that is
1341 referenced, not the Message-ID: header line.
1343 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1344 string_format() function, because snprintf() does not exist on all
1347 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1348 ${stat: expansion item.
1350 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1351 protocol synchronization error", to keep the pedants happy.
1353 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1354 well as for IRIX systems, when gcc is being used. See the host.c source
1357 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1359 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1362 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1363 transport and to the smtp transport in LMTP mode.
1365 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1367 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1368 run a filter in a subprocess. This could lead to confusion in subsequent
1369 lookups in the parent process. There should also be a search_tidyup() at
1370 the end of the subprocess.
1372 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1373 only if the host matched helo_try_verify_hosts, which caused the
1374 verification to occur when the EHLO/HELO command was issued. The ACL just
1375 tested the remembered result. Now, if a previous verification attempt has
1376 not happened, "verify = helo" does it there and then.
1378 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1380 TK/03 Fix log output including CR from clamd.
1382 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1383 exist provoked a memory error which could cause a segfault.
1385 PH/15 Installed PCRE 6.2
1387 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1389 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1390 of the problem. Specifically, suggested +O2 rather than +O1 for the
1393 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1395 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1396 record, the port was not correctly logged when the outgoing_port log
1397 selector was set (it logged the transort's default port).
1399 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1400 fallback_hosts, and "hosts" in the smtp transport.
1402 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1403 host errors such as "Connection refused".
1405 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1406 authentication with radiusclient 0.4.9:
1408 - Error returned from rc_read_config was caught wrongly
1409 - Username/password not passed on to radius server due to wrong length.
1411 The presumption is that some radiusclient API changes for 4.51/PH/17
1412 were not taken care of correctly. The code is still untested by me (my
1413 Linux distribution still has 0.3.2 of radiusclient), but it was
1414 contributed by a Radius user.
1416 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1417 expanding the "port" option of the smtp transport.
1419 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1420 while reading a MIME header. Thanks to Tom Hughes for a patch.
1422 PH/24 Include config.h inside local_scan.h so that configuration settings are
1425 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1426 that in an ACL for RCPT (for example), you can examine exactly what was
1429 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1430 commands, but it was not correctly comparing the address with the actual
1431 client host address. Thus, it would show the EHLO address in Received:
1432 header lines when this was not necessary.
1434 PH/27 Added the % operator to ${eval:}.
1436 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1437 it should be ignoring failures (because with -C, for example, it has lost
1438 privilege). It wasn't ignoring creation failures other than "already
1441 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1444 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1445 because an input line was too long, either on its own, or by virtue of
1446 too many continuations, the temporary file was not being removed, and the
1447 return code was incorrect.
1449 PH/31 Missing "BOOL" in function definition in filtertest.c.
1451 PH/32 Applied Sieve patches from the maintainer.
1453 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1455 PH/33 Added "verify = not_blind".
1457 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1458 Local/Makefile (with some defaults set). These are used in built scripts
1459 such as exicyclog, but they have never been used in the exim_install
1460 script (though there are many overriding facilities there). I have
1461 arranged that the exim_install script now takes note of these two
1464 PH/35 Installed configuration files for Dragonfly.
1466 PH/36 When a locally submitted message by a trusted user did not contain a
1467 From: header, and the sender address was obtained from -f or from an SMTP
1468 MAIL command, and the trusted user did not use -F to supply a sender
1469 name, $originator_name was incorrectly used when constructing a From:
1470 header. Furthermore, $originator_name was used for submission mode
1471 messages from external hosts without From: headers in a similar way,
1472 which is clearly wrong.
1474 PH/37 Added control=suppress_local_fixups.
1476 PH/38 When log_selector = +received_sender was set, and the addition of the
1477 sender made the log line's construction buffer exactly full, or one byte
1478 less than full, an overflow happened when the terminating "\n" was
1481 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1482 when the result of a list match is failure because a DNS lookup failed.
1484 PH/40 RM_COMMAND is now used in the building process.
1486 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1487 the "build-* directories that it finds.
1489 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1490 address in a domain literal was a prefix of an interface address.
1492 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1493 when verifying a sender address, unless rewrite_headers is false.
1495 PH/44 Wrote a long comment about why errors_to addresses are verified as
1496 recipients, not senders.
1498 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1499 the ratelimit ACL was added.
1501 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1503 PH/46 Added extra information about PostgreSQL errors to the error string.
1505 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1506 avoiding the need to install special zones in a real server. This is
1507 backwards compatible; if it can't find the fake resolver, it drops back.
1508 Thus, both old and new test suites can be run.
1510 TF/02 Added util/ratelimit.pl
1512 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1513 clock is set back in time.
1515 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1516 Candler <B.Candler@pobox.com>.
1518 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1519 OK for addresses that are the result of redirections.
1521 PH/48 A number of further additions for the benefit of the new test suite,
1522 including a fake gethostbyname() that interfaces to the fake DNS resolver
1525 TF/06 The fix for widen_domains has also been applied to qualify_single and
1526 search_parents which are the other dnslookup options that can cause
1529 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1532 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1534 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1535 service" if it encountered an MX record with an empty target hostname.
1536 The message is now "an MX or SRV record indicated no SMTP service".
1538 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1539 if the DNS of the sending site is misconfigured. This is quite a
1540 common situation. This change restores the behaviour of treating a
1541 helo verification defer as a failure.
1543 PH/51 If self=fail was set on a router, the bounce message did not include the
1544 actual error message.
1550 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1552 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1553 command itself ended in error, the underlying message about the transport
1554 filter timeout was being overwritten with the pipe command error. Now the
1555 underlying error message should be appended to the second error message.
1557 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1559 PH/02 Reduce the amount of output that "make" produces by default. Full output
1560 can still be requested.
1562 PH/03 The warning log line about a condition test deferring for a "warn" verb
1563 was being output only once per connection, rather than after each
1564 occurrence (because it was using the same function as for successful
1565 "warn" verbs). This seems wrong, so I have changed it.
1567 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1568 it should not have, which might have caused a crash in the right
1569 circumstances, but probably never did.
1571 PH/04 Installed a modified version of Tony Finch's patch to make submission
1572 mode fix the return path as well as the Sender: header line, and to
1573 add a /name= option so that you can make the user's friendly name appear
1576 TF/03 Added the control = fakedefer ACL modifier.
1578 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1579 Mark Lowes for thorough testing.
1581 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1583 TK/03 Merged latest SRS patch from Miles Wilton.
1585 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1586 with the definition in sysexits.h (which is #included earlier).
1587 Fortunately, Exim does not actually use EX_OK. The code used to try to
1588 preserve the sysexits.h value, by assumimg that macro definitions were
1589 scanned for macro replacements. I have been disabused of this notion,
1590 so now the code just undefines EX_OK before #including unistd.h.
1592 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1593 in the smtp transport. When a block could not be written in a single
1594 write() function, the timeout was being re-applied to each part-write.
1595 This seems wrong - if the receiver was accepting one byte at a time it
1596 would take for ever. The timeout is now adjusted when this happens. It
1597 doesn't have to be particularly precise.
1599 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1600 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1602 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1603 without a domain if the check to <postmaster@domain> fails.
1605 SC/01 Eximstats: added -xls and the ability to specify output files
1606 (patch written by Frank Heydlauf).
1608 SC/02 Eximstats: use FileHandles for outputing results.
1610 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1612 SC/04 Eximstats: fixed display of large numbers with -nvr option
1614 SC/05 Eximstats: fixed merging of reports with empty tables.
1616 SC/06 Eximstats: added the -include_original_destination flag
1618 SC/07 Eximstats: removed tabs and trailing whitespace.
1620 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1622 TK/06 MBOX spool code: Add real "From " MBOX separator line
1623 so the .eml file is really in mbox format (even though
1624 most programs do not really care). Patch from Alex Miller.
1626 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1627 The latter is generated from $received_to and is only set if the
1628 message has one envelope recipient. SA can use these headers,
1629 obviously out-of-the-box. Patch from Alex Miller.
1631 PH/08 The ${def test on a variable was returning false if the variable's
1632 value was "0", contrary to what the specification has always said!
1633 The result should be true unless the variable is empty.
1635 PH/09 The syntax error of a character other than { following "${if
1636 def:variable_name" (after optional whitespace) was not being diagnosed.
1637 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1638 accidental colon was present, for example, could give incorrect results.
1640 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1641 result is used (not including appendfile, where other changes are about
1644 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1645 This involved changing a lot of size variables from int to off_t. It
1646 should work with maildirs and everything.
1648 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1649 spamd dying while we are connected to it.
1651 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1654 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1655 function for BDB 4.3.
1657 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1659 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1660 the sources for PCRE have been re-arranged and more files are now
1663 PH/15 The code I had for printing potentially long long variables in PH/11
1664 above was not the best (it lost precision). The length of off_t variables
1665 is now inspected at build time, and an appropriate printing format (%ld
1666 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1667 to be "long long int" or "long int". This is needed for the internal
1668 formatting function string_vformat().
1670 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1671 the configuration file to be ":syslog", then the script "guesses" where
1672 the logs files are, rather than using the compiled in default. In our
1673 case the guess is not the same as the compiled default, so the script
1674 suddenly stopped working when I started to use syslog. The patch checks
1675 to see if log_file_path is "". If so, it attempts to read it from exim
1676 with no configuration file to get the compiled in version, before it
1677 falls back to the previous guessing code."
1679 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1680 implementing BATV in an Exim configuration. See NewStuff for the gory
1683 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1684 Makefile that are specific to HP-UX.
1686 PH/18 If the "use_postmaster" option was set for a recipient callout together
1687 with the "random" option, the postmaster address was used as the MAIL
1688 FROM address for the random test, but not for the subsequent recipient
1689 test. It is now used for both.
1691 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1692 patch removes a few documentation additions to RFC 3028, because the
1693 latest draft now contains them. It adds the new en;ascii-case comparator
1694 and a new error check for 8bit text in MIME parts. Comparator and
1695 require names are now matched exactly. I enabled the subaddress
1696 extension, but it is not well tested yet (read: it works for me)."
1698 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1699 rework some of the code of TK/09 above to avoid the hardwired use of
1700 "%lld" and "long long". Replaced the call to snprintf() with a call to
1703 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1704 records point to non-existent hosts", "retry timeout exceeded", and
1705 "retry time not reached for any host after a long failure period".
1707 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1708 experimental DomainKeys support:
1710 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1711 (2) On an error such as an illegally used "control", the wrong name for
1712 the control was given.
1714 These problems did NOT occur unless DomainKeys support was compiled.
1716 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1718 PH/24 Added ${if match_ip condition.
1720 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1721 databases so that it will be absolutely obvious if a crash occurs in the
1722 DB library. This is a regular occurrence (often caused by mis-matched
1725 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1726 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1727 picked up on a user's system that detects such things. There doesn't seem
1728 to be a gcc warning option for this - only an attribute that has to be
1729 put on the function's prototype. It seems that in Fedora Core 4 they have
1730 set this on a number of new functions. No doubt there will be more in due
1733 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1734 specify a transport. However, if an address that was verified by such a
1735 router was the subject of a callout, Exim crashed because it tried to
1736 read the rcpt_include_affixes from the non-existent transport. Now it
1737 just assumes that the setting of that option is false. This bug was
1738 introduced by 4.51/PH/31.
1740 PH/28 Changed -d+all to exclude +memory, because that information is very
1741 rarely of interest, but it makes the output a lot bigger. People tend to
1742 do -d+all out of habit.
1744 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1745 code in os-type was giving problems when libc.so lives in lib64, like on
1748 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1749 aren't the modern standard, and it seems that some systems' include files
1750 don't always have them. Exim was already checking for some of the newer
1751 ones like T_AAAA, and defining it itself. I've added checks for all the
1752 record types that Exim uses.
1754 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1755 not automatically generating a new one, as it is supposed to. This
1756 prevented TLS from working. If the file did exist, but contained invalid
1757 data, a new version was generated, as expected. It was only the case of a
1758 non-existent file that was broken.
1760 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1761 with a change in libdomainkeys > 0.64.
1763 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1764 from DNS. If the selector record carries the flag, it now has
1765 precedence over the domain-wide flag.
1767 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1769 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1770 the use of a port name that isn't defined in /etc/services) occurred, the
1771 message was deferred as in a normal delivery, and thus remained on the
1772 spool, instead of being failed because of the mua_wrapper setting. This
1773 is now fixed, and I tidied up some of the mua_wrapper messages at the
1776 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1777 the messages in a hash of arrays rather than using individual hashes.
1778 This is a bit cleaner and results in dramatic memory savings, albeit
1779 at a slight CPU cost.
1781 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1782 as requested by Marc Sherman.
1784 SC/10 Eximstats: added histograms for user specified patterns as requested
1787 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1789 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1790 fopen() in the content-scanning modules that did not already have it.
1796 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1797 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1799 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1801 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1803 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1804 to test Sieve filters that use "vacation".
1806 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1807 that changes the way the GnuTLS parameters are stored in the cache file.
1808 The new format can be generated externally. For backward compatibility,
1809 if the data in the cache doesn't make sense, Exim assumes it has read an
1810 old-format file, and it generates new data and writes a new file. This
1811 means that you can't go back to an older release without removing the
1814 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1815 work if there are any delivery delays because "one_time" forces the
1816 parent to be marked "delivered", so its unseen clone is never tried
1817 again. For this reason, Exim now forbids the simultaneous setting of
1820 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1821 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1822 entry if you want to know the details. The fix, however, neglected to
1823 consider the case when local delivery batching is involved. The test for
1824 "previously delivered" was not happening when checking to see if an
1825 address could be batched with a previous (undelivered) one; under
1826 certain circumstances this could lead to multiple deliveries to the same
1829 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1830 in its include files, and this causes problems building Exim.
1832 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1833 header_syntax) but Exim was just ignoring anything given after a slash.
1834 In particular, this caused confusion with an attempt to use "verify =
1835 reverse_host_lookup/defer_ok". An error is now given when options are
1836 supplied for verify items that do not have them. (Maybe reverse_host_
1837 lookup should have a defer_ok option, but that's a different point.)
