1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.463 2007/01/25 15:51:28 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.
67 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
68 fixed by 4.65/MH/01 (is this a record?) are fixed:
70 (i) An empty string was always treated as zero by the numeric comparison
71 operators. This behaviour has been restored.
73 (ii) It is documented that the numeric comparison operators always treat
74 their arguments as decimal numbers. This was broken in that numbers
75 starting with 0 were being interpreted as octal.
77 While fixing these problems I realized that there was another issue that
78 hadn't been noticed. Values of message_size_limit (both the global option
79 and the transport option) were treated as octal if they started with 0.
80 The documentation was vague. These values are now always treated as
81 decimal, and I will make that clear in the documentation.
87 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
88 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
91 MH/01 Don't check that the operands of numeric comparison operators are
92 integers when their expansion is in "skipping" mode (fixes bug
93 introduced by 4.64-PH/07).
95 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
96 child addresses, Exim now panics and dies. Previously, because the count
97 is held in a short int, deliveries were likely to be lost. As such a
98 large number of recipients for a single message is ridiculous
99 (performance will be very, very poor), I have chosen to impose a limit
100 rather than extend the field.
106 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
107 leftover -K file (the existence of which was triggered by #402).
108 While we were at it, introduced process PID as part of the -K
109 filename. This should rule out race conditions when creating
112 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
113 processing considerably. Previous code took too long for large mails,
114 triggering a timeout which in turn triggers #401.
116 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
117 in the DK code in transports.c. sendfile() is not really portable,
118 hence the _LINUX specificness.
120 TF/01 In the add_headers option to the mail command in an Exim filter,
121 there was a bug that Exim would claim a syntax error in any
122 header after the first one which had an odd number of characters
125 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
126 callout verification, Exim cached a "reject" for the entire domain. This
127 is correct for most verifications, but it is not correct for a recipient
128 verification with use_sender or use_postmaster set, because in that case
129 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
130 case of MAIL FROM:<> rejection from other early rejections (e.g.
131 rejection of HELO). When verifying a recipient using a non-null MAIL
132 address, the cache is ignored if it shows MAIL FROM:<> rejection.
133 Whatever the result of the callout, the value of the domain cache is
134 left unchanged (for any other kind of callout, getting as far as trying
135 RCPT means that the domain itself is ok).
137 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
140 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
141 manifest itself as EPIPE rather than ECONNECT. When tidying away a
142 session, the daemon ignores ECONNECT errors and logs others; it now
143 ignores EPIPE as well.
145 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
146 (quoted-printable decoding).
148 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
149 later the small subsequent patch to fix an introduced bug.
151 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
153 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
155 PH/08 An error is now given if message_size_limit is specified negative.
157 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
158 to be given (somewhat) arbitrary names.
160 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
163 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
164 miscellaneous code fixes
166 PH/10 Added the log_reject_target ACL modifier to specify where to log
169 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
170 hostname. This is wrong, because it relates to the incoming message (and
171 probably the interface on which it is arriving) and not to the outgoing
172 callout (which could be using a different interface). This has been
173 changed to use the value of the helo_data option from the smtp transport
174 instead - this is what is used when a message is actually being sent. If
175 there is no remote transport (possible with a router that sets up host
176 addresses), $smtp_active_hostname is used.
178 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
179 tweaks were necessary in order to get it to work (see also 21 below):
180 (a) The code assumed that strncpy() returns a negative number on buffer
181 overflow, which isn't the case. Replaced with Exim's string_format()
183 (b) There were several signed/unsigned issues. I just did the minimum
184 hacking in of casts. There is scope for a larger refactoring.
185 (c) The code used strcasecmp() which is not a standard C function.
186 Replaced with Exim's strcmpic() function.
187 (d) The code set only $1; it now sets $auth1 as well.
188 (e) A simple test gave the error "authentication client didn't specify
189 service in request". It would seem that Dovecot has changed its
190 interface. Fortunately there's a specification; I followed it and
191 changed what the client sends and it appears to be working now.
193 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
196 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
197 address A is aliased to B and C, where B exists and C does not. Without
198 -v the output is "A verified" because verification stops after a
199 successful redirection if more than one address is generated. However,
200 with -v the child addresses are also verified. Exim was outputting "A
201 failed to verify" and then showing the successful verification for C,
202 with its parentage. It now outputs "B failed to verify", showing B's
203 parentage before showing the successful verification of C.
205 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
206 look up a TXT record in a specific list after matching in a combined
209 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
210 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
211 they consult the DNS. I had assumed they would set it the way they
212 wanted; and indeed my experiments on Linux seem to show that in some
213 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
214 To be on the safe side, however, I have now made the interface to
215 host_find_byname() similar to host_find_bydns(), with an argument
216 containing the DNS resolver options. The host_find_byname() function now
217 sets these options at its start, just as host_find_bydns() does. The smtp
218 transport options dns_qualify_single and dns_search_parents are passed to
219 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
220 of host_find_byname() use the default settings of RES_DEFNAMES
221 (qualify_single) but not RES_DNSRCH (search_parents).
223 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
224 spool_read_header() do less string testing, by means of a preliminary
225 switch on the second character of optional "-foo" lines. (This is
226 overdue, caused by the large number of possibilities that now exist.
227 Originally there were few.) While I was there, I also converted the
228 str(n)cmp tests so they don't re-test the leading "-" and the first
229 character, in the hope this might squeeze out yet more improvement.
231 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
232 flag allowing group syntax was set by the header_syntax check but not
233 turned off, possible causing trouble later; (2) The flag was not being
234 set at all for the header_verify test, causing "group"-style headers to
235 be rejected. I have now set it in this case, and also caused header_
236 verify to ignore an empty address taken from a group. While doing this, I
237 came across some other cases where the code for allowing group syntax
238 while scanning a header line wasn't quite right (mostly, not resetting
239 the flag correctly in the right place). These bugs could have caused
240 trouble for malformed header lines. I hope it is now all correct.
242 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
243 with the "reply" argument non-NULL. The code, however (which originally
244 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
245 but it didn't always do it. This confused somebody who was copying the
246 code for some other use. I have removed all the tests.
248 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
249 feature that was used to support insecure browsers during the U.S. crypto
250 embargo. It requires special client support, and Exim is probably the
251 only MTA that supported it -- and would never use it because real RSA is
252 always available. This code has been removed, because it had the bad
253 effect of slowing Exim down by computing (never used) parameters for the
254 RSA_EXPORT functionality.
256 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
257 authenticator to fail if there's a tab character in the incoming data
258 (there should never be unless someone is messing about, as it's supposed
259 to be base64-encoded). Also added, on Timo's advice, the "secured" option
260 if the connection is using TLS or if the remote IP is the same as the
261 local IP, and the "valid-client-cert option" if a client certificate has
264 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
265 authenticators. This can be used for authorization after authentication
266 succeeds. (In the case of plaintext, it servers for both authentication
269 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
270 if any retry times were supplied.
272 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
273 connection, where there is no client IP address to check. In this
274 situation, the verify now always succeeds.
276 PH/25 Applied John Jetmore's -Mset patch.
278 PH/26 Added -bem to be like -Mset, but loading a message from a file.
280 PH/27 In a string expansion for a processed (not raw) header when multiple
281 headers of the same name were present, leading whitespace was being
282 removed from all of them, but trailing whitespace was being removed only
283 from the last one. Now trailing whitespace is removed from each header
284 before concatenation. Completely empty headers in a concatenation (as
287 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
288 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
290 PH/29 [Removed. This was a change that I later backed out, and forgot to
291 correct the ChangeLog entry (that I had efficiently created) before
292 committing the later change.]