1839 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1840 defined by RFC 821) to 2048, because there were problems with some AUTH
1841 commands, and RFC 1869 says the size should be increased for extended
1842 SMTP commands that take arguments.
1844 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1847 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1848 "unknown" error; now it says that the functionality isn't in the binary.
1850 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1851 an address' error message when a string expansion fails (syntax or
1852 whatever). Otherwise the password may appear in the log. Following change
1853 PH/42 below, there is no longer a chance of it appearing in a bounce
1856 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1858 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1859 of its addresses was ever tried. (Bugzilla bug #2.)
1861 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1862 the result incorrectly in the debug output. (It correctly added a newline
1863 to what was transported.)
1865 TF/01 Added $received_time.
1867 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1868 commented out examples of how to interface to a virus scanner and to
1869 SpamAssassin. Also added commented examples of av_scanner and
1870 spamd_address settings.
1872 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1873 and controls are allowed in which ACLs. There were a couple of minor
1874 errors. Some of the entries in the conditions table (which is a table of
1875 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1876 negation of where the condition IS allowed.
1878 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1880 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1881 header file does not have a version number, so I've had to invent a new
1882 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1883 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1884 radiusclient), but it was contributed by a Radius user.
1886 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1887 files or directories, for OpenSSL.
1889 PH/19 When an Exim process that is running as root has to create an Exim log
1890 file, it does so in a subprocess that runs as exim:exim so as to get the
1891 ownership right at creation (otherwise, other Exim processes might see
1892 the file with the wrong ownership). There was no test for failure of this
1893 fork() call, which would lead to the process getting stuck as it waited
1894 for a non-existent subprocess. Forks do occasionally fail when resources
1895 run out. I reviewed all the other calls to fork(); they all seem to check
1898 PH/20 When checking for unexpected SMTP input at connect time (before writing
1899 the banner), Exim was not dealing correctly with a non-positive return
1900 from the read() function. If the client had disconnected by this time,
1901 the result was a log entry for a synchronization error with an empty
1902 string after "input=" when read() returned zero. If read() returned -1
1903 (an event I could not check), uninitialized data bytes were printed.
1904 There were reports of junk text (parts of files, etc) appearing after
1907 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1909 PH/22 Added support for macro redefinition, and (re)definition in between
1910 driver and ACL definitions.
1912 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1913 forgetting to use the resulting value; it was using the unexpanded value.
1915 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1916 hadn't been configured. The fix is from Juergen Kreileder, who
1917 understands it better than I do:
1919 "Here's what I see happening with three configured cyrus_sasl
1920 authenticators configured (plain, login, cram-md5):
1922 On startup auth_cyrus_sasl_init() gets called for each of these.
1923 This means three calls to sasl_listmech() without a specified mech_list.
1924 => SASL tests which mechs of all available mechs actually work
1925 => three warnings about OTP not working
1926 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1928 With the patch, sasl_listmech() also gets called three times. But now
1929 SASL's mech_list option is set to the server_mech specified in the the
1930 authenticator. Or in other words, the answer from sasl_listmech()
1931 gets limited to just the mech you're testing for (which is different
1933 => the return list contains just 'plain' or 'login', 'cram-md5' or
1934 nothing depending on the value of ob->server_mech.
1936 I've just tested the patch: Authentication still works fine,
1937 unavailable mechs specified in the exim configuration are still
1938 caught, and the auth.log warnings about OTP are gone."
1940 PH/25 When debugging is enabled, the contents of the command line are added
1941 to the debugging output, even when log_selector=+arguments is not
1944 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1945 answer is "GNU", and only if the return is "GNU/something" is the answer
1948 PH/27 $acl_verify_message is now set immediately after the failure of a
1949 verification in an ACL, and so is available in subsequent modifiers. In
1950 particular, the message can be preserved by coding like this:
1952 warn !verify = sender
1953 set acl_m0 = $acl_verify_message
1955 Previously, $acl_verify_message was set only while expanding "message"
1956 and "log_message" when a very denied access.
1958 PH/28 Modified OS/os.c-Linux with
1960 -#ifndef OS_LOAD_AVERAGE
1961 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1963 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1964 nomenclature these days.)
1966 PH/29 Installed patch from the Sieve maintainer that adds the options
1967 sieve_useraddress and sieve_subaddress to the redirect router.
1969 PH/30 In these circumstances:
1970 . Two addresses routed to the same list of hosts;
1971 . First host does not offer TLS;
1972 . First host accepts first address;
1973 . First host gives temporary error to second address;
1974 . Second host offers TLS and a TLS session is established;
1975 . Second host accepts second address.
1976 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1977 and peerdn, if requested) that were in fact used only for the second
1980 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1981 attention to any local part prefix or suffix that was matched by the
1982 router that accepted the address. It now behaves in the same way as it
1983 does for delivery: the affixes are removed from the local part unless
1984 rcpt_include_affixes is set on the transport.
1986 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1987 timeout during the DATA phase of an incoming message.
1989 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1990 applied a patch sanctioned by the Sieve maintainer.
1992 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1993 the uid or gid is negative. A case of a negative gid caused this to be
1994 noticed. The fix allows for either to be negative.
1996 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1997 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2000 PH/36 The change PH/12 above was broken. Fixed it.
2002 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2003 the grounds that routing the same address twice would always produce the
2004 same answer. This might have been true once, but it is certainly no
2005 longer true now. Routing a child address may depend on the previous
2006 routing that produced that child. Some complicated redirection strategies
2007 went wrong when messages had multiple recipients, and made Exim's
2008 behaviour dependent on the order in which the addresses were given.
2010 I have moved the duplicate checking until after the routing is complete.
2011 Exim scans the addresses that are assigned to local and remote
2012 transports, and removes any duplicates. This means that more work will be
2013 done, as duplicates will always all be routed, but duplicates are
2014 presumably rare, so I don't expect this is of any significance.
2016 For deliveries to pipes, files, and autoreplies, the duplicate checking
2017 still happens during the routing process, since they are not going to be
2020 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2021 It corrects a timeout issue with spamd. This is Ian's comment: "The
2022 background is that sometimes spamd either never reads data from a
2023 connection it has accepted, or it never writes response data. The exiscan
2024 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2025 blindly assumes that writes won't block so it may never time out."
2027 PH/39 Allow G after quota size as well as K and M.
2029 PH/40 The value set for $authenticated_id in an authenticator may not contain
2030 binary zeroes or newlines because the value is written to log lines and
2031 to spool files. There was no check on this. Now the value is run through
2032 the string_printing() function so that such characters are converted to
2033 printable escape sequences.
2035 PH/41 $message_linecount is a new variable that contains the total number of
2036 lines in the message. Compare $body_linecount, which is the count for the
2039 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2040 bounce and delay warning messages, except in certain special cases, which
2043 (a) An SMTP error message from a remote host;
2044 (b) A message specified in a :fail: redirection;
2045 (c) A message specified in a "fail" command in a system filter;
2046 (d) A message specified in a FAIL return from the queryprogram router;
2047 (e) A message specified by the cannot_route_message router option.
2049 In these cases only, Exim does include the error details in bounce and
2050 warning messages. There are also a few cases where bland messages such
2051 as "unrouteable address" or "local delivery error" are given.
2053 PH/43 $value is now also set for the "else" part of a ${run expansion.
2055 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2056 being worked on, but at least Exim now implements the latest version to
2059 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2060 process to complete was treated as a delivery failure, a timeout while
2061 writing the message to the pipe was logged, but erroneously treated as a
2062 successful delivery. Such timeouts include transport filter timeouts. For
2063 consistency with the overall process timeout, these timeouts are now
2064 treated as errors, giving rise to delivery failures by default. However,
2065 there is now a new Boolean option for the pipe transport called
2066 timeout_defer, which, if set TRUE, converts the failures into defers for
2067 both kinds of timeout. A transport filter timeout is now identified in
2070 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2071 systems where "make" and "gmake" are different, calling "gmake" at top
2072 level broke things. I've arranged for the value of $(MAKE) to be passed
2073 from the Makefile to this script so that it can call the same version of
2077 A note about Exim versions 4.44 and 4.50
2078 ----------------------------------------
2080 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2081 changes of various kinds. As a consequence, a big documentation update was
2082 needed. This delayed the release for rather longer than seemed good, especially
2083 in the light of a couple of (minor) security issues. Therefore, the changes
2084 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2085 release. So 4.44 and 4.50 are in effect two different branches that both start
2088 I have left the 4.50 change log unchanged; it contains all the changes since
2089 4.43. The change log for 4.44 is below; many of its items are identical to
2090 those for 4.50. This seems to be the most sensible way to preserve the
2091 historical information.
2097 1. Minor wording change to the doc/README.SIEVE file.
2099 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2100 computation of the current number of files was incorrect.
2102 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2103 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2106 4. Give more explanation in the error message when the command for a transport
2107 filter fails to execute.
2109 5. There are several places where Exim runs a non-Exim command in a
2110 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2111 was being done only for the command run by the queryprogram router. It is
2112 now done for all such subprocesses. The other cases are: ${run, transport
2113 filters, and the commands run by the lmtp and pipe transports.
2115 6. Added CONFIGURE_GROUP build-time option.
2117 7. Some older OS have a limit of 256 on the maximum number of file
2118 descriptors. Exim was using setrlimit() to set 1000 as a large value
2119 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2120 systems. I've change it so that if it can't get 1000, it tries for 256.
2122 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2123 was an oversight, and furthermore, ever since the addition of extra
2124 controls (e.g. 4.43/32), the checks on when to allow different forms of
2125 "control" were broken. There should now be diagnostics for all cases when a
2126 control that does not make sense is encountered.
2128 9. Added the /retain_sender option to "control=submission".
2130 10. $recipients is now available in the predata ACL (oversight).
2132 11. Tidy the search cache before the fork to do a delivery from a message
2133 received from the command line. Otherwise the child will trigger a lookup
2134 failure and thereby defer the delivery if it tries to use (for example) a
2135 cached ldap connection that the parent has called unbind on.
2137 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2138 of $address_data from the recipient verification was clobbered by the
2139 sender verification.
2141 13. The value of address_data from a sender verification is now available in
2142 $sender_address_data in subsequent conditions in the ACL statement.
2144 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2146 15. Added a new option "connect=<time>" to callout options, to set a different
2149 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2150 was its contents. (It was OK if the option was not defined at all.)
2152 17. A "Completed" log line is now written for messages that are removed from
2153 the spool by the -Mrm option.
2155 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2156 information about exactly what failed.
2158 19. Added -dd to debug only the daemon process.
2160 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2161 handles timeouts, both on the server side and network timeouts. Renamed the
2162 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2164 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2165 It is now set to "smtps".
2167 22. $host_address is now set to the target address during the checking of
2168 ignore_target_hosts.
2170 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2171 being passed; this would have caused $sender_host_name to have been used if
2172 matching the list had actually called for a host name (not very likely,
2173 since this list is usually IP addresses). A host name is now passed as
2176 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2177 code that specifies a non-restarting handler (typically sigaction() in
2178 modern systems) in an attempt to fix a rare and obscure crash bug.
2180 25. Narrowed the window for a race in the daemon that could cause it to ignore
2181 SIGCHLD signals. This is not a major problem, because they are used only to
2182 wake it up if nothing else does.
2184 26. A malformed maildirsize file could cause Exim to calculate negative values
2185 for the mailbox size or file count. Odd effects could occur as a result.
2186 The maildirsize information is now recalculated if the size or filecount
2189 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2190 support for a long time. Removed HAVE_SYS_VFS_H.
2192 28. Installed the latest version of exipick from John Jetmore.
2194 29. In an address list, if the pattern was not a regular expression, an empty
2195 subject address (from a bounce message) matched only if the pattern was an
2196 empty string. Non-empty patterns were not even tested. This was the wrong
2197 because it is perfectly reasonable to use an empty address as part of a
2198 database query. An empty address is now tested by patterns that are
2199 lookups. However, all the other forms of pattern expect the subject to
2200 contain a local part and a domain, and therefore, for them, an empty
2201 address still always fails if the pattern is not itself empty.
2203 30. Exim went into a mad DNS loop when attempting to do a callout where the
2204 host was specified on an smtp transport, and looking it up yielded more
2205 than one IP address.
2207 31. Re-factored the code for checking spool and log partition space into a
2208 function that finds that data and another that does the check. The former
2209 is then used to implement four new variables: $spool_space, $log_space,
2210 $spool_inodes, and $log_inodes.
2212 32. The RFC2047 encoding function was originally intended for short strings
2213 such as real names; it was not keeping to the 75-character limit for
2214 encoded words that the RFC imposes. It now respects the limit, and
2215 generates multiple encoded words if necessary. To be on the safe side, I
2216 have increased the buffer size for the ${rfc2047: expansion operator from
2219 33. It is now permitted to omit both strings after an "if" condition; if the
2220 condition is true, the result is "true". As before, when the second string
2221 is omitted, a false condition yields an empty string. This makes it less
2222 cumbersome to write custom ACL and router conditions.
2224 34. Failure to deliver a bounce message always caused it to be frozen, even if
2225 there was an errors_to setting on the router. The errors_to setting is now
2228 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2229 canonical form (fully expanded) before being placed in
2230 $sender_host_address.
2232 36. The table in the code that translates DNS record types into text (T_A to
2233 "A" for instance) was missing entries for NS and CNAME. It is just possible
2234 that this could have caused confusion if both these types were looked up
2235 for the same domain, because the text type is used as part of Exim's
2236 per-process caching. But the chance of anyone hitting this buglet seems
2239 37. The dnsdb lookup has been extended in a number of ways.
2241 (1) There is a new type, "zns", which walks up the domain tree until it
2242 finds some nameserver records. It should be used with care.
2244 (2) There is a new type, "mxh", which is like "mx" except that it returns
2245 just the host names, not the priorities.
2247 (3) It is now possible to give a list of domains (or IP addresses) to be
2248 looked up. The behaviour when one of the lookups defers can be
2249 controlled by a keyword.
2251 (4) It is now possible to specify the separator character for use when
2252 multiple records are returned.
2254 38. The dnslists ACL condition has been extended: it is now possible to supply
2255 a list of IP addresses and/or domains to be looked up in a particular DNS
2258 39. Added log_selector=+queue_time_overall.