294 PH/30 Exim was sometimes attempting to deliver messages that had suffered
295 address errors (4xx response to RCPT) over the same connection as other
296 messages routed to the same hosts. Such deliveries are always "forced",
297 so retry times are not inspected. This resulted in far too many retries
298 for the affected addresses. The effect occurred only when there were more
299 hosts than the hosts_max_try setting in the smtp transport when it had
300 the 4xx errors. Those hosts that it had tried were not added to the list
301 of hosts for which the message was waiting, so if all were tried, there
302 was no problem. Two fixes have been applied:
304 (i) If there are any address or message errors in an SMTP delivery, none
305 of the hosts (tried or untried) are now added to the list of hosts
306 for which the message is waiting, so the message should not be a
307 candidate for sending over the same connection that was used for a
308 successful delivery of some other message. This seems entirely
309 reasonable: after all the message is NOT "waiting for some host".
310 This is so "obvious" that I'm not sure why it wasn't done
311 previously. Hope I haven't missed anything, but it can't do any
312 harm, as the worst effect is to miss an optimization.
314 (ii) If, despite (i), such a delivery is accidentally attempted, the
315 routing retry time is respected, so at least it doesn't keep
316 hammering the server.
318 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
319 in ${readsocket because some servers need this prod.
321 PH/32 Added some extra debug output when updating a wait-xxx database.
323 PH/33 The hint "could be header name not terminated by colon", which has been
324 given for certain expansion errors for a long time, was not being given
325 for the ${if def:h_colon_omitted{... case.
327 PH/34 The spec says: "With one important exception, whenever a domain list is
328 being scanned, $domain contains the subject domain." There was at least
329 one case where this was not true.
331 PH/35 The error "getsockname() failed: connection reset by peer" was being
332 written to the panic log as well as the main log, but it isn't really
333 panic-worthy as it just means the connection died rather early on. I have
334 removed the panic log writing for the ECONNRESET error when getsockname()
337 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
338 runs only) independently of the message's sender address. This meant
339 that, if the 4xx error was in fact related to the sender, a different
340 message to the same recipient with a different sender could confuse
341 things. In particualar, this can happen when sending to a greylisting
342 server, but other circumstances could also provoke similar problems.
343 I have changed the default so that the retry time for these errors is now
344 based a combination of the sender and recipient addresses. This change
345 can be overridden by setting address_retry_include_sender=false in the
348 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
349 remote server are returned as part of bounce messages. This was not
350 happening for LMTP over a pipe (the lmtp transport), but now it is the
351 same for both kinds of LMTP.
353 PH/38 Despite being documented as not happening, Exim was rewriting addresses
354 in header lines that were in fact CNAMEs. This is no longer the case.
356 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
357 and queue runs started by the daemon processed all messages. This has
358 been fixed so that -R and -S can now usefully be given with -q<time>.
360 PH/40 Import PCRE release 6.7 (fixes some bugs).
362 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
364 PH/42 Give an error if -q is specified more than once.
366 PH/43 Renamed the variables $interface_address and $interface_port as
367 $received_ip_address and $received_port, to make it clear that these
368 values apply to message reception, and not to the outgoing interface when
369 a message is delivered. (The old names remain recognized, of course.)
371 PH/44 There was no timeout on the connect() call when using a Unix domain
372 socket in the ${readsocket expansion. There now is.
374 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
375 be meaningful with "accept".
377 SC/01 Eximstats V1.43
378 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
380 SC/02 Eximstats V1.44
381 Use a glob alias rather than an array ref in the generated
382 parser. This improves both readability and performance.
384 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
385 Collect SpamAssassin and rejection statistics.
386 Don't display local sender or destination tables unless
387 there is data to show.
388 Added average volumes into the top table text output.
390 SC/04 Eximstats V1.46
391 Collect data on the number of addresses (recipients)
392 as well as the number of messages.
394 SC/05 Eximstats V1.47
395 Added 'Message too big' to the list of mail rejection
396 reasons (thanks to Marco Gaiarin).
398 SC/06 Eximstats V1.48
399 Mainlog lines which have GMT offsets and are too short to
400 have a flag are now skipped.
402 SC/07 Eximstats V1.49 (Alain Williams)
403 Added the -emptyok flag.
405 SC/08 Eximstats V1.50
406 Fixes for obtaining the IP address from reject messages.
408 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
409 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
410 whitesspace changes from 4.64-PH/27
412 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
415 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
416 are found, allow negative numbers in numeric criteria)
418 JJ/06 exipick.20061117.2, added new $message_body_missing variable
420 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
421 to match changes made in 4.64-PH/43
423 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
425 PH/47 Put in an explicit test for a DNS lookup of an address record where the
426 "domain" is actually an IP address, and force a failure. This locks out
427 those revolvers/nameservers that support "A-for-A" lookups, in
428 contravention of the specifications.
430 PH/48 When a host name was looked up from an IP address, and the subsequent
431 forward lookup of the name timed out, the host name was left in
432 $sender_host_name, contrary to the specification.
434 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
435 restricted to single-key lookups, Exim was not diagnosing an error if
436 * or *@ was used with a query-style lookup.
438 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
440 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
441 long ago, but noone interested enough thought of it. Let's just say that
442 the "1.1" means that there are some new functions that weren't there at
443 some point in the past.
445 PH/51 Error processing for expansion failure of helo_data from an smtp
446 transport during callout processing was broken.
448 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
449 tested/used via the -bh/-bhc/-bs options.
451 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
452 bug, fixed in subsequent PCRE releases).
454 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
455 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
457 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
463 SC/01 Use a glob alias rather than an array ref in eximstats generated
464 parser. This improves both readability and performance.
466 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
467 Don't display local sender or destination tables in eximstats unless
468 there is data to show.
469 Added average volumes into the eximstats top table text output.
471 SC/03 Collect data on the number of addresses (recipients) as well
472 as the number of messages in eximstats.
474 TF/01 Correct an error in the documentation for the redirect router. Exim
475 does (usually) call initgroups() when daemonizing.
477 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
478 with consistent privilege compared to when running as a daemon.
480 TF/03 Note in the spec that $authenticated_id is not set for local
481 submissions from trusted users.
483 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
484 Thanks to Dean Brooks <dean@iglou.com> for the patch.
486 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
487 by adding some example configuration directives to the default
488 configuration file. A little bit of work is required to uncomment the
489 directives and define how usernames and passwords are checked, but
490 there is now a framework to start from.
492 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
493 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
494 without this. I don't know how relevant this is to other LDAP libraries.
496 PH/02 Add the verb name to the "unknown ACL verb" error.
498 PH/03 Magnus Holmgren's patch for filter_prepend_home.
500 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
502 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
503 directory not expanded when it should be if an expanded home directory
504 was set for the address (which is overridden by the transport).
506 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
509 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
510 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
511 because it is too late at that time, and has no effect.
513 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
514 security issue with \' (bugzilla #107). I could not use the
515 PQescapeStringConn() function, because it needs a PGconn value as one of
518 PH/08 When testing addresses using -bt, indicate those final addresses that
519 are duplicates that would not cause an additional delivery. At least one
520 person was confused, thinking that -bt output corresponded to deliveries.
521 (Suppressing duplicates isn't a good idea as you lose the information
522 about possibly different redirections that led to the duplicates.)
524 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
525 systems where poll() doesn't work, in particular OS X.