2260 40. When running the queue in the test harness, wait just a tad after forking a
2261 delivery process, to get repeatability of debugging output.
2263 41. Include certificate and key file names in error message when GnuTLS fails
2264 to set them up, because the GnuTLS error message doesn't include the name
2265 of the failing file when there is a problem reading it.
2267 42. Allow both -bf and -bF in the same test run.
2269 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2271 44. The "Exiscan patch" is now merged into the mainline Exim source.
2273 45. Sometimes the final signoff response after QUIT could fail to get
2274 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2275 < 0 before doing a fflush(). This bug looks as though it goes back to the
2276 introduction of TLS in release 3.20, but "sometimes" must have been rare
2277 because the tests only now provoked it.
2279 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2280 (this can affect the format of dates).
2282 47. exim_tidydb, when checking for the continued existence of a message for
2283 which it has found a message-specific retry record, was not finding
2284 messages that were in split spool directories. Consequently, it was
2285 deleting retry records that should have stayed in existence.
2287 48. Steve fixed some bugs in eximstats.
2289 49. The SPA authentication driver was not abandoning authentication and moving
2290 on to the next authenticator when an expansion was forced to fail,
2291 contradicting the general specification for all authenticators. Instead it
2292 was generating a temporary error. It now behaves as specified.
2294 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2295 (the order specifies the preference for clients). The order is now AES256,
2296 AES128, 3DES, ARCFOUR128.
2298 51. Small patch to Sieve code - explicitly set From: when generating an
2301 52. Exim crashed if a remote delivery caused a very long error message to be
2302 recorded - for instance if somebody sent an entire SpamAssassin report back
2303 as a large number of 550 error lines. This bug was coincidentally fixed by
2304 increasing the size of one of Exim's internal buffers (big_buffer) that
2305 happened as part of the Exiscan merge. However, to be on the safe side, I
2306 have made the code more robust (and fixed the comments that describe what
2309 53. Now that there can be additional text after "Completed" in log lines (if
2310 the queue_time_overall log selector is set), a one-byte patch to exigrep
2311 was needed to allow it to recognize "Completed" as not the last thing in
2314 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2315 patch that reportedly fixes this has been added. I am not expert enough to
2316 create a test for it. This is what the patch creator wrote:
2318 "I found a little strange behaviour of ldap code when working with
2319 Windows 2003 AD Domain, where users was placed in more than one
2320 Organization Units. When I tried to give exim partial DN, the exit code
2321 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2322 But simultaneously result of request was absolutely normal ldap result,
2323 so I produce this patch..."
2325 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2326 REFERENCE, so I have modified the code to exclude the patch when that macro
2329 55. Some experimental protocols are using DNS PTR records for new purposes. The
2330 keys for these records are domain names, not reversed IP addresses. The
2331 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2332 leaves it alone. Component reversal etc. now happens only for IP addresses.
2335 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2337 57. Double the size of the debug message buffer (to 2048) so that more of very
2338 long debug lines gets shown.
2340 58. The exicyclog utility now does better if the number of log files to keep
2341 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2343 59. Two changes related to the smtp_active_hostname option:
2345 (1) $smtp_active_hostname is now available as a variable.
2346 (2) The default for smtp_banner uses $smtp_active_hostname instead
2347 of $primary_hostname.
2349 60. The host_aton() function is supposed to be passed a string that is known
2350 to be a valid IP address. However, in the case of IPv6 addresses, it was
2351 not checking this. This is a hostage to fortune. Exim now panics and dies
2352 if the condition is not met. A case was found where this could be provoked
2353 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2354 components; fortuitously, this particular loophole had already been fixed
2355 by change 4.50/55 above.
2357 If there are any other similar loopholes, the new check in host_aton()
2358 itself should stop them being exploited. The report I received stated that
2359 data on the command line could provoke the exploit when Exim was running as
2360 exim, but did not say which command line option was involved. All I could
2361 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2362 running as the user.
2365 61. There was a buffer overflow vulnerability in the SPA authentication code
2366 (which came originally from the Samba project). I have added a test to the
2367 spa_base64_to_bits() function which I hope fixes it.
2370 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2371 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2373 63. The daemon start-up calls getloadavg() while still root for those OS that
2374 need the first call to be done as root, but it missed one case: when
2375 deliver_queue_load_max is set with deliver_drop_privilege. This is
2376 necessary for the benefit of the queue runner, because there is no re-exec
2377 when deliver_drop_privilege is set.
2379 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2380 This has been fixed.
2382 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2383 was in use, was not putting the data itself into the right store pool;
2384 consequently, it could be overwritten for a subsequent message in the same
2385 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2388 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2390 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2391 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2392 the function were treating the return as a boolean value, which happened to
2393 work because 0=false and not-0=true, but is not correct code.
2395 68. The host_aton() function was not handling scoped IPv6 addresses (those
2396 with, for example, "%eth0" on the end) correctly.
2398 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2399 negated items (that is, ~something) in unsigned ints. Some compilers
2400 apparently mutter when there is no cast.
2402 70. If an address verification called from an ACL failed, and did not produce a
2403 user-specific message (i.e. there was only a "system" message), nothing was
2404 put in $acl_verify_message. In this situation, it now puts the system
2407 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2408 session; change 4.31/43 added the unwanted input to the log line - except
2409 that it did not do this in the start of session case. It now does.
2411 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2412 This is wrong and can cause the other end to generate a synchronization
2413 error if it is another Exim or anything else that does the synchronization
2414 check. A QUIT command is no longer sent after a timeout.
2416 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2417 during host lookups.
2419 74. The defer_ok option of callout verification was not working if it was used
2420 when verifying addresses in header lines, that is, for this case:
2422 verify = header_sender/callout=defer_ok
2424 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2425 those file descriptors could be used for SMTP connections. If anything
2426 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2427 could be sent to the SMTP client, causing chaos. The daemon now opens
2428 stdin, stdout, and stderr to /dev/null when it puts itself into the
2431 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2432 log by default. The user can override this with suitable Perl magic.
2434 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2435 the log message when discard triggers, was not working for the DATA ACL or
2436 for the non-SMTP ACL.
2438 78. Error message wording change in sieve.c.
2440 79. If smtp_accept_max_per_host was set, the number of connections could be
2441 restricted to fewer than expected, because the daemon was trying to set up
2442 a new connection before checking whether the processes handling previous
2443 connections had finished. The check for completed processes is now done
2444 earlier. On busy systems, this bug wouldn't be noticed because something
2445 else would have woken the daemon, and it would have reaped the completed
2448 80. If a message was submitted locally by a user whose login name contained one
2449 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2450 It caused a spool format error. I have fixed the spool reading code. A
2451 related problem was that the "from" clause in the Received: line became
2452 illegal because of the space(s). It is now covered by ${quote_local_part.
2454 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2457 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2459 83. Patch from Sieve maintainer to fix unterminated string problem in
2460 "vacation" handling.
2462 84. Some minor changes to the Linux configuration files to help with other
2463 OS variants using glibc.
2465 85. One more patch for Sieve to update vacation handling to latest spec.
2468 ----------------------------------------------------
2469 See the note above about the 4.44 and 4.50 releases.
2470 ----------------------------------------------------
2476 1. Change 4.43/35 introduced a bug that caused file counts to be
2477 incorrectly computed when quota_filecount was set in an appendfile
2480 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2481 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2484 3. Give more explanation in the error message when the command for a transport
2485 filter fails to execute.
2487 4. There are several places where Exim runs a non-Exim command in a
2488 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2489 was being done only for the command run by the queryprogram router. It is
2490 now done for all such subprocesses. The other cases are: ${run, transport
2491 filters, and the commands run by the lmtp and pipe transports.
2493 5. Some older OS have a limit of 256 on the maximum number of file
2494 descriptors. Exim was using setrlimit() to set 1000 as a large value
2495 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2496 systems. I've change it so that if it can't get 1000, it tries for 256.
2498 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2499 was an oversight, and furthermore, ever since the addition of extra
2500 controls (e.g. 4.43/32), the checks on when to allow different forms of
2501 "control" were broken. There should now be diagnostics for all cases when a
2502 control that does not make sense is encountered.
2504 7. $recipients is now available in the predata ACL (oversight).
2506 8. Tidy the search cache before the fork to do a delivery from a message
2507 received from the command line. Otherwise the child will trigger a lookup
2508 failure and thereby defer the delivery if it tries to use (for example) a
2509 cached ldap connection that the parent has called unbind on.
2511 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2512 of $address_data from the recipient verification was clobbered by the
2513 sender verification.
2515 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2516 was its contents. (It was OK if the option was not defined at all.)
2518 11. A "Completed" log line is now written for messages that are removed from
2519 the spool by the -Mrm option.
2521 12. $host_address is now set to the target address during the checking of
2522 ignore_target_hosts.
2524 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2525 being passed; this would have caused $sender_host_name to have been used if
2526 matching the list had actually called for a host name (not very likely,
2527 since this list is usually IP addresses). A host name is now passed as
2530 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2531 code that specifies a non-restarting handler (typically sigaction() in
2532 modern systems) in an attempt to fix a rare and obscure crash bug.
2534 15. Narrowed the window for a race in the daemon that could cause it to ignore
2535 SIGCHLD signals. This is not a major problem, because they are used only to
2536 wake it up if nothing else does.
2538 16. A malformed maildirsize file could cause Exim to calculate negative values
2539 for the mailbox size or file count. Odd effects could occur as a result.
2540 The maildirsize information is now recalculated if the size or filecount
2543 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2544 support for a long time. Removed HAVE_SYS_VFS_H.
2546 18. Updated exipick to current release from John Jetmore.
2548 19. Allow an empty sender to be matched against a lookup in an address list.
2549 Previously the only cases considered were a regular expression, or an
2552 20. Exim went into a mad DNS lookup loop when doing a callout where the
2553 host was specified on the transport, if the DNS lookup yielded more than
2556 21. The RFC2047 encoding function was originally intended for short strings
2557 such as real names; it was not keeping to the 75-character limit for
2558 encoded words that the RFC imposes. It now respects the limit, and
2559 generates multiple encoded words if necessary. To be on the safe side, I
2560 have increased the buffer size for the ${rfc2047: expansion operator from
2563 22. Failure to deliver a bounce message always caused it to be frozen, even if
2564 there was an errors_to setting on the router. The errors_to setting is now
2567 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2568 canonical form (fully expanded) before being placed in
2569 $sender_host_address.
2571 24. Updated eximstats to version 1.33
2573 25. Include certificate and key file names in error message when GnuTLS fails
2574 to set them up, because the GnuTLS error message doesn't include the name
2575 of the failing file when there is a problem reading it.
2577 26. Expand error message when OpenSSL has problems setting up cert/key files.
2580 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2581 (this can affect the format of dates).
2583 28. exim_tidydb, when checking for the continued existence of a message for
2584 which it has found a message-specific retry record, was not finding
2585 messages that were in split spool directories. Consequently, it was
2586 deleting retry records that should have stayed in existence.
2588 29. eximstats updated to version 1.35
2589 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2590 1.35 - bugfix such that pie charts by volume are generated correctly
2592 30. The SPA authentication driver was not abandoning authentication and moving
2593 on to the next authenticator when an expansion was forced to fail,
2594 contradicting the general specification for all authenticators. Instead it
2595 was generating a temporary error. It now behaves as specified.
2597 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2598 (the order specifies the preference for clients). The order is now AES256,
2599 AES128, 3DES, ARCFOUR128.
2601 31. Small patch to Sieve code - explicitly set From: when generating an
2604 32. Exim crashed if a remote delivery caused a very long error message to be
2605 recorded - for instance if somebody sent an entire SpamAssassin report back
2606 as a large number of 550 error lines. This bug was coincidentally fixed by
2607 increasing the size of one of Exim's internal buffers (big_buffer) that
2608 happened as part of the Exiscan merge. However, to be on the safe side, I
2609 have made the code more robust (and fixed the comments that describe what
2612 33. Some experimental protocols are using DNS PTR records for new purposes. The
2613 keys for these records are domain names, not reversed IP addresses. The
2614 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2615 leaves it alone. Component reversal etc. now happens only for IP addresses.
2618 34. The host_aton() function is supposed to be passed a string that is known
2619 to be a valid IP address. However, in the case of IPv6 addresses, it was
2620 not checking this. This is a hostage to fortune. Exim now panics and dies
2621 if the condition is not met. A case was found where this could be provoked
2622 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2623 components; fortuitously, this particular loophole had already been fixed
2624 by change 4.50/55 or 4.44/33 above.
2626 If there are any other similar loopholes, the new check in host_aton()
2627 itself should stop them being exploited. The report I received stated that
2628 data on the command line could provoke the exploit when Exim was running as
2629 exim, but did not say which command line option was involved. All I could
2630 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2631 running as the user.
2634 35. There was a buffer overflow vulnerability in the SPA authentication code
2635 (which came originally from the Samba project). I have added a test to the
2636 spa_base64_to_bits() function which I hope fixes it.
2639 36. The daemon start-up calls getloadavg() while still root for those OS that
2640 need the first call to be done as root, but it missed one case: when
2641 deliver_queue_load_max is set with deliver_drop_privilege. This is
2642 necessary for the benefit of the queue runner, because there is no re-exec
2643 when deliver_drop_privilege is set.
2645 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2646 was in use, was not putting the data itself into the right store pool;
2647 consequently, it could be overwritten for a subsequent message in the same
2648 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2651 38. Sometimes the final signoff response after QUIT could fail to get
2652 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2653 < 0 before doing a fflush(). This bug looks as though it goes back to the
2654 introduction of TLS in release 3.20, but "sometimes" must have been rare
2655 because the tests only now provoked it.
2661 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2662 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2663 Now it returns an integer. A number of calls were still expecting a T/F
2664 return. Fortuitously, in all cases, the tests worked in OK situations,
2665 which is the norm. However, things would have gone wrong on any write
2666 failures on the smtp file descriptor. This function is used when sending
2667 messages over SMTP and also when doing verify callouts.
2669 2. When Exim is called to do synchronous delivery of a locally submitted
2670 message (the -odf or -odi options), it no longer closes stderr before doing
2673 3. Implemented the mua_wrapper option.
2675 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2677 5. Implemented the functions header_remove(), header_testname(),
2678 header_add_at_position(), and receive_remove_recipient(), and exported them
2681 6. If an ACL "warn" statement specified the addition of headers, Exim already
2682 inserted X-ACL-Warn: at the start if there was no header name. However, it
2683 was not making this test for the second and subsequent header lines if
2684 there were newlines in the string. This meant that an invalid header could
2685 be inserted if Exim was badly configured.
2687 7. Allow an ACL "warn" statement to add header lines at the start or after all
2688 the Received: headers, as well as at the end.
2690 8. Added the rcpt_4xx retry error code.
2692 9. Added postmaster_mailfrom=xxx to callout verification option.
2694 10. Added mailfrom=xxxx to the callout verification option, for verify=
2697 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2698 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2700 12. Inserted some casts to stop certain compilers warning when using pointer
2701 differences as field lengths or precisions in printf-type calls (mostly
2702 affecting debugging statements).
2704 13. Added optional readline() support for -be (dynamically loaded).
2706 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2707 same clock tick as a message's arrival, so that its received time was the
2708 same as the "first fail" time on the retry record, and that message
2709 remained on the queue past the ultimate address timeout, every queue runner
2710 would try a delivery (because it was past the ultimate address timeout) but
2711 after another failure, the ultimate address timeout, which should have then
2712 bounced the address, did not kick in. This was a "< instead of <=" error;
2713 in most cases the first failure would have been in the next clock tick
2714 after the received time, and all would be well.
2716 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2717 being recognized when the domain list was tested by the match_domain
2718 condition in an expansion string.
2720 16. Added the ${str2b64: operator.
2722 17. Exim was always calling setrlimit() to set a large limit for the number of
2723 processes, without checking whether the existing limit was already
2724 adequate. (It did check for the limit on file descriptors.) Furthermore,
2725 errors from getrlimit() and setrlimit() were being ignored. Now they are
2726 logged to the main and panic logs, but Exim does carry on, to try to do its
2727 job under whatever limits there are.
2729 18. Imported PCRE 5.0.
2731 19. Trivial typo in log message " temporarily refused connection" (the leading
2734 20. If the log selector return_path_on_delivery was set and an address was
2735 redirected to /dev/null, the delivery process crashed because it assumed
2736 that a return path would always be set for a "successful" delivery. In this
2737 case, the whole delivery is bypassed as an optimization, and therefore no
2740 21. Internal re-arrangement: the function for sending a challenge and reading
2741 a response while authentication was assuming a zero-terminated challenge
2742 string. It's now changed to take a pointer and a length, to allow for
2743 binary data in such strings.
2745 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2747 23. Exim was not respecting finduser_retries when seeking the login of the
2748 uid under which it was called; it was always trying 10 times. (The default
2749 setting of finduser_retries is zero.) Also, it was sleeping after the final
2750 failure, which is pointless.
2752 24. Implemented tls_on_connect_ports.
2754 25. Implemented acl_smtp_predata.
2756 26. If the domain in control=submission is set empty, Exim assumes that the
2757 authenticated id is a complete email address when it generates From: or
2758 Sender: header lines.
2760 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2761 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2762 chown and chgrp in /bin and hostname in /usr/bin.
2764 28. Exim was keeping the "process log" file open after each use, just as it
2765 does for the main log. This opens the possibility of it remaining open for
2766 long periods when the USR1 signal hits a daemon. Occasional processlog
2767 errors were reported, that could have been caused by this. Anyway, it seems
2768 much more sensible not to leave this file open at all, so that is what now
2771 29. The long-running daemon process does not normally write to the log once it
2772 has entered its main loop, and it closes the log before doing so. This is
2773 so that log files can straightforwardly be renamed and moved. However,
2774 there are a couple of unusual error situations where the daemon does write
2775 log entries, and I had neglected to close the log afterwards.
2777 30. The text of an SMTP error response that was received during a remote
2778 delivery was being truncated at 512 bytes. This is too short for some of
2779 the long messages that one sometimes sees. I've increased the limit to
2782 31. It is now possible to make retry rules that apply only when a message has a
2783 specific sender, in particular, an empty sender.
2785 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2786 it possible to be selective about when SMTP synchronization is enforced.
2788 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2790 32. Implemented hosts_connection_nolog.
2792 33. Added an ACL for QUIT.
2794 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2797 35. Added mailbox_size and mailbox_filecount to appendfile.
2799 36. Added control = no_multiline_responses to ACLs.
2801 37. There was a bug in the logic of the code that waits for the clock to tick
2802 in the case where the clock went backwards by a substantial amount such
2803 that the microsecond fraction of "now" was more than the microsecond
2804 fraction of "then" (but the whole seconds number was less).
2806 38. Added support for the libradius Radius client library this is found on
2807 FreeBSD (previously only the radiusclient library was supported).
2813 1. When certain lookups returned multiple values in the form name=value, the
2814 quoting of the values was not always being done properly. Specifically:
2815 (a) If the value started with a double quote, but contained no whitespace,
2817 (b) If the value contained whitespace other than a space character (i.e.
2818 tabs or newlines or carriage returns) it was not quoted.
2819 This fix has been applied to the mysql and pgsql lookups by writing a
2820 separate quoting function and calling it from the lookup code. The fix
2821 should probably also be applied to nisplus, ibase and oracle lookups, but
2822 since I cannot test any of those, I have not disturbed their existing code.
2824 2. A hit in the callout cache for a specific address caused a log line with no
2825 reason for rejecting RCPT. Now it says "Previous (cached) callout
2826 verification failure".
2828 3. There was an off-by-one bug in the queryprogram router. An over-long
2829 return line was truncated at 256 instead of 255 characters, thereby
2830 overflowing its buffer with the terminating zero. As well as fixing this, I
2831 have increased the buffer size to 1024 (and made a note to document this).
2833 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2834 when Exim is waiting for an SMTP response from a remote server, Exim
2835 restarts its select() call on the socket, thereby resetting its timeout.
2836 This is not a problem when such interrupts are rare. Somebody set up a cron
2837 job to run exiwhat every 2 minutes, which is less than the normal select()
2838 timeout (5 or 10 minutes). This meant that the select() timeout never
2839 kicked in because it was always reset. I have fixed this by comparing the
2840 time when an interrupt arrives with the time at the start of the first call
2841 to select(). If more time than the timeout has elapsed, the interrupt is
2842 treated as a timeout.
2844 5. Some internal re-factoring in preparation for the addition of Sieve
2845 extensions (by MH). In particular, the "personal" test is moved to a
2846 separate function, and given an option for scanning Cc: and Bcc: (which is
2847 not set for Exim filters).
2849 6. When Exim created an email address using the login of the caller as the
2850 local part (e.g. when creating a From: or Sender: header line), it was not
2851 quoting the local part when it contained special characters such as @.
2853 7. Installed new OpenBSD configuration files.
2855 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2856 try to make them clearer.
2858 9. Callout options, other than the timeout value, were being ignored when
2859 verifying sender addresses in header lines. For example, when using
2861 verify = header_sender/callout=no_cache
2863 the cache was (incorrectly) being used.
2865 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2866 only the Cygwin environment.
2868 11. When return_path_on_delivery was set as a log selector, if different remote
2869 addresses in the same message used different return paths and parallel
2870 remote delivery occurred, the wrong values would sometimes be logged.
2871 (Whenever a remote delivery process finished, the return path value from
2872 the most recently started remote delivery process was logged.)
2874 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2875 lines when AUTH and/or TLS are in use. This is the "received protocol"
2876 field. Exim used to use "asmtp" for authenticated SMTP, without any
2877 indication (in the protocol name) for TLS use. Now it follows the RFC and
2878 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2879 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2880 names appear in log lines as well as in Received: header lines.
2882 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2883 extensions, and comparison tests, and to fix some bugs.
2885 14. Changes to the "personal" filter test:
2887 (1) The test was buggy in that it was just doing the equivalent of
2888 "contains" tests on header lines. For example, if a user's address was
2889 anne@some.where, the "personal" test would incorrectly be true for
2891 To: susanne@some.where
2893 This test is now done by extracting each address from the header in turn,
2894 and checking the entire address. Other tests that are part of "personal"
2895 are now done using regular expressions (for example, to check local parts
2896 of addresses in From: header lines).
2898 (2) The list of non-personal local parts in From: addresses has been
2899 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2900 taken from the Sieve specification recommendations.
2902 (3) If the message contains any header line starting with "List-" it is
2903 treated as non-personal.
2905 (4) The test for "circular" in the Subject: header line has been removed
2906 because it now seems ill-conceived.
2908 15. Minor typos in src/EDITME comments corrected.
2910 16. Installed latest exipick from John Jetmore.
2912 17. If headers_add on a router specified a text string that was too long for
2913 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2914 of string_sprintf() is now avoided.
2916 18. $message_body_size was not set (it was always zero) when running the DATA
2917 ACL and the local_scan() function.
2919 19. For the "mail" command in an Exim filter, no default was being set for
2920 the once_repeat time, causing a random time value to be used if "once" was
2921 specified. (If the value happened to be <= 0, no repeat happened.) The
2922 default is now 0s, meaning "never repeat". The "vacation" command was OK
2923 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2924 (I found it when inspecting the code).
2926 20. There is now an overall timeout for performing a callout verification. It
2927 defaults to 4 times the callout timeout, which applies to individual SMTP
2928 commands during the callout. The overall timeout applies when there is more
2929 than one host that can be tried. The timeout is checked before trying the
2930 next host. This prevents very long delays if there are a large number of
2931 hosts and all are timing out (e.g. when the network connections are timing
2932 out). The value of the overall timeout can be changed by specifying an
2933 additional sub-option for "callout", called "maxwait". For example:
2935 verify = sender/callout=5s,maxwait=20s
2937 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2938 to the end before writing, but this should make it even safer).
2940 22. Exim was forgetting that it had advertised PIPELINING for the second and
2941 subsequent messages on an SMTP connection. It was also not resetting its
2942 memory on STARTTLS and an internal HELO.
2944 23. When Exim logs an SMTP synchronization error within a session, it now
2945 records whether PIPELINING has been advertised or not.
2947 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2948 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2949 rather than long int.
2951 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2953 26. Added the never_mail option to autoreply.
2959 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2960 crash if the getsockname() call failed; this can happen if a connection is
2961 closed very soon after it is established. The problem was simply in the
2962 order in which certain operations were done, causing Exim to try to write
2963 to the SMTP stream before it had set up the file descriptor. The bug has
2964 been fixed by making things happen in the correct order.
2970 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2971 before the connection was closed, thus losing the rejection response.
2973 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2974 some early Solaris releases, but causes trouble in current releases where
2975 socklen_t is defined.
2977 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2980 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2983 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2984 that releasing the top of it at the end releases what was used for sub-
2985 expansions (unless the block got too big). However, discard this block if
2986 the first thing is a variable or header, so that we can use its block when
2987 it is dynamic (useful for very large $message_headers, for example).
2989 6. Lookups now cache *every* query, not just the most recent. A new, separate
2990 store pool is used for this. It can be recovered when all lookup caches are
2991 flushed. Lookups now release memory at the end of their result strings.
2992 This has involved some general refactoring of the lookup sources.
2994 7. Some code has been added to the store_xxx() functions to reduce the amount
2995 of flapping under certain conditions.
2997 8. log_incoming_interface used to affect only the <= reception log lines. Now
2998 it causes the local interface and port to be added to several more SMTP log
2999 lines, for example "SMTP connection from", and rejection lines.
3001 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3003 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3005 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3006 could be overwritten at the end of the current message (or the start of a
3007 new message if it was set in a HELO ACL). The value is now preserved for
3008 the duration of the SMTP connection.
3010 12. If a transport had a headers_rewrite setting, and a matching header line
3011 contained an unqualified address, that address was qualified, even if it
3012 did not match any rewriting rules. The underlying bug was that the values
3013 of the flags that permit the existence of unqualified sender and recipient
3014 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3015 non-local messages, and by -bnq for local messages) were not being
3016 preserved with the message after it was received.
3018 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3019 "next input=" as part of the text comprising the host identity instead of
3020 the correct text. The code was using the same buffer for two different
3021 strings. However, depending on which order the printing function evaluated
3022 its arguments, the bug did not always show up. Under Linux, for example, my
3023 test suite worked just fine.
3025 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3026 doesn't work with some older versions of Perl. It has been changed to "my",
3027 which in any case is probably the better facility to use.
3029 15. A really picky compiler found some instances of statements for creating
3030 error messages that either had too many or two few arguments for the format
3033 16. The size of the buffer for calls to the DNS resolver has been increased
3034 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3035 for addresses that have a lot of PTR records. This alleviates a problem; it
3036 does not fully solve it.
3038 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3039 buffer now truncates the list and logs the incident, which is the same
3040 action as happens when Exim is looking up a host name and its aliases.
3041 Previously in this situation something unpredictable would happen;
3042 sometimes it was "internal error: store_reset failed".
3044 18. If a server dropped the connection unexpectedly when an Exim client was
3045 using GnuTLS and trying to read a response, the client delivery process
3046 crashed while trying to generate an error log message.
3048 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3049 string, for example:
3051 warn message = H1: something\nH2: something
3053 the text was added as a single header line from Exim's point of view
3054 though it ended up OK in the delivered message. However, searching for the
3055 second and subsequent header lines using $h_h2: did not work. This has been
3056 fixed. Similarly, if a system filter added multiple headers in this way,
3057 the routers could not see them.
3059 20. Expanded the error message when iplsearch is called with an invalid key to
3060 suggest using net-iplsearch in a host list.
3062 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3063 ACLs are no longer actually imposed (and a message to that effect is
3066 22. If a "gecos" field in a passwd entry contained escaped characters, in
3067 particular, if it contained a \" sequence, Exim got it wrong when building
3068 a From: or a Sender: header from that name. A second bug also caused
3069 incorrect handling when an unquoted " was present following a character
3070 that needed quoting.