527 PH/10 Added more information to debugging output for retry time not reached.
529 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
530 operations in malware.c.
532 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
535 PH/13 If write_rejectlog was set false when logging was sent to syslog with
536 syslog_duplication set false, log lines that would normally be written
537 both the the main log and to the reject log were not written to syslog at
540 PH/14 In the default configuration, change the use of "message" in ACL warn
541 statements to "add_header".
543 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
544 not followed by a command (e.g. "seen endif").
546 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
547 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
550 PH/17 Added extra conditions to the default value of delay_warning_condition
554 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
555 { match{$h_precedence:}{(?i)bulk|list|junk} } \
556 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
559 The Auto-Submitted: and various List- headers are standardised, whereas I
560 don't think Precedence: ever was.
562 PH/18 Refactored debugging code in route_finduser() to show more information,
563 in particular, the error code if getpwnam() issues one.
565 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
566 This is apparently needed in addition to the PH/07 change above to avoid
567 any possible encoding problems.
569 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
570 but not after initializing Perl.
572 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
573 output them only if debugging. By default they are written stderr,
574 apparently, which is not desirable.
576 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
579 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
582 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
584 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
585 authenticated or an ident call has been made. Suppress the default
586 values for $authenticated_id and $authenticated_sender (but permit -oMai
587 and -oMas) when testing with -bh.
589 PH/24 Re-jigged the order of the tests in the default configuration so that the
590 tests for valid domains and recipients precede the DNS black list and CSA
591 tests, on the grounds that those ones are more expensive.
593 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
594 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
595 This bug exists in every version of Exim that I still have, right back to
598 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
599 However, an attempt to turn on case-sensitivity in a regex key by
600 including (?-i) didn't work because the subject string was already
601 lowercased, and the effects were non-intuitive. It turns out that a
602 one-line patch can be used to allow (?-i) to work as expected.
608 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
609 other effects) broke the use of negated acl sub-conditions.
611 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
614 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
615 "Deny" causes Exim to reject the incoming connection with a 554 error.
616 Unfortunately, if there is a major crisis, such as a disk failure,
617 tcp-wrappers gives "deny", whereas what one would like would be some
618 kind of temporary error. A kludge has been added to help with this.
619 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
620 554 error is used if errno is still zero or contains ENOENT (which occurs
621 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
624 PH/03 Add -lutil to the default FreeBSD LIBS setting.
626 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
627 errors. Otherwise a message that provokes a temporary error (when other
628 messages do not) can cause a whole host to time out.
630 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
631 addresses were routed directly to files or pipes from a redirect router.
632 File deliveries just didn't batch; pipe deliveries might have suffered
635 PH/06 A failure to get a lock for a hints database would erroneously always say
636 "Failed to get write lock", even when it was really a read lock.
638 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
639 of 0600. This has been changed to use the value of the lockfile_mode
640 option (which defaults to 0600).
642 PH/08 Applied small patch from the Sieve maintainer.
644 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
645 folder from quota calculations, a direct delivery into this folder messed
646 up the contents of the maildirsize file. This was because the regex was
647 used only to exclude .Trash (or whatever) when the size of the mailbox
648 was calculated. There was no check that a delivery was happening into an
649 excluded directory. This bug has been fixed by ignoring all quota
650 processing for deliveries into excluded directories.
652 PH/10 Added the maildirfolder_create_regex option to appendfile.
658 PH/01 The code for finding all the local interface addresses on a FreeBSD
659 system running IPv6 was broken. This may well have applied to all BSD
660 systems, as well as to others that have similar system calls. The broken
661 code found IPv4 interfaces correctly, but gave incorrect values for the
662 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
663 that it would not match correctly against @[] and not recognize the IPv6
666 PH/02 The ipliteral router was not recognizing addresses of the form user@
667 [ipv6:....] because it didn't know about the "ipv6:" prefix.
669 PH/03 Added disable_ipv6.
671 PH/04 Changed $reply_address to use the raw form of the headers instead of the
672 decoded form, because it is most often used to construct To: headers
673 lines in autoreplies, and the decoded form may well be syntactically
674 invalid. However, $reply_address has leading white space removed, and all
675 newlines turned into spaces so that the autoreply transport does not
678 PH/05 If group was specified without a user on a router, and no group or user
679 was specified on a transport, the group from the router was ignored.
681 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
682 for visible compile-time settings that can be used to change these
683 numbers, for those that want even more. Backwards compatibility with old
684 spool files has been maintained. However, going back to a previous Exim
685 release will lost any variables that are in spool files.
687 PH/07 Two small changes when running in the test harness: increase delay when
688 passing a TCP/IP connection to a new process, in case the original
689 process has to generate a bounce, and remove special handling of
690 127.0.0.2 (sic), which is no longer necessary.
692 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
693 be the same on different OS.
695 PH/09 Moved a debug statement in filter processing to avoid a race problem when
698 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
699 whether --show-vars was specified or not
701 JJ/02 exipick: Added support for new ACL variable spool format introduced
704 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
705 syntactically invalid From: or Reply-to: line, and a filter used this to
706 generate an autoreply, and therefore failed to obtain an address for the
707 autoreply, Exim could try to deliver to a non-existent relative file
708 name, causing unrelated and misleading errors. What now happens is that
709 it logs this as a hard delivery error, but does not attempt to create a
712 PH/11 The exinext utility has a -C option for testing purposes, but although
713 the given file was scanned by exinext itself; it wasn't being passed on
714 when Exim was called.
716 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
717 an end-of-file indication when reading a command response.
719 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
720 compiled. In many other places in Exim, IPv6 addresses are always
721 recognized, so I have changed this. It also means that IPv4 domain
722 literals of the form [IPV4:n.n.n.n] are now always recognized.
724 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
725 used if the router is not running as root, for example, when verifying at
726 ACL time, or when using -bh. The debugging output from this situation was
727 non-existent - all you got was a failure to exec. I have made two
730 (a) Failures to set uid/gid, the current directory, or a process leader
731 in a subprocess such as that created by queryprogram now generate
732 suitable debugging ouput when -d is set.
734 (b) The queryprogram router detects when it is not running as root,
735 outputs suitable debugging information if -d is set, and then runs
736 the subprocess without attempting to change uid/gid.
738 PH/15 Minor change to Makefile for building test_host (undocumented testing
741 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
742 additional section of a DNS packet that returns MX or SRV records.
743 Instead, it always explicitly searches for A/AAAA records. This avoids
744 major problems that occur when a DNS server includes only records of one
745 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
746 fixed another bug: if SRV records were looked up and the corresponding
747 address records were *not* found in the additional section, the port
748 values from the SRV records were lost.
750 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
751 using the correct key (the original address) when searching the retry
752 rules in order to find which one to use for generating the retry hint.
754 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
755 adding the default one. Similarly, if it contains a Reply-To: header, the
756 errors_reply_to option, if set, is not used.
758 PH/19 When calculating a retry time, Exim used to measure the "time since
759 failure" by looking at the "first failed" field in the retry record. Now
760 it does not use this if it is later than than the arrival time of the
761 message. Instead it uses the arrival time. This makes for better
762 behaviour in cases where some deliveries succeed, thus re-setting the
763 "first failed" field. An example is a quota failure for a huge message
764 when small messages continue to be delivered. Without this change, the
765 "time since failure" will always be short, possible causing more frequent
766 delivery attempts for the huge message than are intended.
767 [Note: This change was subsequently modified - see PH/04 for 4.62.]