3072 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3073 was not being matched caselessly.
3075 24. Arranged for all hyphens in the exim.8 source to be escaped with
3078 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3079 properly. Recipient callout cache records were still being keyed to include
3080 the sender, even when use_sender was set false. This led to far more
3081 callouts that were necessary. The sender is no longer included in the key
3082 when use_sender is false.
3084 26. Added "control = submission" modifier to ACLs.
3086 27. Added the ${base62d: operator to decode base 62 numbers.
3088 28. dnsdb lookups can now access SRV records.
3090 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3091 the configuration file.
3093 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3094 (-v) mode. This makes the output for a verbose queue run more intelligible.
3096 31. Added a use_postmaster feature to recipient callouts.
3098 32. Added the $body_zerocount variable, containing the number of binary zero
3099 bytes in the message body.
3101 33. The time of last modification of the "new" subdirectory is now used as the
3102 "mailbox time last read" when there is a quota error for a maildir
3105 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3107 35. Added +ignore_unknown as a special item in host lists.
3109 36. Code for decoding IPv6 addresses in host lists is now included, even if
3110 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3111 address was recognized as an IP address, but was then not correctly decoded
3112 into binary, causing unexpected and incorrect effects when compared with
3119 1. Very minor rewording of debugging text in manualroute to say "list of
3120 hosts" instead of "hostlist".
3122 2. If verify=header_syntax was set, and a header line with an unqualified
3123 address (no domain) and a large number of spaces between the end of the
3124 name and the colon was received, the reception process suffered a buffer
3125 overflow, and (when I tested it) crashed. This was caused by some obsolete
3126 code that should have been removed. The fix is to remove it!
3128 3. When running in the test harness, delay a bit after writing a bounce
3129 message to get a bit more predictability in the log output.
3131 4. Added a call to search_tidyup() just before forking a reception process. In
3132 theory, someone could use a lookup in the expansion of smtp_accept_max_
3133 per_host which, without the tidyup, could leave open a database connection.
3135 5. Added the variables $recipient_data and $sender_data which get set from a
3136 lookup success in an ACL "recipients" or "senders" condition, or a router
3137 "senders" option, similar to $domain_data and $local_part_data.
3139 6. Moved the writing of debug_print from before to after the "senders" test
3142 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3143 problems for message scanning, either using a data ACL, or using
3144 local_scan() because the Received: header was not generated till after they
3145 were called (in order to set the time as the time of reception completion).
3146 I have revised the way this works. The header is now generated after the
3147 body is received, but before the ACL or local_scan() are called. After they
3148 are run, the timestamp in the header is updated.
3154 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3155 before starting a queue runner without re-exec. This happened only when
3156 deliver_drop_privilege was set or when the Exim user was set to root. The
3157 effect of the bug was that timeouts during subsequent deliveries caused
3158 crashes instead of being properly handled. The handler is now left at its
3159 default (and expected) setting.
3161 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3162 message, again when deliver_drop_privilege is set or Exim is run as root.
3163 The bug described in (1) was not present in this case, but the tidying up
3164 of the other signals was missing. I have made the two cases consistent.
3166 3. The ignore_target_hosts setting on a manualroute router was being ignored
3167 for hosts that were looked up using the /MX notation.
3169 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3172 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3173 operated on the sender address. After changing the $sender_address to <>
3174 for the sender address verify, Exim was re-instated it as the original
3175 (before rewriting) address, but remembering that it had rewritten it, so it
3176 wasn't rewriting it again. This bug also had the effect of breaking the
3177 sender address verification caching when the sender address was rewritten.
3179 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3180 This has been changed so that if the ip literal address matches
3181 ignore_target_hosts, the router declines.
3183 7. Added expansion conditions match_domain, match_address, and match_local_
3184 part (NOT match_host).
3186 8. The placeholder for the Received: header didn't have a length field set.
3188 9. Added code to Exim itself and to exim_lock to test for a specific race
3189 condition that could lead to file corruption when using MBX delivery. The
3190 issue is with the lockfile that is created in /tmp. If this file is removed
3191 after a process has opened it but before that process has acquired a lock,
3192 there is the potential for a second process to recreate the file and also
3193 acquire a lock. This could lead to two Exim processes writing to the file
3194 at the same time. The added code performs the same test as UW imapd; it
3195 checks after acquiring the lock that its file descriptor still refers to
3196 the same named file.
3198 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3199 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3202 11. Added the smtp_active_hostname option. If used, this will typically be made
3203 to depend on the incoming interface address. Because $interface_address is
3204 not set up until the daemon has forked a reception process, error responses
3205 that can happen earlier (such as "too many connections") no longer contain
3208 12. If an expansion in a condition on a "warn" statement fails because a lookup
3209 defers, the "warn" statement is abandoned, and the next ACL statement is
3210 processed. Previously this caused the whole ACL to be aborted.
3212 13. Added the iplsearch lookup type.
3214 14. Added ident_timeout as a log selector.
3216 15. Added tls_certificate_verified as a log selector.
3218 16. Added a global option tls_require_ciphers (compare the smtp transport
3219 option of the same name). This controls incoming TLS connections.
3221 17. I finally figured out how to make tls_require_ciphers do a similar thing
3222 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3223 before starting the TLS session.
3225 18. Tabs are now shown as \t in -bP output.
3227 19. If the log selector return_path_on_delivery was set, Exim crashed when
3228 bouncing a message because it had too many Received: header lines.
3230 20. If two routers both had headers_remove settings, and the first one included
3231 a superfluous trailing colon, the final name in the first list and the
3232 first name in the second list were incorrectly joined into one item (with a
3233 colon in the middle).
3239 1. Added -C and -D options to the exinext utility, mainly to make it easier
3240 to include in the automated testing, but these could be helpful when
3241 multiple configurations are in use.
3243 2. The exinext utility was not formatting the output nicely when there was
3244 an alternate port involved in the retry record key, nor when there was a
3245 message id as well (for retries that were specific to a specific message
3246 and a specific host). It was also confused by IPv6 addresses, because of
3247 the additional colons they contain. I have fixed the IPv4 problem, and
3248 patched it up to do a reasonable job for IPv6.
3250 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3251 delivery, the log line now contains "pipelined" if PIPELINING was used.
3253 4. An SMTP transport process used to panic and die if the bind() call to set
3254 an explicit outgoing interface failed. This has been changed; it is now
3255 treated in the same way as a connect() failure.
3257 5. A reference to $sender_host_name in the part of a conditional expansion
3258 that was being skipped was still causing a DNS lookup. This no longer
3261 6. The def: expansion condition was not recognizing references to header lines
3262 that used bh_ and bheader_.
3264 7. Added the _cache feature to named lists.
3266 8. The code for checking quota_filecount in the appendfile transport was
3267 allowing one more file than it should have been.
3269 9. For compatibility with Sendmail, the command line option
3277 and sets the incoming protocol and host name (for trusted callers). The
3278 host name and its colon can be omitted when only the protocol is to be set.
3279 Note the Exim already has two private options, -pd and -ps, that refer to
3280 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3281 "s", but I don't think that's a major issue.
3283 10. A number of refactoring changes to the code, none of which should affect
3286 (a) The number of logging options was getting close to filling up the
3287 32-bit word that was used as a bit map. I have split them into two classes:
3288 those that are passed in the argument to log_write(), and those that are
3289 only ever tested independently outside of that function. These are now in
3290 separate 32-bit words, so there is plenty of room for expansion again.
3291 There is no change in the user interface or the logging behaviour.
3293 (b) When building, for example, log lines, the code previously used a
3294 macro that called string_cat() twice, in order to add two strings. This is
3295 not really sufficiently general. Furthermore, there was one instance where
3296 it was actually wrong because one of the argument was used twice, and in
3297 one call a function was used. (As it happened, calling the function twice
3298 did not affect the overall behaviour.) The macro has been replaced by a
3299 function that can join an arbitrary number of extra strings onto a growing
3302 (c) The code for expansion conditions now uses a table and a binary chop
3303 instead of a serial search (which was left over from when there were very
3304 few conditions). Also, it now recognizes conditions like "pam" even when
3305 the relevant support is not compiled in: a suitably worded error message is
3306 given if an attempt is made to use such a condition.
3308 11. Added ${time_interval:xxxxx}.
3310 12. A bug was causing one of the ddress fields not to be passed back correctly
3311 from remote delivery subprocesses. The field in question was not being
3312 subsequently used, so this caused to problems in practice.
3314 13. Added new log selectors queue_time and deliver_time.
3316 14. Might have fixed a bug in maildirsizefile handling that threw up
3317 "unexpected character" debug warnings, and recalculated the data
3318 unnecessarily. In any case, I expanded the warning message to give more
3321 15. Added the message "Restricted characters in address" to the statements in
3322 the default ACL that block characters like @ and % in local parts.
3324 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3325 Three changes have been made:
3327 (a) There was a serious bug; a negative response to MAIL caused the whole
3328 recipient domain to be cached as invalid, thereby blocking all messages
3329 to all local parts at the same domain, from all senders. This bug has
3330 been fixed. The domain is no longer cached after a negative response to
3331 MAIL if the sender used is not empty.
3333 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3336 (c) A new callout option, "use_sender" has been added for people who want
3337 the modified behaviour.
3343 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3346 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3347 indeed breaks things for older releases.
3349 3. Added additional logging to the case where there is a problem reading data
3350 from a filter that is running in a subprocess using a pipe, in order to
3351 try to track down a specific problem.
3353 4. Testing facility fudge: when running in the test harness and attempting
3354 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3355 getting "No route to host". Convert this to a timeout.
3357 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3360 6. Some OS don't have socklen_t but use size_t instead. This affects the
3361 fifth argument of getsockopt() amongst other things. This is now
3362 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3363 can be set for individual OS. I have set it for SunOS5, OSF1, and
3364 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3365 some earlier ones do not.
3367 7. Change 4.30/15 was not doing the test caselessly.
3369 8. The standard form for an IPv6 address literal was being rejected by address
3370 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3371 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3372 this, as well as the form without the "IPv6" on the front (but only when
3373 address literals are enabled, of course).
3375 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3377 10. Exim crashed if a message with an empty sender address specified by -f
3378 encountered a router with an errors_to setting. This could be provoked only
3379 by a command such as
3383 where an empty string was supplied; "<>" did not hit this bug.
3385 11. Installed PCRE release 4.5.
3387 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3388 remained set. It is now erased.
3390 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3391 times from message ids (which are base 36 rather than the normal 62).
3393 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3394 were being counted as actual protocol errors, and logged if the log
3395 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3396 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3397 and DATA following a set of rejected RCPTs do not count as protocol errors.
3398 In other words, Exim assumes they were pipelined, though this may not
3399 actually be the case. Of course, in all cases the client gets an
3400 appropriate error code.
3402 15. If a lookup fails in an ACL condition, a message about the failure may
3403 be available; it is used if testing the ACL cannot continue, because most
3404 such messages specify what the cause of the deferral is. However, some
3405 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3406 that caused an old message to be retained and used if a later statement
3407 caused a defer, replacing the real cause of the deferral.
3409 16. If an IP address had so many PTR records that the DNS lookup buffer
3410 was not large enough to hold them, Exim could crash while trying to process
3411 the truncated data. It now detects and logs this case.
3413 17. Further to 4.21/58, another change has been made: if (and only if) the
3414 first line of a message (the first header line) ends with CRLF, a bare LF
3415 in a subsequent header line has a space inserted after it, so as not to
3416 terminate the header.
3418 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3419 unnecessarily, used atoi() instead of strtol(), and didn't check the
3420 termination when getting file sizes from file names by regex.
3422 19. Completely re-implemented the support for maildirsize files, in the light
3423 of a number of problems with the previous contributed implementation
3424 (4.30/29). In particular:
3426 . If the quota is zero, the maildirsize file is maintained, but no quota is
3429 . If the maildir directory does not exist, it is created before any attempt
3430 to write a maildirsize file.
3432 . The quota value in the file is just a cache; if the quota is changed in
3433 the transport, the new value overrides.
3435 . A regular expression is available for excluding directories from the
3438 20. The autoreply transport checks the characters in options that define the
3439 message's headers; it allows continued headers, but it was checking with
3440 isspace() after an embedded newline instead of explicitly looking for a
3443 21. If all the "regular" hosts to which an address was routed had passed their
3444 expiry times, and had not reached their retry times, the address was
3445 bounced, even if fallback hosts were defined. Now Exim should go on to try
3448 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3449 equivalent code in the SMTP transport. Some hosts send humungous responses
3450 to HELO/EHLO, more than 1024 it seems.
3452 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3453 gives compiler warnings in some environments. I've now done it "properly",
3456 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3457 (because of problems with the built-in one) was declared to return uschar *
3458 instead of char *, causing compiler failure.
3460 25. Fixed a file descriptor leak when processing alias/forward files.
3462 26. Fixed a minor format string issue in dbfn.c.
3464 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3466 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3467 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3470 29. When a sender address is verified, it is cached, to save repeating the test
3471 when there is more than one recipient in a message. However, when the
3472 verification involves a callout, it is possible for different callout
3473 options to be set for different recipients. It is too complicated to keep
3474 track of this in the cache, so now Exim always runs a verification when a
3475 callout is required, relying on the callout cache for the optimization.
3476 The overhead is duplication of the address routing, but this should not be
3479 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3480 to be verified with callout=postmaster, and the main callout worked but the
3481 postmaster check failed, the verification correctly failed. However, if a
3482 subsequent RCPT command asked for sender verification *without* the
3483 postmaster check, incorrect caching caused this verification also to fail,
3486 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3487 it was not caching the DNS options (qualify_single, search_parents) that
3488 were used when the lookup failed. A subsequent lookup with different
3489 options therefore always gave the same answer, though there were cases
3490 where it should not have. (Example: a "domains = !$mx_any" option on a
3491 dnslookup router: the "domains" option is always processed without any
3492 widening, but the router might have qualify_single set.) Now Exim uses the
3493 cached value only when the same options are set.
3495 32. Added John Jetmore's "exipick" utility to the distribution.
3497 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3498 than a timeout (e.g. a certificate is required, and is not provided), an
3499 Exim server now closes the connection immediately. Previously it waited for
3500 the client to close - but if the client is SSL, it seems that they each
3501 wait for each other, leading to a delay before one of them times out.
3503 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3504 maintained 0.8.x compatibility because I don't think many are using it, and
3505 it is clearly obsolete.
3507 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3510 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3511 client certificate was expired. A simple patch fixes this, though I don't
3512 understand the full logic of why the verify callback is called multiple
3515 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3516 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3517 which causes problems with some clients (such as the Certicom SSL Plus
3518 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3519 disables the coutermeasure allowing Eudora to connect."
3521 38. Exim was not checking that a write() to a log file succeeded. This could
3522 lead to Bad Things if a log got too big, in particular if it hit a file
3523 size limit. Exim now panics and dies if it cannot write to a log file, just
3524 as it does if it cannot open a log file.
3526 39. Modified OS/Makefile-Linux so that it now contains
3528 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3530 The two -D definitions ensure that Exim is compiled with large file
3531 support, which makes it possible to handle log files that are bigger than
3534 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3535 instance) a domain was checked against a named list that involved a lookup,
3536 causing $domain_data to be set, then another domain was checked against the
3537 same list, then the first domain was re-checked, the value of $domain_data
3538 after the final check could be wrong. In particular, if the second check
3539 failed, it could be set empty. This bug probably also applied to
3542 41. The strip_trailing_dot option was not being applied to the address given
3543 with the -f command-line option.
3545 42. The code for reading a message's header from the spool was incrementing
3546 $received_count, but never initializing it. This meant that the value was
3547 incorrect (doubled) while delivering a message in the same process in which
3548 it was received. In the most common configuration of Exim, this never
3549 happens - a fresh exec is done - but it can happen when
3550 deliver_drop_privilege is set.
3552 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3553 it now includes up to 150 characters of the unexpected data in the log
3556 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3557 and building data strings. The size of both of these buffers was 10 000
3558 bytes - far larger than anybody would *ever* want, thought I. Needless to
3559 say, somebody hit the limit. I have increased the maximum line length to
3560 20 000 and the maximum data length of concatenated lines to 100 000. I have
3561 also fixed two bugs, because there was no checking on these buffers. Tsk,
3562 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3563 buffer is too small.
3565 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3566 lsearch lookups. Now it does.
3568 46. When parsing a route_list item in a manualroute router, a fixed-length
3569 buffer was used for the list of hosts. I made this 1024 bytes long,
3570 thinking that nobody would ever have a list of hosts that long. Wrong.
3571 Somebody had a whole pile of complicated expansion conditions, and the
3572 string was silently truncated, leading to an expansion error. It turns out
3573 that it is easier to change to an unlimited length (owing to other changes
3574 that have happened since this code was originally written) than to build
3575 structure for giving a limitation error. The length of the item that
3576 expands into the list of hosts is now unlimited.
3578 47. The lsearch lookup could not handle data where the length of text line was
3579 more than 4095 characters. Such lines were truncated, leading to shortened
3580 data being returned. It should now handle lines of any length.
3582 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3583 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3586 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3587 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3588 no longer happens in comment lines. A list of replacements is now placed
3589 at the head of all of the source files, except those whose only change is
3590 to replace PERL_COMMAND in the very first #! line.
3592 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3593 messages on the queue, with a bottom-up merge sort, using code contributed
3594 by Michael Haardt. This should make operations like -bp somewhat faster on
3595 large queues. It won't affect queue runners, except when queue_run_in_order
3598 51. Installed eximstats 1.31 in the distribution.
3600 52. Added support for SRV lookups to the dnslookup router.
3602 53. If an ACL referred to $message_body or $message_body_end, the value was not
3603 reset for any messages that followed in the same SMTP session.
3605 54. The store-handling optimization for building very long strings was not
3606 differentiating between the different store pools. I don't think this
3607 actually made any difference in practice, but I've tidied it.
3609 55. While running the routers to verify a sender address, $sender_address
3610 was still set to the sender address. This is wrong, because when routing to
3611 send a bounce to the sender, it would be empty. Therefore, I have changed
3612 it so that, while verifying a sender address, $sender_address is set to <>.
3613 (There is no change to what happens when verifying a recipient address.)
3615 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3616 target A or AAAA records (if not already returned) without resetting the
3617 qualify_single or search_parents options of the DNS resolver. These are
3618 inappropriate in this case because the targets of MX and SRV records must
3619 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3620 target that, when qualified, matched something in the local domain. These
3621 two options are now turned off when doing these lookups.
3623 57. It seems that at least some releases of Reiserfs (which does not have the
3624 concept of a fixed number of inodes) returns zero and not -1 for the
3625 number of available inodes. This interacted badly with check_spool_inodes,
3626 which assumed that -1 was the "no such thing" setting. What I have done is
3627 to check that the total number of inodes is greater than zero before doing
3628 the test of how many are available.
3630 58. When a "warn" ACL statement has a log_message modifier, the message is
3631 remembered, and not repeated. This is to avoid a lot of repetition when a
3632 message has many recipients that cause the same warning to be written.
3633 Howewer, Exim was preserving the list of already written lines for an
3634 entire SMTP session, which doesn't seem right. The memory is now reset if a
3635 new message is started.
3637 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3638 debugging output unless log_rewrite was also set.
3640 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3641 of (int)(handle) when we know that handle contains (void *)(-1).
3643 61. The Exim daemon panic-logs an error return when it closes the incoming
3644 connection. However "connection reset by peer" seems to be common, and
3645 isn't really an error worthy of noting specially, so that particular error
3648 62. When Exim is trying to find all the local interfaces, it used to panic and
3649 die if the ioctl to get the interface flags failed. However, it seems that
3650 on at least one OS (Solaris 9) it is possible to have an interface that is
3651 included in the list of interfaces, but for which you get a failure error
3652 for this call. This happens when the interface is not "plumbed" into a
3653 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3654 failure of the "get flags" call assumes that the interface is down.
3656 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3657 makes life easier for people who are doing arithmetic on fields extracted
3658 from dates, where you often get leading zeros that should not be
3659 interpreted as octal.
3661 64. Added qualify_domain to the redirect router, to override the global
3664 65. If a pathologically long header line contained very many addresses (the
3665 report of this problem mentioned 10 000) and each of them was rewritten,
3666 Exim could use up a very large amount of memory. (It kept on making new
3667 copies of the header line as it rewrote, and never released the old ones.)
3668 At the expense of a bit more processing, the header rewriting function has
3669 been changed so that it no longer eats memory in this way.
3671 66. The generation of the Received: header has been moved from the time that a
3672 message starts to be received, to the time that it finishes. The timestamp
3673 in the Received: header should now be very close to that of the <= log
3674 line. There are two side-effects of this change:
3676 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3677 logged header lines no longer include the local Received: line, because
3678 it has not yet been created. The same applies to a copy of the message
3679 that is returned to a non-SMTP sender when a message is rejected.
3681 (b) When a filter file is tested using -bf, no additional Received: header
3682 is added to the test message. After some thought, I decided that this
3685 This change does not affect the value of $received_for. It is still set
3686 after address rewriting, but before local_scan() is called.
3688 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3690 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3691 gave an unhelpful panic error message, and a defer error. I have managed to
3692 change this behaviour so that it now rejects any supplied certificate,
3693 which seems right, as the list of acceptable certificates is empty.
3695 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3696 gave an unhelpful defer error. I have not managed to make this reject any
3697 supplied certificates, but the error message it gives is "no certificate
3698 supplied", which is not helpful.
3700 70. exigrep's output now also includes lines that are not associated with any
3701 message, but which match the given pattern. Implemented by a patch from
3702 Martin Sluka, which also tidied up the Perl a bit.
3704 71. Recipient callout verification, like sender verification, was using <> in
3705 the MAIL FROM command. This isn't really the right thing, since the actual
3706 sender may affect whether the remote host accepts the recipient or not. I
3707 have changed it to use the actual sender in the callout; this means that
3708 the cache record is now keyed on a recipient/sender pair, not just the
3709 recipient address. There doesn't seem to be a real danger of callout loops,
3710 since a callout by the remote host to check the sender would use <>.
3711 [SEE ABOVE: changed after hitting problems.]
3713 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3714 temporary errors. However, in the case of such a code being given after
3715 the end of a data transmission (i.e. after ".") Exim was failing to write
3716 a retry record for the message. (Yes, there was some broken host that was
3717 actually sending 8xx at this point.)
3719 73. An unknown lookup type in a host list could cause Exim to panic-die when
3720 the list was checked. (An example that provoked this was putting <; in the
3721 middle of a list instead of at the start.) If this happened during a DATA
3722 ACL check, a -D file could be left lying around. This kind of configuration
3723 error no longer causes Exim to die; instead it causes a defer errror. The
3724 incident is still logged to the main and panic logs.
3726 74. Buglet left over from Exim 3 conversion. The message "too many messages
3727 in one connection" was written to the rejectlog but not the mainlog, except
3728 when address rewriting (yes!) was being logged.
3730 75. Added write_rejectlog option.
3732 76. When a system filter was run not as root (that is, when system_filter_user
3733 was set), the values of the $n variables were not being returned to the
3734 main process; thus, they were not subsequently available in the $sn
3737 77. Added +return_path_on_delivery log selector.
3739 78. A connection timeout was being treated differently from recipients deferred
3740 when testing hosts_max_try with a message that was older than the host's
3741 retry timeout. (The host should not be counted, thus allowing all hosts to
3742 be tried at least once before bouncing.) This may have been the cause of an
3743 occasionally reported bug whereby a message would remain on the queue
3744 longer than the retry timeout, but would be bounced if a delivery was
3745 forced. I say "may" because I never totally pinned down the problem;
3746 setting up timeout/retry tests is difficult. See also the next item.
3748 79. The ultimate address timeout was not being applied to errors that involved
3749 a combination of host plus message (for example, a timeout on a MAIL
3750 command). When an address resolved to a number of possible hosts, and they
3751 were not all tried for each delivery (e.g. because of hosts_max_try), a
3752 message could remain on the queue longer than the retry timeout.
3754 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3757 81. Fixed an obscure SMTP outgoing bug which required at least the following
3758 conditions: (a) there was another message waiting for the same server;
3759 (b) the server returned 5xx to all RCPT commands in the first message so
3760 that the message was not completed; (c) the server dropped the connection
3761 or gave a negative response to the RSET that Exim sends to abort the
3762 transaction. The observed case was a dropped connection after DATA that had
3763 been sent in pipelining mode. That is, the server had advertised PIPELINING
3764 but was not implementing it correctly. The effect of the bug was incorrect
3765 behaviour, such as trying another host, and this could lead to a crash.
3771 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3772 and daemon.c were passed as pointers to ints; they should have been
3773 pointers to socklen_t variables (which are typically unsigned ints).
3775 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3778 3. Fixed a really odd bug that affected only the testing scheme; patching a
3779 certain fixed string in the binary changed the value of another string that
3780 happened to be identical to the end of the original first string.
3782 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3783 name", it returns that address as the IP address. On some operating
3784 systems (e.g. Solaris), it also passes back the IP address string as the
3785 "host name". However, on others (e.g. Linux), it passes back an empty
3786 string. Exim wasn't checking for this, and was changing the host name to an
3787 empty string, assuming it had been canonicized.
3789 5. Although rare, it is permitted to have more than one PTR record for a given
3790 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3791 all the names associated with an address, because they do in Solaris.
3792 However, it seems that they do not in Linux for data that comes from the
3793 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3794 I found this out when I moved to a new Linux workstation and tried to run
3795 the Exim test suite.
3797 To get round this problem I have changed the code so that it now does its
3798 own call to the DNS to look up PTR records when searching for a host name.
3799 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3800 addresses that are only in /etc/hosts are still found.
3802 This behaviour is, however, controlled by an option called host_lookup_
3803 order, which defaults to "bydns:byaddr". If people want to use the other
3804 order, or indeed, just use one or the other means of lookup, they can
3805 specify it in this variable.
3807 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3808 some operating systems, this comes back from gethostbyaddr() as an empty
3809 string, and this is what Exim used to test for. However, it seems that in
3810 other systems, "." is yielded. Exim now tests for this case too.
3812 7. The values of check_spool_space and check_log_space are now held internally
3813 as a number of kilobytes instead of an absolute number of bytes. If a
3814 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3815 kilobyte. This means that much larger values can be stored.
3817 8. Exim monitor: an attempt to get the action menu when not actually pointing
3818 at a message produces an empty menu entitled "No message selected". This
3819 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3820 no entries in it ("Shell widget menu has zero width and/or height"). So I
3821 have added a single, blank menu entry in this case.
3823 9. Added ${quote_local_part.
3825 10. MIME decoding is now applied to the contents of Subject: header lines when
3828 11. Now that a reference to $sender_host_address automatically causes a reverse
3829 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3830 host lookup before query-style lookups in lists that might use this
3831 variable. This has therefore been abolished, and the "net-" prefix is no
3832 longer necessary for query-style lookups.
3834 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3835 have been a typo for LFLAGS, so it has been changed.
3837 13. The install script calls Exim with "-C /dev/null" in order to find the
3838 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3839 to be output. Howeve, since Exim outputs its version number before the
3840 error, it didn't break the script. It just looked ugly. I fixed this by
3841 always allowing "-C /dev/null" if the caller is root.
3843 14. Ignore overlarge ACL variable number when reading spool file - insurance
3844 against a later release with more variables having written the file.
3846 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3847 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3848 this, as well as the form without the "IPv6" on the front.
3850 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3851 OS/Makefile-Darwin file.
3853 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3854 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3856 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3857 variable. However, some OS use ints for inodes. I've added cast to long int
3858 to get rid of the compiler warning.
3860 19. I had forgotten to lock out "/../" in configuration file names when
3861 ALT_CONFIG_PREFIX was set.
3863 20. Routers used for verification do not need to specify transports. However,
3864 if such a router generated a host list, and callout was configured, Exim
3865 crashed, because it could not find a port number from the (non-existent)
3866 transport. It now assumes port 25 in this circumstance.
3868 21. Added the -t option to exigrep.