769 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
770 $1, $2, $3) because the numerical variables can be reset during some
771 expansion items (e.g. "match"), thereby losing the authentication data.
773 PH/21 Make -bV show the size of off_t variables so that the test suite can
774 decide whether to run tests for quotas > 2G.
776 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
777 mailbox_size, and mailbox_filecount in the appendfile transport. If a
778 filecount value is greater than 2G or if a quota value is greater than 2G
779 on a system where the size of off_t is not greater than 4, a panic error
782 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
783 never match. The debug and -bh output now contains an explicit error
784 message indicating a malformed IPv4 address or mask.
786 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
787 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
790 PH/25 Do not write to syslog when running in the test harness. The only
791 occasion when this arises is a failure to open the main or panic logs
792 (for which there is an explicit test).
794 PH/26 Added the /no_tell option to "control=freeze".
796 PH/27 If a host name lookup failed very early in a connection, for example, if
797 the IP address matched host_lookup and the reverse lookup yielded a name
798 that did not have a forward lookup, an error message of the form "no IP
799 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
800 could be logged. Now it outputs the IP address instead of "NULL".
802 PH/28 An enabling patch from MH: add new function child_open_exim2() which
803 allows the sender and the authenticated sender to be set when
804 submitting a message from within Exim. Since child_open_exim() is
805 documented for local_scan(), the new function should be too.
807 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
808 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
809 results in an empty string is now treated as unset.
811 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
813 PH/31 Added sender_verify_fail logging option.
815 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
816 needed by RFC 822 but not by RFC 2822 was commented out. I have now
817 tidied the source and removed it altogether.
819 PH/33 When a queue run was abandoned because the load average was too high, a
820 log line was always written; now it is written only if the queue_run log
821 selector is set. In addition, the log line for abandonment now contains
822 information about the queue run such as the pid. This is always present
823 in "start" and "stop" lines but was omitted from the "abandon" line.
825 PH/34 Omit spaces between a header name and the colon in the error message that
826 is given when verify = headers_syntax fails (if there are lots of them,
827 the message gets confusing).
829 PH/35 Change the default for dns_check_names_pattern to allow slashes within
830 names, as there are now some PTR records that contain slashes. This check
831 is only to protect against broken name servers that fall over on strange
832 characters, so the fact that it applies to all lookups doesn't matter.
834 PH/36 Now that the new test suite is complete, we can remove some of the
835 special code in Exim that was needed for the old test suite. For example,
836 sorting DNS records because real resolvers return them in an arbitrary
837 order. The new test suite's fake resolver always returns records in the
840 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
841 bounces) except when queue_only is set, to avoid logging races between
842 the different processes.
844 PH/38 Panic-die if .include specifies a non-absolute path.
846 PH/39 A tweak to the "H" retry rule from its user.
848 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
849 a label. They prevented compilation on older perls.
851 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
852 a warning to be raised on newish perls.
854 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
855 on queue. Changes to match documented behaviour of showing count of
856 messages matching specified criteria.
858 PH/40 Changed the default ident timeout from 30s to 5s.
860 PH/41 Added support for the use of login_cap features, on those BSD systems
861 that have them, for controlling the resources used by pipe deliveries.
863 PH/42 The content-scanning code uses fopen() to create files in which to put
864 message data. Previously it was not paying any attention to the mode of
865 the files. Exim runs with umask(0) because the rest of the code creates
866 files with open(), and sets the required mode explicitly. Thus, these
867 files were ending up world-writeable. This was not a big issue, because,
868 being within the spool directory, they were not world-accessible. I have
869 created a function called modefopen, which takes an additional mode
870 argument. It sets umask(777), creates the file, chmods it to the required
871 mode, then resets the umask. All the relevant calls to fopen() in the
872 content scanning code have been changed to use this function.
874 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
875 to 24 hours. This avoids potential overflow problems when processing G
876 and H retry rules. I suspect nobody ever tinkers with this value.
878 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
880 PH/45 When the plaintext authenticator is running as a client, the server's
881 challenges are checked to ensure they are valid base64 strings. By
882 default, the authentication attempt is cancelled if an invalid string is
883 received. Setting client_ignore_invalid_base64 true ignores these errors.
884 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
885 they are received. Thus, the responses can be made to depend on the
886 challenges. If an invalid string is ignored, an empty string is placed in
889 PH/46 Messages that are created by the autoreply transport now contains a
890 References: header, in accordance with RFCs 2822 and 3834.
892 PH/47 Added authenticated_sender_force to the smtp transport.
894 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
896 PH/49 Installed latest patch from the Sieve maintainer.
898 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
899 was also set, the appendfile transport was unnecessarily scanning a
900 directory of message files (e.g. for maildir delivery) to find the count
901 of files (along with the size), even though it did not need this
902 information. It now does the scan only if it needs to find either the
903 size of the count of files.
905 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
907 PH/52 Two bugs concerned with error handling when the smtp transport is
910 (i) Exim was not creating retry information for temporary errors given
911 for individual recipients after the DATA command when the smtp transport
912 was used in LMTP mode. This meant that they could be retried too
913 frequently, and not timed out correctly.
915 (ii) Exim was setting the flag that allows error details to be returned
916 for LMTP errors on RCPT commands, but not for LMTP errors for individual
917 recipients that were returned after the DATA command.
919 PH/53 This is related to PH/52, but is more general: for any failing address,
920 when detailed error information was permitted to be returned to the
921 sender, but the error was temporary, then after the final timeout, only
922 "retry timeout exceeded" was returned. Now it returns the full error as
923 well as "retry timeout exceeded".
925 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
926 do this, and (what is worse) MTAs that accept it.
928 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
929 will now be deprecated.
931 PH/56 New os.c-cygwin from the Cygwin maintainer.
933 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
934 formats (previously only available in exim formats via -bpr, -bpru,
935 and -bpra. Now also available in native and exiqgrep formats)
937 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
938 with very large, slow to parse queues
940 JJ/08 exipick: added ! as generic prefix to negate any criteria format
942 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
944 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
945 responses to authentication challenges, though it was showing the
946 challenges; (ii) I've removed the CR characters from the debug output for
949 PH/58 Allow for the insertion of a newline as well as a space when a string
950 is turned into more than one encoded-word during RFC 2047 encoding. The
951 Sieve code now uses this.
953 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
954 data_4xx, lost_connection, tls_required.
956 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
957 message was being sent as an SMTP response.
959 PH/61 Add -l and -k options to exicyclog.
961 PH/62 When verifying, if an address was redirected to one new address, so that
962 verification continued, and the new address failed or deferred after
963 having set something in $address_data, the value of $address_data was not
964 passed back to the ACL. This was different to the case when no
965 redirection occurred. The value is now passed back in both cases.
967 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
968 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
969 use login_cap.h, so on its own it isn't the distinguishing feature. The
970 new name refers directly to the setclassresources() function.
972 PH/65 Added configuration files for NetBSD3.
974 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
976 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
977 is preferred over IPv4.
979 PH/68 The bounce_return_message and bounce_return_body options were not being
980 honoured for bounces generated during the reception of non-SMTP messages.
981 In particular, this applied to messages rejected by the ACL. This bug has
982 been fixed. However, if bounce_return_message is true and bounce_return_
983 body is false, the headers that are returned for a non-SMTP message
984 include only those that have been read before the error was detected.
985 (In the case of an ACL rejection, they have all been read.)
987 PH/69 The HTML version of the specification is now built in a directory called
988 spec_html instead of spec.html, because the latter looks like a path with
989 a MIME-type, and this confuses some software.
991 PH/70 Catch two compiler warnings in sieve.c.
993 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
994 function verify_get_ident() calls ip_connect() to connect a socket, but
995 if the "connect()" function timed out, ip_connect() used to close the
996 socket. However, verify_get_ident() also closes the socket later, and in
997 between Exim writes to the log, which may get opened at this point. When
998 the socket was closed in ip_connect(), the log could get the same file
999 descriptor number as the socket. This naturally causes chaos. The fix is
1000 not to close the socket in ip_connect(); the socket should be closed by
1001 the function that creates it. There was only one place in the code where
1002 this was missing, in the iplookup router, which I don't think anybody now
1003 uses, but I've fixed it anyway.
1005 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1006 well as to direct DNS lookups. Otherwise the handling of names in host
1007 lists is inconsistent and therefore confusing.
1013 PH/01 Two changes to the default runtime configuration:
1015 (1) Move the checks for relay_from_hosts and authenticated clients from
1016 after to before the (commented out) DNS black list checks.
1018 (2) Add control=submission to the relay_from_hosts and authenticated
1019 clients checks, on the grounds that messages accepted by these
1020 statements are most likely to be submissions.
1022 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1024 (1) Generate an error if the third argument for the ${prvs expansion is
1027 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1030 (3) Reset the variables that are obtained from the first argument of
1031 ${prvscheck and used in the second argument before leaving the code,
1032 because their memory is reclaimed, so using them afterwards may do
1035 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1036 one (it's much easier than Tom thought :-).
1038 (5) Because of (4), we can now allow for the use of $prvscheck_result
1039 inside the third argument.
1041 PH/03 For some reason, the default setting of PATH when running a command from
1042 a pipe transport was just "/usr/bin". I have changed it to
1045 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1046 anything to be listed in the output from -bV.
1048 PH/05 When a filter generated an autoreply, the entire To: header line was
1049 quoted in the delivery log line, like this:
1051 => >A.N.Other <ano@some.domain> <original@ddress> ...
1053 This has been changed so that it extracts the operative address. There
1054 may be more than one such address. If so, they are comma-separated, like
1057 => >ano@some.domain,ona@other.domain <original@ddress> ...
1059 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1060 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1061 not being looked up in the rDNS to get a host name, Exim was showing the
1062 IP address twice in Received: lines, even though the IP addresses were
1063 identical. For example:
1065 Received: from [1.2.3.4] (helo=[1.2.3.4])
1067 However, if the real host name was known, it was omitting the HELO data
1068 if it matched the actual IP address. This has been tidied up so that it
1069 doesn't show the same IP address twice.
1071 PH/07 When both +timestamp and +memory debugging was on, the value given by
1072 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1073 called by the debug printing, thereby overwriting the timestamp buffer.
1074 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1076 PH/08 When the original message was included in an autoreply transport, it
1077 always said "this is a copy of the message, including all the headers",
1078 even if body_only or headers_only was set. It now gives an appropriate
1081 PH/09 Applied a patch from the Sieve maintainer which:
1083 o fixes some comments
1084 o adds the (disabled) notify extension core
1085 o adds some debug output for the result of if/elsif tests
1086 o points to the current vacation draft in the documentation
1087 and documents the missing references header update
1091 o fixes a bug in processing the envelope test (when testing
1092 multiple envelope elements, the last element determinted the
1095 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1096 Electronic Mail") by including:
1098 Auto-submitted: auto-generated
1100 in the messages that it generates (bounce messages and others, such as
1101 warnings). In the case of bounce messages for non-SMTP mesages, there was
1102 also a typo: it was using "Auto_submitted" (underscore instead of
1103 hyphen). Since every message generated by Exim is necessarily in response
1104 to another message, thes have all been changed to:
1106 Auto-Submitted: auto-replied
1108 in accordance with these statements in the RFC:
1110 The auto-replied keyword:
1112 - SHOULD be used on messages sent in direct response to another
1113 message by an automatic process,
1115 - MUST NOT be used on manually-generated messages,
1117 - MAY be used on Delivery Status Notifications (DSNs) and Message
1118 Disposition Notifications (MDNs),
1120 - MUST NOT be used on messages generated by automatic or periodic
1121 processes, except for messages which are automatic responses to
1124 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1125 to the default Received: header definition.
1127 PH/12 Added log selector acl_warn_skipped (default on).
1129 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1130 variables because (a) they are in the wrong storage pool and (b) even if
1131 they were copied, it wouldn't work properly because of the caching.
1133 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1134 checking when decoding. Apparently there are clients that generate
1135 overlong encoded strings. Why am I not surprised?
1137 PH/15 If the first argument of "${if match_address" was not empty, but did not
1138 contain an "@" character, Exim crashed. Now it writes a panic log message
1139 and treats the condition as false.
1141 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1143 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1144 extension "envelope-auth". The code is finished and in agreement with
1145 other implementations, but there is no documentation so far and in fact,
1146 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1147 not changing the active code.
1149 Print executed "if" and "elsif" statements when debugging is used. This
1150 helps a great deal to understand what a filter does.
1152 Document more things not specified clearly in RFC3028. I had all this
1153 sorted out, when out of a sudden new issues came to my mind. Oops."
1155 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1158 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1159 canonical form (as documented). However, after a host name lookup from
1160 the IP address, check_host() was doing a simple string comparison with
1161 addresses acquired from the DNS when checking that the found name did
1162 have the original IP as one of its addresses. Since any found IPv6
1163 addresses are likely to be in abbreviated form, the comparison could
1164 fail. Luckily, there already exists a function for doing the comparison
1165 by converting both addresses to binary, so now that is used instead of
1166 the text comparison.
1168 PH/20 There was another similar case to PH/19, when a complete host name was
1169 given in a host list; looking up its IP address could give an abbreviated
1170 form, whereas the current host's name might or might not be abbreviated.
1171 The same fix has been applied.
1177 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1178 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1181 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1182 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1184 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1186 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1187 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1188 Also, for "Auto-Submitted", treat anything other than "no" as
1189 non-personal, in accordance with RFC 3834. (Previously it treated
1190 anything starting "auto-" as non-personal.)
1192 TF/01 The control=submission/name=... option had a problem with syntax
1193 errors if the name included a slash character. The /name= option
1194 now slurps the rest of the string, so it can include any characters
1195 but it must come last in the list of options (after /sender_retain
1198 PH/05 Some modifications to the interface to the fake nameserver for the new
1206 TK/01 Added the "success_on_redirect" address verification option. See
1207 NewStuff for rationale and an example.
1209 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1211 PH/02 Patch to exigrep to allow it to work on syslog lines.
1213 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1214 fread() to skip over the body file's header line, because in Cygwin the
1215 header line is locked and is inaccessible.
1217 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1218 co-exist for some time) to make it clear that it is the Exim ID that is
1219 referenced, not the Message-ID: header line.
1221 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1222 string_format() function, because snprintf() does not exist on all
1225 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1226 ${stat: expansion item.
1228 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1229 protocol synchronization error", to keep the pedants happy.
1231 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1232 well as for IRIX systems, when gcc is being used. See the host.c source
1235 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1237 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1240 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1241 transport and to the smtp transport in LMTP mode.
1243 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1245 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1246 run a filter in a subprocess. This could lead to confusion in subsequent
1247 lookups in the parent process. There should also be a search_tidyup() at
1248 the end of the subprocess.
1250 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1251 only if the host matched helo_try_verify_hosts, which caused the
1252 verification to occur when the EHLO/HELO command was issued. The ACL just
1253 tested the remembered result. Now, if a previous verification attempt has
1254 not happened, "verify = helo" does it there and then.
1256 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1258 TK/03 Fix log output including CR from clamd.
1260 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1261 exist provoked a memory error which could cause a segfault.
1263 PH/15 Installed PCRE 6.2
1265 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1267 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1268 of the problem. Specifically, suggested +O2 rather than +O1 for the
1271 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1273 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1274 record, the port was not correctly logged when the outgoing_port log
1275 selector was set (it logged the transort's default port).
1277 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1278 fallback_hosts, and "hosts" in the smtp transport.
1280 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1281 host errors such as "Connection refused".
1283 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1284 authentication with radiusclient 0.4.9:
1286 - Error returned from rc_read_config was caught wrongly
1287 - Username/password not passed on to radius server due to wrong length.
1289 The presumption is that some radiusclient API changes for 4.51/PH/17
1290 were not taken care of correctly. The code is still untested by me (my
1291 Linux distribution still has 0.3.2 of radiusclient), but it was
1292 contributed by a Radius user.
1294 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1295 expanding the "port" option of the smtp transport.
1297 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1298 while reading a MIME header. Thanks to Tom Hughes for a patch.
1300 PH/24 Include config.h inside local_scan.h so that configuration settings are
1303 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1304 that in an ACL for RCPT (for example), you can examine exactly what was
1307 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1308 commands, but it was not correctly comparing the address with the actual
1309 client host address. Thus, it would show the EHLO address in Received:
1310 header lines when this was not necessary.
1312 PH/27 Added the % operator to ${eval:}.
1314 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1315 it should be ignoring failures (because with -C, for example, it has lost
1316 privilege). It wasn't ignoring creation failures other than "already
1319 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1322 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1323 because an input line was too long, either on its own, or by virtue of
1324 too many continuations, the temporary file was not being removed, and the
1325 return code was incorrect.
1327 PH/31 Missing "BOOL" in function definition in filtertest.c.
1329 PH/32 Applied Sieve patches from the maintainer.
1331 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1333 PH/33 Added "verify = not_blind".
1335 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1336 Local/Makefile (with some defaults set). These are used in built scripts
1337 such as exicyclog, but they have never been used in the exim_install
1338 script (though there are many overriding facilities there). I have
1339 arranged that the exim_install script now takes note of these two
1342 PH/35 Installed configuration files for Dragonfly.
1344 PH/36 When a locally submitted message by a trusted user did not contain a
1345 From: header, and the sender address was obtained from -f or from an SMTP
1346 MAIL command, and the trusted user did not use -F to supply a sender
1347 name, $originator_name was incorrectly used when constructing a From:
1348 header. Furthermore, $originator_name was used for submission mode
1349 messages from external hosts without From: headers in a similar way,
1350 which is clearly wrong.
1352 PH/37 Added control=suppress_local_fixups.
1354 PH/38 When log_selector = +received_sender was set, and the addition of the
1355 sender made the log line's construction buffer exactly full, or one byte
1356 less than full, an overflow happened when the terminating "\n" was
1359 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1360 when the result of a list match is failure because a DNS lookup failed.
1362 PH/40 RM_COMMAND is now used in the building process.
1364 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1365 the "build-* directories that it finds.
1367 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1368 address in a domain literal was a prefix of an interface address.
1370 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1371 when verifying a sender address, unless rewrite_headers is false.
1373 PH/44 Wrote a long comment about why errors_to addresses are verified as
1374 recipients, not senders.
1376 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1377 the ratelimit ACL was added.
1379 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1381 PH/46 Added extra information about PostgreSQL errors to the error string.
1383 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1384 avoiding the need to install special zones in a real server. This is
1385 backwards compatible; if it can't find the fake resolver, it drops back.
1386 Thus, both old and new test suites can be run.
1388 TF/02 Added util/ratelimit.pl
1390 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1391 clock is set back in time.
1393 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1394 Candler <B.Candler@pobox.com>.
1396 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1397 OK for addresses that are the result of redirections.
1399 PH/48 A number of further additions for the benefit of the new test suite,
1400 including a fake gethostbyname() that interfaces to the fake DNS resolver
1403 TF/06 The fix for widen_domains has also been applied to qualify_single and
1404 search_parents which are the other dnslookup options that can cause
1407 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1410 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1412 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1413 service" if it encountered an MX record with an empty target hostname.
1414 The message is now "an MX or SRV record indicated no SMTP service".
1416 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1417 if the DNS of the sending site is misconfigured. This is quite a
1418 common situation. This change restores the behaviour of treating a
1419 helo verification defer as a failure.
1421 PH/51 If self=fail was set on a router, the bounce message did not include the
1422 actual error message.
1428 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1430 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1431 command itself ended in error, the underlying message about the transport
1432 filter timeout was being overwritten with the pipe command error. Now the
1433 underlying error message should be appended to the second error message.
1435 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1437 PH/02 Reduce the amount of output that "make" produces by default. Full output
1438 can still be requested.
1440 PH/03 The warning log line about a condition test deferring for a "warn" verb
1441 was being output only once per connection, rather than after each
1442 occurrence (because it was using the same function as for successful
1443 "warn" verbs). This seems wrong, so I have changed it.
1445 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1446 it should not have, which might have caused a crash in the right
1447 circumstances, but probably never did.
1449 PH/04 Installed a modified version of Tony Finch's patch to make submission
1450 mode fix the return path as well as the Sender: header line, and to
1451 add a /name= option so that you can make the user's friendly name appear
1454 TF/03 Added the control = fakedefer ACL modifier.
1456 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1457 Mark Lowes for thorough testing.
1459 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1461 TK/03 Merged latest SRS patch from Miles Wilton.
1463 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1464 with the definition in sysexits.h (which is #included earlier).
1465 Fortunately, Exim does not actually use EX_OK. The code used to try to
1466 preserve the sysexits.h value, by assumimg that macro definitions were
1467 scanned for macro replacements. I have been disabused of this notion,
1468 so now the code just undefines EX_OK before #including unistd.h.
1470 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1471 in the smtp transport. When a block could not be written in a single
1472 write() function, the timeout was being re-applied to each part-write.
1473 This seems wrong - if the receiver was accepting one byte at a time it
1474 would take for ever. The timeout is now adjusted when this happens. It
1475 doesn't have to be particularly precise.
1477 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1478 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1480 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1481 without a domain if the check to <postmaster@domain> fails.
1483 SC/01 Eximstats: added -xls and the ability to specify output files
1484 (patch written by Frank Heydlauf).
1486 SC/02 Eximstats: use FileHandles for outputing results.
1488 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1490 SC/04 Eximstats: fixed display of large numbers with -nvr option
1492 SC/05 Eximstats: fixed merging of reports with empty tables.
1494 SC/06 Eximstats: added the -include_original_destination flag
1496 SC/07 Eximstats: removed tabs and trailing whitespace.
1498 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1500 TK/06 MBOX spool code: Add real "From " MBOX separator line
1501 so the .eml file is really in mbox format (even though
1502 most programs do not really care). Patch from Alex Miller.
1504 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1505 The latter is generated from $received_to and is only set if the
1506 message has one envelope recipient. SA can use these headers,
1507 obviously out-of-the-box. Patch from Alex Miller.
1509 PH/08 The ${def test on a variable was returning false if the variable's
1510 value was "0", contrary to what the specification has always said!
1511 The result should be true unless the variable is empty.
1513 PH/09 The syntax error of a character other than { following "${if
1514 def:variable_name" (after optional whitespace) was not being diagnosed.
1515 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1516 accidental colon was present, for example, could give incorrect results.
1518 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1519 result is used (not including appendfile, where other changes are about
1522 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1523 This involved changing a lot of size variables from int to off_t. It
1524 should work with maildirs and everything.
1526 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1527 spamd dying while we are connected to it.
1529 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1532 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1533 function for BDB 4.3.
1535 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1537 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1538 the sources for PCRE have been re-arranged and more files are now
1541 PH/15 The code I had for printing potentially long long variables in PH/11
1542 above was not the best (it lost precision). The length of off_t variables
1543 is now inspected at build time, and an appropriate printing format (%ld
1544 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1545 to be "long long int" or "long int". This is needed for the internal
1546 formatting function string_vformat().
1548 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1549 the configuration file to be ":syslog", then the script "guesses" where
1550 the logs files are, rather than using the compiled in default. In our
1551 case the guess is not the same as the compiled default, so the script
1552 suddenly stopped working when I started to use syslog. The patch checks
1553 to see if log_file_path is "". If so, it attempts to read it from exim
1554 with no configuration file to get the compiled in version, before it
1555 falls back to the previous guessing code."
1557 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1558 implementing BATV in an Exim configuration. See NewStuff for the gory
1561 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1562 Makefile that are specific to HP-UX.
1564 PH/18 If the "use_postmaster" option was set for a recipient callout together
1565 with the "random" option, the postmaster address was used as the MAIL
1566 FROM address for the random test, but not for the subsequent recipient
1567 test. It is now used for both.
1569 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1570 patch removes a few documentation additions to RFC 3028, because the
1571 latest draft now contains them. It adds the new en;ascii-case comparator
1572 and a new error check for 8bit text in MIME parts. Comparator and
1573 require names are now matched exactly. I enabled the subaddress
1574 extension, but it is not well tested yet (read: it works for me)."
1576 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1577 rework some of the code of TK/09 above to avoid the hardwired use of
1578 "%lld" and "long long". Replaced the call to snprintf() with a call to
1581 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1582 records point to non-existent hosts", "retry timeout exceeded", and
1583 "retry time not reached for any host after a long failure period".
1585 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1586 experimental DomainKeys support:
1588 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1589 (2) On an error such as an illegally used "control", the wrong name for
1590 the control was given.
1592 These problems did NOT occur unless DomainKeys support was compiled.
1594 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1596 PH/24 Added ${if match_ip condition.
1598 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1599 databases so that it will be absolutely obvious if a crash occurs in the
1600 DB library. This is a regular occurrence (often caused by mis-matched
1603 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1604 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1605 picked up on a user's system that detects such things. There doesn't seem
1606 to be a gcc warning option for this - only an attribute that has to be
1607 put on the function's prototype. It seems that in Fedora Core 4 they have
1608 set this on a number of new functions. No doubt there will be more in due
1611 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1612 specify a transport. However, if an address that was verified by such a
1613 router was the subject of a callout, Exim crashed because it tried to
1614 read the rcpt_include_affixes from the non-existent transport. Now it
1615 just assumes that the setting of that option is false. This bug was
1616 introduced by 4.51/PH/31.
1618 PH/28 Changed -d+all to exclude +memory, because that information is very
1619 rarely of interest, but it makes the output a lot bigger. People tend to
1620 do -d+all out of habit.
1622 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1623 code in os-type was giving problems when libc.so lives in lib64, like on
1626 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1627 aren't the modern standard, and it seems that some systems' include files
1628 don't always have them. Exim was already checking for some of the newer
1629 ones like T_AAAA, and defining it itself. I've added checks for all the
1630 record types that Exim uses.
1632 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1633 not automatically generating a new one, as it is supposed to. This
1634 prevented TLS from working. If the file did exist, but contained invalid
1635 data, a new version was generated, as expected. It was only the case of a
1636 non-existent file that was broken.
1638 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1639 with a change in libdomainkeys > 0.64.
1641 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1642 from DNS. If the selector record carries the flag, it now has
1643 precedence over the domain-wide flag.
1645 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1647 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1648 the use of a port name that isn't defined in /etc/services) occurred, the
1649 message was deferred as in a normal delivery, and thus remained on the
1650 spool, instead of being failed because of the mua_wrapper setting. This
1651 is now fixed, and I tidied up some of the mua_wrapper messages at the
1654 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1655 the messages in a hash of arrays rather than using individual hashes.
1656 This is a bit cleaner and results in dramatic memory savings, albeit
1657 at a slight CPU cost.
1659 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1660 as requested by Marc Sherman.
1662 SC/10 Eximstats: added histograms for user specified patterns as requested
1665 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1667 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1668 fopen() in the content-scanning modules that did not already have it.
1674 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1675 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1677 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1679 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1681 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1682 to test Sieve filters that use "vacation".
1684 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1685 that changes the way the GnuTLS parameters are stored in the cache file.
1686 The new format can be generated externally. For backward compatibility,
1687 if the data in the cache doesn't make sense, Exim assumes it has read an
1688 old-format file, and it generates new data and writes a new file. This
1689 means that you can't go back to an older release without removing the
1692 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1693 work if there are any delivery delays because "one_time" forces the
1694 parent to be marked "delivered", so its unseen clone is never tried
1695 again. For this reason, Exim now forbids the simultaneous setting of
1698 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1699 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1700 entry if you want to know the details. The fix, however, neglected to
1701 consider the case when local delivery batching is involved. The test for
1702 "previously delivered" was not happening when checking to see if an
1703 address could be batched with a previous (undelivered) one; under
1704 certain circumstances this could lead to multiple deliveries to the same
1707 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1708 in its include files, and this causes problems building Exim.
1710 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1711 header_syntax) but Exim was just ignoring anything given after a slash.
1712 In particular, this caused confusion with an attempt to use "verify =
1713 reverse_host_lookup/defer_ok". An error is now given when options are
1714 supplied for verify items that do not have them. (Maybe reverse_host_
1715 lookup should have a defer_ok option, but that's a different point.)
1717 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1718 defined by RFC 821) to 2048, because there were problems with some AUTH
1719 commands, and RFC 1869 says the size should be increased for extended
1720 SMTP commands that take arguments.
1722 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1725 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1726 "unknown" error; now it says that the functionality isn't in the binary.
1728 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1729 an address' error message when a string expansion fails (syntax or
1730 whatever). Otherwise the password may appear in the log. Following change
1731 PH/42 below, there is no longer a chance of it appearing in a bounce
1734 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1736 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1737 of its addresses was ever tried. (Bugzilla bug #2.)
1739 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1740 the result incorrectly in the debug output. (It correctly added a newline
1741 to what was transported.)
1743 TF/01 Added $received_time.
1745 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1746 commented out examples of how to interface to a virus scanner and to
1747 SpamAssassin. Also added commented examples of av_scanner and
1748 spamd_address settings.
1750 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1751 and controls are allowed in which ACLs. There were a couple of minor
1752 errors. Some of the entries in the conditions table (which is a table of
1753 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1754 negation of where the condition IS allowed.
1756 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1758 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1759 header file does not have a version number, so I've had to invent a new
1760 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1761 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1762 radiusclient), but it was contributed by a Radius user.
1764 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1765 files or directories, for OpenSSL.
1767 PH/19 When an Exim process that is running as root has to create an Exim log
1768 file, it does so in a subprocess that runs as exim:exim so as to get the
1769 ownership right at creation (otherwise, other Exim processes might see
1770 the file with the wrong ownership). There was no test for failure of this
1771 fork() call, which would lead to the process getting stuck as it waited
1772 for a non-existent subprocess. Forks do occasionally fail when resources
1773 run out. I reviewed all the other calls to fork(); they all seem to check
1776 PH/20 When checking for unexpected SMTP input at connect time (before writing
1777 the banner), Exim was not dealing correctly with a non-positive return
1778 from the read() function. If the client had disconnected by this time,
1779 the result was a log entry for a synchronization error with an empty
1780 string after "input=" when read() returned zero. If read() returned -1
1781 (an event I could not check), uninitialized data bytes were printed.
1782 There were reports of junk text (parts of files, etc) appearing after
1785 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1787 PH/22 Added support for macro redefinition, and (re)definition in between
1788 driver and ACL definitions.
1790 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1791 forgetting to use the resulting value; it was using the unexpanded value.
1793 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1794 hadn't been configured. The fix is from Juergen Kreileder, who
1795 understands it better than I do:
1797 "Here's what I see happening with three configured cyrus_sasl
1798 authenticators configured (plain, login, cram-md5):
1800 On startup auth_cyrus_sasl_init() gets called for each of these.
1801 This means three calls to sasl_listmech() without a specified mech_list.
1802 => SASL tests which mechs of all available mechs actually work
1803 => three warnings about OTP not working
1804 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1806 With the patch, sasl_listmech() also gets called three times. But now
1807 SASL's mech_list option is set to the server_mech specified in the the
1808 authenticator. Or in other words, the answer from sasl_listmech()
1809 gets limited to just the mech you're testing for (which is different
1811 => the return list contains just 'plain' or 'login', 'cram-md5' or
1812 nothing depending on the value of ob->server_mech.
1814 I've just tested the patch: Authentication still works fine,
1815 unavailable mechs specified in the exim configuration are still
1816 caught, and the auth.log warnings about OTP are gone."
1818 PH/25 When debugging is enabled, the contents of the command line are added
1819 to the debugging output, even when log_selector=+arguments is not
1822 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1823 answer is "GNU", and only if the return is "GNU/something" is the answer
1826 PH/27 $acl_verify_message is now set immediately after the failure of a
1827 verification in an ACL, and so is available in subsequent modifiers. In
1828 particular, the message can be preserved by coding like this:
1830 warn !verify = sender
1831 set acl_m0 = $acl_verify_message
1833 Previously, $acl_verify_message was set only while expanding "message"
1834 and "log_message" when a very denied access.
1836 PH/28 Modified OS/os.c-Linux with
1838 -#ifndef OS_LOAD_AVERAGE
1839 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1841 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1842 nomenclature these days.)
1844 PH/29 Installed patch from the Sieve maintainer that adds the options
1845 sieve_useraddress and sieve_subaddress to the redirect router.
1847 PH/30 In these circumstances:
1848 . Two addresses routed to the same list of hosts;
1849 . First host does not offer TLS;
1850 . First host accepts first address;
1851 . First host gives temporary error to second address;
1852 . Second host offers TLS and a TLS session is established;
1853 . Second host accepts second address.
1854 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1855 and peerdn, if requested) that were in fact used only for the second
1858 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1859 attention to any local part prefix or suffix that was matched by the
1860 router that accepted the address. It now behaves in the same way as it
1861 does for delivery: the affixes are removed from the local part unless
1862 rcpt_include_affixes is set on the transport.
1864 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1865 timeout during the DATA phase of an incoming message.
1867 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1868 applied a patch sanctioned by the Sieve maintainer.
1870 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1871 the uid or gid is negative. A case of a negative gid caused this to be
1872 noticed. The fix allows for either to be negative.
1874 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1875 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1878 PH/36 The change PH/12 above was broken. Fixed it.
1880 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1881 the grounds that routing the same address twice would always produce the
1882 same answer. This might have been true once, but it is certainly no
1883 longer true now. Routing a child address may depend on the previous
1884 routing that produced that child. Some complicated redirection strategies
1885 went wrong when messages had multiple recipients, and made Exim's
1886 behaviour dependent on the order in which the addresses were given.
1888 I have moved the duplicate checking until after the routing is complete.
1889 Exim scans the addresses that are assigned to local and remote
1890 transports, and removes any duplicates. This means that more work will be
1891 done, as duplicates will always all be routed, but duplicates are
1892 presumably rare, so I don't expect this is of any significance.
1894 For deliveries to pipes, files, and autoreplies, the duplicate checking
1895 still happens during the routing process, since they are not going to be
1898 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1899 It corrects a timeout issue with spamd. This is Ian's comment: "The
1900 background is that sometimes spamd either never reads data from a
1901 connection it has accepted, or it never writes response data. The exiscan
1902 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1903 blindly assumes that writes won't block so it may never time out."
1905 PH/39 Allow G after quota size as well as K and M.
1907 PH/40 The value set for $authenticated_id in an authenticator may not contain
1908 binary zeroes or newlines because the value is written to log lines and
1909 to spool files. There was no check on this. Now the value is run through
1910 the string_printing() function so that such characters are converted to
1911 printable escape sequences.
1913 PH/41 $message_linecount is a new variable that contains the total number of
1914 lines in the message. Compare $body_linecount, which is the count for the
1917 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1918 bounce and delay warning messages, except in certain special cases, which
1921 (a) An SMTP error message from a remote host;
1922 (b) A message specified in a :fail: redirection;
1923 (c) A message specified in a "fail" command in a system filter;
1924 (d) A message specified in a FAIL return from the queryprogram router;
1925 (e) A message specified by the cannot_route_message router option.
1927 In these cases only, Exim does include the error details in bounce and
1928 warning messages. There are also a few cases where bland messages such
1929 as "unrouteable address" or "local delivery error" are given.
1931 PH/43 $value is now also set for the "else" part of a ${run expansion.
1933 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1934 being worked on, but at least Exim now implements the latest version to
1937 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1938 process to complete was treated as a delivery failure, a timeout while
1939 writing the message to the pipe was logged, but erroneously treated as a
1940 successful delivery. Such timeouts include transport filter timeouts. For
1941 consistency with the overall process timeout, these timeouts are now
1942 treated as errors, giving rise to delivery failures by default. However,
1943 there is now a new Boolean option for the pipe transport called
1944 timeout_defer, which, if set TRUE, converts the failures into defers for
1945 both kinds of timeout. A transport filter timeout is now identified in
1948 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1949 systems where "make" and "gmake" are different, calling "gmake" at top
1950 level broke things. I've arranged for the value of $(MAKE) to be passed
1951 from the Makefile to this script so that it can call the same version of
1955 A note about Exim versions 4.44 and 4.50
1956 ----------------------------------------
1958 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1959 changes of various kinds. As a consequence, a big documentation update was
1960 needed. This delayed the release for rather longer than seemed good, especially
1961 in the light of a couple of (minor) security issues. Therefore, the changes
1962 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1963 release. So 4.44 and 4.50 are in effect two different branches that both start
1966 I have left the 4.50 change log unchanged; it contains all the changes since
1967 4.43. The change log for 4.44 is below; many of its items are identical to
1968 those for 4.50. This seems to be the most sensible way to preserve the
1969 historical information.
1975 1. Minor wording change to the doc/README.SIEVE file.
1977 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1978 computation of the current number of