3870 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3871 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3872 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3873 them is set, LOOKUP_LSEARCH is forced.
3875 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3877 24. Added sender and host information to the "rejected by local_scan()" log
3878 line; previously there was no indication of these.
3880 25. Added .include_if_exists.
3882 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3883 Linux. It turns out that not all file systems support this. Apparently some
3884 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3885 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3886 not supported on the file descriptor, is now ignored when Exim is trying to
3887 sync a directory. This applies only to Linux.
3889 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3891 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3892 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3895 29. Added support for maildirsize files from supplied patch (modified a bit).
3897 30. The use of :fail: followed by an empty string could lead Exim to respond to
3898 sender verification failures with (e.g.):
3900 550 Verification failed for <xxx>
3901 550 Sender verify failed
3903 where the first response line was missing the '-' that indicates it is not
3904 the final line of the response.
3906 31. The loop for finding the name of the user that called Exim had a hardwired
3907 limit of 10; it now uses the value of finduser_retries, which is used for
3908 all other user lookups.
3910 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3913 33. Exim was neglecting to zero errno before one call of strtol() when
3914 expanding a string and expecting an integer value. On some systems this
3915 resulted in spurious "integer overflow" errors. Also, it was casting the
3916 result into an int without checking.
3918 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3919 did not work. The code looks as if it has *never* worked, though it appears
3920 to have been documented since at least releast 1.62. I have made it work.
3922 35. The "timeout_DNS" error in retry rules, also documented since at least
3923 1.62, also never worked. As it isn't clear exactly what this means, and
3924 clearly it isn't a major issue, I have abolished the feature by treating it
3925 as "timeout", and writing a warning to the main and panic logs.
3927 36. The display of retry rules for -brt wasn't always showing the error code
3930 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3931 timeout_connect_A, timeout_connect_MX.
3933 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3934 to the empty sender.
3936 39. The daemon was not analysing the content of -oX till after it had closed
3937 stderr and disconnected from the controlling terminal. This meant that any
3938 syntax errors were only noted on the panic log, and the return code from
3939 the command was 0. By re-arranging the code a little, I've made the
3940 decoding happen first, so such errors now appear on stderr, and the return
3941 code is 1. However, the actual setting up of the sockets still happens in
3942 the disconnected process, so errors there are still only recorded on the
3945 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3946 connections (as happens on some IP stacks) was logged at start up time as
3947 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3948 it from "IPv6 and IPv4", which means that two separate sockets are being
3951 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3952 says whether AF_INET or AF_INET6 was passed as an argument.
3954 42. Exiwhat output was messed up when time zones were included in log
3957 43. Exiwhat now gives more information about the daemon's listening ports,
3958 and whether -tls-on-connect was used.
3960 44. The "port" option of the smtp transport is now expanded.
3962 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3963 silently ignored. Now an error message is written to the main and panic
3966 46. There's a new ACL modifier called "logwrite" which writes to a log file
3967 as soon as it is encountered.
3969 47. Added $local_user_uid and $local_user_gid at routing time.
3971 48. Exim crashed when trying to verify a sender address that was being
3974 49. Exim was recognizing only a space character after ".include". It now also
3975 recognizes a tab character.
3977 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3978 extracting the relevant information from the specification. The man page no
3979 longer contains scrambled data for the -d option, and I've added a section
3980 at the front about calling Exim under different names.
3982 51. Added "extra_headers" argument to the "mail" command in filter files.
3984 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3987 53. Installed eximstats 1.29.
3989 54. Added transport_filter_timeout as a generic transport option.
3991 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3992 Cc: header lines. This was required by RFC 822, but it not required by RFC
3995 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3996 incoming messages that did not have them. Now it does so only if the
3997 message originates locally, that is, if there is no associated remote host
3998 address. When Resent- header lines are present, this applies to the Resent-
3999 lines rather than the non-Resent- lines.
4001 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4002 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4004 58. Messages for configuration errors now include the name of the main
4005 configuration file - useful now that there may be more than one file in a
4006 list (.included file names were always shown).
4008 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4009 for those rare installations that do not start the daemon as root or run it
4010 setuid root. I've cut out the call to initgroups() if the daemon is not
4013 60. The Exim user and group can now be bound into the binary as text strings
4014 that are looked up at the start of Exim's processing.
4016 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4018 62. Added $mailstore_basename variable.
4020 63. Installed patch to sieve.c from Michael Haardt.
4022 64. When Exim failed to open the panic log after failing to open the main log,
4023 the original message it was trying to log was written to stderr and debug
4024 output, but if they were not available (the usual case in production), it
4025 was lost. Now it is written to syslog before the two lines that record the
4026 failures to open the logs.
4028 65. Users' Exim filters run in subprocesses under the user's uid. It is
4029 possible for a "deliver" command or an alias in a "personal" command to
4030 provoke an address rewrite. If logging of address rewriting is configured,
4031 this fails because the process is not running as root or exim. There may be
4032 a better way of dealing with this, but for the moment (because 4.30 needs
4033 to be released), I have disabled address rewrite logging when running a
4034 filter in a non-root, non-exim process.
4040 1. The buildconfig auxiliary program wasn't quoting the value set for
4041 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4042 not defined. This bug was masked in 4.22 by the effect that was fixed in
4045 2. Some messages that were rejected after a message id was allocated were
4046 shown as "incomplete" by exigrep. It no longer does this for messages that
4047 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4049 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4050 have allow_domain_literals set, the ID did not get logged in the <= line.
4051 Domain literals are now always recognized in Message-ID: header lines.
4053 4. The first argument for a ${extract expansion item is the key name or field
4054 number. Leading and trailing spaces in this item were not being ignored,
4055 causing some misleading effects.
4057 5. When deliver_drop_privilege was set, single queue runner processes started
4058 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4059 same command in the process it spins off) were not dropping privilege.
4061 6. When the daemon running as "exim" started a queue runner, it always
4062 re-executed Exim in the spun-off process. This is a waste of effort when
4063 deliver_drop_privilege is set. The new process now just calls the
4064 queue-runner function directly.
4070 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4073 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4074 ignored. Though the use of -oP was forcing the writing of a pid file, it
4075 was always written to the default place.
4077 3. If the message "no IP address found for host xxxx" is generated during
4078 incoming verification, it is now followed by identification of the incoming
4079 connection (so you can more easily find what provoked it).
4081 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4083 5. Added some features to "harden" Exim a bit more against certain attacks:
4085 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4086 be put in Local/Makefile. This is like the never_users runtime option,
4087 but it cannot be overridden. The default setting is "root".
4089 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4090 prefix string with which any file named in a -C command line option
4093 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4094 is retained for -C and -D only if the caller of Exim is root. Without
4095 it, the exim user may also use -C and -D and retain privilege.
4097 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4098 command line option is disabled.
4100 6. Macro names set by the -D option must start with an upper case letter, just
4101 like macro names defined in the configuration file.
4103 7. Added "dereference=" facility to LDAP.
4105 8. Two instances of the typo "uknown" in the source files are fixed.
4107 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4108 the Configure-Makefile script screwed up while processing it.
4110 10. Incorporated PCRE 4.4.
4112 11. The SMTP synchronization check was not operating right at the start of an
4113 SMTP session. For example, it could not catch a HELO sent before the client
4114 waited for the greeting. There is now a check for outstanding input at the
4115 point when the greeting is written. Because of the duplex, asynchronous
4116 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4117 way, but not yet received, when the check is performed.
4119 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4120 on TCP/IP sockets, because this apparently causes some broken clients to
4123 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4124 unchanged) from the Cygwin maintainer.
4126 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4127 was supported instead of testing for mbx. Effectively a typo.
4129 15. The spa authenticator server code was not checking that the input it
4130 received was valid base64.
4132 16. The debug output line for the "set" modifier in ACLs was not showing the
4133 name of the variable that was being set.
4135 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4137 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4138 lookup on its own. Something else was needed to trigger the lookup. For
4139 example, a match in host_lookup or the need for a host name in a host list.
4140 Now, if $sender_host_name is referenced and the host name has not yet been
4141 looked up, a lookup is performed. If the lookup fails, the variable remains
4142 empty, and $host_lookup_failed is set to "1".
4144 19. Added "eqi" as a case-independent comparison operator.
4146 20. The saslauthd authentication condition could segfault if neither service
4147 nor realm was specified.
4149 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4150 error message that was logged was misleading, and incoming SMTP
4151 connections were dropped. The message is now more accurate, and temporary
4152 errors are given to SMTP connections.
4154 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4155 (except RSET) with a 421 error, until QUIT is received. However, it was
4156 failing to send a response to QUIT.
4158 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4159 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4160 ACL is now run after the helo_try_verify_hosts code.
4162 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4163 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4164 case-independent, so other case variants are also recognized). Apparently
4165 some systems use these upper case variants.
4167 25. If more than two messages were waiting for the same host, and a transport
4168 filter was specified for the transport, Exim sent two messages over the
4169 same TCP/IP connection, and then failed with "socket operation on non-
4170 socket" when it tried to send the third.
4172 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4174 27. The extern definition of crypt16() in expand.c was not being excluded when
4175 the OS had its own crypt16() function.
4177 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4178 as a preferred synonym for return_size_limit, both as an option and as an
4181 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4183 30. Changed the default configuration ACL to relax the local part checking rule
4184 for addresses that are not in any local domains. For these addresses,
4185 slashes and pipe symbols are allowed within local parts, but the sequence
4186 /../ is explicitly forbidden.
4188 31. SPA server authentication was not clearing the challenge buffer before
4191 32. log_message in a "warn" ACL statement was writing to the reject log as
4192 well as to the main log, which contradicts the documentation and doesn't
4193 seem right (because no rejection is happening). So I have stopped it.
4195 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4196 However, I am unable to do any testing of this.
4198 34. Fixed an infelicity in the appendfile transport. When checking directories
4199 for a mailbox, to see if any needed to be created, it was accidentally
4200 using path names with one or more superfluous leading slashes; tracing
4201 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4203 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4204 contents are added to the log line that is written for every discarded
4205 recipient. (Previously a log_message setting was ignored.)
4207 36. The ${quote: operator now quotes the string if it is empty.
4209 37. The install script runs exim in order to find its version number. If for
4210 some reason other than non-existence or emptiness, which it checks, it
4211 could not run './exim', it was installing it with an empty version number,
4212 i.e. as "exim-". This error state is now caught, and the installation is
4215 38. An argument was missing from the function that creates an error message
4216 when Exim fails to connect to the socket for saslauthd authentication.
4217 This could cause Exim to crash, or give a corrupted message.
4219 39. Added isip, isip4, and isip6 to ${if conditions.
4221 40. The ACL variables $acl_xx are now saved with the message, and can be
4222 accessed later in routers, transports, and filters.
4224 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4225 strings in the file are not string-expanded.
4227 42. If a MAIL command specified a SIZE value that was too large to fit into an
4228 int variable, the check against message_size_limit failed. Such values are
4229 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4230 day this will have to be increased, but I don't think I want to be around
4231 when emails are that large.
4238 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4239 iconv() is not standard in FreeBSD.
4241 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4242 IPv6 enabled. The observed symptom was a segmentation fault on return from
4243 the function os_common_find_running_interfaces() in src/os.c.
4245 3. In the check_special_case() function in daemon.c I had used "errno" as an
4246 argument name, which causes warnings on some systems. This was basically a
4247 typo, since it was named "eno" in the comments!
4249 4. The code that waits for the clock to tick (at a resolution of some fraction
4250 of a second) so as to ensure message-id uniqueness was always waiting for
4251 at least one whole tick, when it could have waited for less. [This is
4252 almost certainly not relevant at current processor speeds, where it is
4253 unlikely to ever wait at all. But we try to future-proof.]
4255 5. The function that sleeps for a time interval that includes fractions of a
4256 second contained a race. It did not block SIGALRM between setting the
4257 timer, and suspending (a couple of lines later). If the interval was short
4258 and the sigsuspend() was delayed until after it had expired, the suspension
4259 never ended. On busy systems this could lead to processes getting stuck for
4262 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4263 process, before it forks any delivery processes. The open lookup caching
4264 mechanism meant that the open file or database connection was passed into
4265 the delivery process. The problem was that delivery processes always tidy
4266 up cached lookup data. This could cause a problem for the next delivery
4267 process started by the queue runner, because the external queue runner
4268 process does not know about the closure. So the next delivery process
4269 still has data in the lookup cache. In the case of a file lookup, there was
4270 no problem because closing a file descriptor in a subprocess doesn't affect
4271 the parent. However, if the lookup was caching a connection to a database,
4272 the connection was closed, and the second delivery process was likely to
4273 see errors such as "PGSQL: query failed: server closed the connection
4274 unexpectedly". The problem has been fixed by closing all cached lookups
4275 in a queue runner before running a delivery process.
4277 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4278 seem to have the "const" qualifier which it has on other OS. I've
4281 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4282 *of the same type* (client or server) with the same public name that an
4283 error should be diagnosed.
4285 9. When Exim looked up a host name for an IP address, but failed to find the
4286 original IP address when looking up the host name (a safety check), it
4287 output the message "<ip address> does not match any IP for NULL", which was
4288 confusing, to say the least. The bug was that the host name should have
4289 appeared instead of "NULL".
4291 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4292 user that is built into the binary, and the -C or -D options is used, root
4293 privilege is dropped before the configuration file is read. In addition,
4294 logging is switched to stderr instead of the normal log files. If the
4295 configuration then re-defines the Exim user, the unprivileged environment
4296 is probably not what is expected, so Exim logs a panic warning message (but
4299 However, if deliver_drop_privilege is set, the unprivileged state may well
4300 be exactly what is intended, so the warning has been cut out in that case,
4301 and Exim is allowed to try to write to its normal log files.
4307 1. smtp_return_error_details was not giving details for temporary sender
4308 or receiver verification errors.
4310 2. Diagnose a configuration error if two authenticators have the same public
4313 3. Exim used not to create the message log file for a message until the first
4314 delivery attempt. This could be confusing when incoming messages were held
4315 for policy or load reasons. The message log file is now created at the time
4316 the message is received, and an initial "Received" line is written to it.
4318 4. The automatically generated man page for command line options had a minor
4319 bug that caused no ill effects; however, a more serious problem was that
4320 the procedure for building the man page automatically didn't always
4321 operate. Consequently, release 4.20 contains an out-of-date version. This
4322 shouldn't happen again.
4324 5. When building Exim with embedded Perl support, the script that builds the
4325 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4326 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4328 6. The freeze_tell option was not being used for messages that were frozen on
4329 arrival, either by an ACL or by local_scan().
4331 7. Added the smtp_incomplete_transaction log selector.
4333 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4334 was accepting AUTH without a new EHLO.
4336 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4337 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4340 10. Logging of TCP/IP connections (when configured) now happens in the main
4341 daemon process instead of the child process, so that the TCP/IP connection
4342 count is more accurate (but it can never be perfect).
4344 11. The use of "drop" in a nested ACL was not being handled correctly in the
4345 outer ACL. Now, if condition failure induced by the nested "drop" causes
4346 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4347 or "require"), the connection is dropped.
4349 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4350 that yield "discard" can now be used with an "accept" or a "discard" verb,
4351 but an error is generated for any others (because I can't see a useful way
4352 to define what should happen).
4354 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4355 are now processed in the same way as lines in the Exim configuration file.
4356 In particular, continuation lines are supported.
4358 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4360 15. Added -ti meaning -t -i.
4362 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4363 domains, and warn by logging if others are found.
4365 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4366 returned from GIFCONF when Exim is trying to find the list of interfaces on
4367 a host. The code in os.c has been modified to copy each ifreq to an aligned
4368 structure in all cases.
4370 Also, in some cases, the returned ifreq's were being copied to a 'struct
4371 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4372 means the last couple of bytes of an IPv6 address could be chopped if the
4373 ifreq contained only a normal sockaddr (14 bytes storage).
4375 18. Named domain lists were not supported in the hosts_treat_as_local option.
4376 An entry such as +xxxx was not recognized, and was treated as a literal
4379 19. Ensure that header lines added by a DATA ACL are included in the reject log
4380 if the ACL subsequently rejects the message.
4382 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4383 MD5 (which is deprecated).
4385 21. When testing a filter file using -bf, Exim was writing a message when it
4386 took the sender from a "From " line in the message, but it was not doing so
4387 when it took $return_path from a Return-Path: header line. It now does.
4389 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4390 with a valid header line field name (a series of printing characters
4391 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4393 23. Changed "disc" in the source to "disk" to conform to the documentation and
4394 the book and for uniformity.
4396 24. Ignore Sendmail's -Ooption=value command line item.
4398 25. When execve() failed while trying to run a command in a pipe transport,
4399 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4400 could be confused with a return value of 69 from the command itself. This
4401 has been changed to 127, the value the shell returns if it is asked to run
4402 a non-existent command. The wording for the related log line suggests a
4403 non-existent command as the problem.
4405 26. If received_header_text expands to an empty string, do not add a Received:
4406 header line to the message. (Well, it adds a token one on the spool, but
4407 marks it "old" so that it doesn't get used or transmitted.)
4409 27. Installed eximstats 1.28 (addition of -nt option).
4411 28. There was no check for failure on the call to getsockname() in the daemon
4412 code. This can fail if there is a shortage of resources on the system, with
4413 ENOMEM, for example. A temporary error is now given on failure.
4415 29. Contrary to the C standard, it seems that in some environments, the
4416 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4417 program. Exim now does this explicitly; it affects the formatting of
4418 timestamps using strftime().
4420 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4421 complaints. I've now initialized all the variables, to avoid this.
4423 32. Header lines added by a system filter were not being "seen" during
4424 transport-time rewrites.
4426 33. The info_callback() function passed to OpenSSL is set up with type void
4427 (*)(SSL *, int, int), as described somewhere. However, when calling the
4428 function (actually a macro) that sets it up, the type void(*)() is
4429 expected. I've put in a cast to prevent warnings from picky compilers.
4431 34. If a DNS black list lookup found a CNAME record, but there were no A
4432 records associated with the domain it pointed at, Exim crashed.
4434 35. If a DNS black list lookup returned more than one A record, Exim ignored
4435 all but the first. It now scans all returned addresses if a particular IP
4436 value is being sought. In this situation, the contents of the
4437 $dnslist_value variable are a list of all the addresses, separated by a
4440 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4441 to accept a host name and all its aliases if the forward lookup for any of
4442 them yielded the IP address of the incoming connection. Now it accepts only
4443 those names whose forward lookup yields the correct IP address. Any other
4444 names are discarded. This closes a loophole whereby a rogue DNS
4445 administrator could create reverse DNS records to break through a
4446 wildcarded host restriction in an ACL.
4448 37. If a user filter or a system filter that ran in a subprocess used any of
4449 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4450 the wrong values were passed to the pipe command ($thisaddress had the
4451 value of $0, $0 had the value of $1, etc). This bug was introduced by
4452 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4454 38. Improved the line breaking for long SMTP error messages from ACLs.
4455 Previously, if there was no break point between 40 and 75 characters, Exim
4456 left the rest of the message alone. Two changes have been made: (a) I've
4457 reduced the minimum length to 35 characters; (b) if it can't find a break
4458 point between 35 and 75 characters, it looks ahead and uses the first one
4459 that it finds. This may give the occasional overlong line, but at least the
4460 remaining text gets split now.
4462 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4463 file descriptors might be low, and that setting 1000 would always raise it.
4464 It turns out that in some environments, the limit is already over 1000 and
4465 that lowering it causes trouble. So now Exim takes care not to decrease it.
4467 40. When delivering a message, the value of $return_path is set to $sender_
4468 address at the start of routing (routers may change the value). By an
4469 oversight, this default was not being set up when an address was tested by
4470 -bt or -bv, which affected the outcome if any router or filter referred to
4473 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4474 log only when "message" or "log_message" are set. However, if one of the
4475 conditions was an address verification, or a call to a nested ACL, the
4476 messages generated by the underlying test were being passed through. This
4477 no longer happens. The underlying message is available in $acl_verify_
4478 message for both "message" and "log_message" expansions, so it can be
4479 passed through if needed.
4481 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4482 new expansion $bh_ to give the encoded byte string without charset
4483 translation. Translation happens only if iconv() is available; HAVE_ICONV
4484 indicates this at build time. HEADERS_CHARSET gives the charset to
4485 translate to; headers_charset can change it in the configuration, and
4486 "headers charset" can change it in an individual filter file.
4488 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4489 that creates RFC 2047 encoded "words" labels them as that charset instead
4490 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4491 expansion operator; (ii) when Exim creates a From: line for a local
4492 message; (iii) when a header line is rewritten to include a "phrase" part.
4494 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4495 buggy, causing it to skip the first lines of messages whose message ID
4496 ended in 'D'. This would not have bitten before Exim release 4.14, because
4497 message IDs were unlikely to end in 'D' before then. The effect was to have
4498 incorrect size information for certain domains.
4500 45. #include "config.h" was missing at the start of the crypt16.c module. This
4501 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4504 46. If there was a timeout during a "random" callout check, Exim treated it as
4505 a failure of the random address, and carried on sending RSET and the real
4506 address. If the delay was just some slowness somewhere, the response to the
4507 original RCPT would be taken as a response to RSET and so on, causing
4508 mayhem of various kinds.
4510 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4511 when I implemented it. It didn't allow for the fact that some option values
4512 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4513 the right test for positive values.
4515 48. Domain names in DNS records are case-independent. Exim always looks them up
4516 in lower case. Some resolvers return domain names in exactly the case they
4517 appear in the zone file, that is, they may contain uppercase letters. Not
4518 all resolvers do this - some return always lower case. Exim was treating a
4519 change of case by a resolver as a change of domain, similar to a widening
4520 of a domain abbreviation. This triggered its re-routing code and so it was
4521 trying to route what was effectively the same domain again. This normally
4522 caused routing to fail (because the router wouldn't handle the domain
4523 twice). Now Exim checks for this case specially, and just changes the
4524 casing of the domain that it ultimately uses when it transmits the message
4527 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4530 50. If a filter generated a file delivery with a non-absolute name (possible if
4531 no home directory exists for the router), the forbid_file option was not
4534 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4535 the existing equality matching.
4537 52. Exim was using ints instead of ino_t variables in some places where it was
4538 dealing with inode numbers.
4540 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4541 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4542 called TMPDIR, and if it finds it is different, it changes its value.
4544 54. The smtp_printf() function is now made available to local_scan() so
4545 additional output lines can be written before returning. There is also an
4546 smtp_fflush() function to enable the detection of a dropped connection.
4547 The variables smtp_input and smtp_batched_input are exported to
4550 55. Changed the default runtime configuration: the message "Unknown user"
4551 has been removed from the ACL, and instead placed on the localuser router,
4552 using the cannot_route_message feature. This means that any verification
4553 failures that generate their own messages won't get overridden. Similarly,
4554 the "Unrouteable address" message that was in the ACL for unverifiable
4555 relay addresses has also been removed.
4557 56. Added hosts_avoid_esmtp to the smtp transport.
4559 57. The exicyclog script was not checking for the esoteric option
4560 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4561 will work only if exicyclog is run under the appropriate euid.
4563 58. Following a discussion on the list, the rules by which Exim recognises line
4564 endings on incoming messages have been changed. The -dropcr and drop_cr
4565 options are now no-ops, retained only for backwards compatibility. The
4566 following line terminators are recognized: LF CRLF CR. However, special
4567 processing applies to CR:
4569 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4570 nor a local message in the state where . is a terminator.
4572 (ii) If a bare CR is encountered in a header line, an extra space is added
4573 after the line terminator so as not to end the header. The reasoning
4574 behind this is that bare CRs in header lines are most likely either
4575 to be mistakes, or people trying to play silly games.
4577 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4578 was being incorrectly given as 18 bytes larger than it should have been.
4579 This is a VOB (very old bug).
4581 60. This may never have affected anything current, but just in case it has:
4582 When the local host is found other than at the start of a list of hosts,
4583 the local host, those with the same MX, and any that follow, are discarded.
4584 When the list in question was part of a longer list of hosts, the following
4585 hosts (not currently being processed) were also being discarded. This no
4586 longer happens. I'm not sure if this situation could ever has previously
4589 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4592 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4593 header. This is something that is recommended in a new Internet Draft, and
4594 is something that is documented as being done by Sendmail. There are two
4595 possible values. For messages generated by the autoreply transport, Exim
4598 Auto-Submitted: auto-replied
4600 whereas for all other generated messages (e.g. bounces) it adds
4602 Auto-Submitted: auto-generated
4604 63. The "personal" condition in filters now includes a test for the
4605 Auto-Submitted: header. If it contains the string "auto-" the message it
4606 not considered personal.
4608 64. Added rcpt_include_affixes as a generic transport option.
4610 65. Added queue_only_override (default true).
4612 66. Added the syslog_duplication option.
4614 67. If what should have been the first header line of a message consisted of
4615 a space followed by a colon, Exim was mis-interpreting it as a header line.
4616 It isn't of course - it is syntactically invalid and should therefore be
4617 treated as the start of the message body. The misbehaviour could have
4618 caused a number of strange effects, including loss of data in subsequent
4619 header lines, and spool format errors.
4621 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4622 client host had authenticated. This control can now be exercised by an ACL
4623 for more flexibility.
4625 69. By default, callouts do not happen when testing with -bh. There is now a
4626 variant, -bhc, which does actually run the callout code, including
4627 consulting and updating the callout cache.
4629 70. Added support for saslauthd authentication, courtesy of Alexander
4632 71. If statvfs() failed on the spool or log directories while checking their
4633 size for availability, Exim confusingly gave the error "space shortage".
4634 Furthermore, in debugging mode it crashed with a floating point exception.
4635 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4636 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4637 serious problem, Exim now writes to the main and panic logs when this
4638 happens, with details of the failure. It then refuses to accept the
4639 incoming message, giving the message "spool directory problem" or "log
4640 directory problem" with a 421 code for SMTP messages.
4642 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4643 0, 1, and 2 exist, because some OS complain for execs without them (see
4644 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4645 descriptors. However, the code omitted to check that the open succeeded,
4646 causing mysterious errors if for some reason the permissions on /dev/null
4647 got screwed. Now Exim writes a message to the main and panic logs, and
4648 bombs out if it can't open /dev/null.
4650 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4651 interact so that it is all more flexible. It is supposed to remain
4652 backwards compatible. Also added extra_local_interfaces.
4654 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4655 to bomb out with an assertion failure - to the client this appears as a
4656 connection drop. This problem occurs in the part of the code that was taken
4657 from the Samba project. Fortunately, the assertion is in a very simple
4658 function, so I have fixed this by reproducing the function inline in the
4659 one place where it is called, and arranging for authentication to fail
4660 instead of killing the process with assert().
4662 75. The SPA client code was not working when the server requested OEM rather
4663 than Unicode encoding.
4665 76. Added code to make require_files with a specific uid setting more usable in
4666 the case where statting the file as root fails - usually a non-root-mounted
4667 NFS file system. When this happens and the failure is EACCES, Exim now
4668 forks a subprocess and does the per-uid checking as the relevant uid.
4670 77. Added process_log_path.
4672 78. If log_file_path was not explicitly set, a setting of check_log_space or
4673 check_log_inodes was ignored.
4675 79. If a space check for the spool or log partitions fails, the incident is now
4676 logged. Of course, in the latter case the data may get lost...
4678 80. Added the %p formatting code to string_format() so that it can be used to
4679 print addresses in debug_print(). Adjusted all the address printing in the
4680 debugging in store.c to use %p rather than %d.
4682 81. There was a concern that a line of code in smtp_in.c could overflow a
4683 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4684 initially expressed, the concern was not well-founded, because trailing
4685 spaces are removed early. However, if the trailing spaces were followed by
4686 a NULL, they did not get removed, so the overflow was possible. Two fixes
4689 (a) I re-wrote the offending code in a cleaner fashion.
4690 (b) If an incoming SMTP command contains a NULL character, it is rejected
4693 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4694 runs initgroups(), so that if the Exim user is in any additional groups,
4695 they will be used during message reception.
4701 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob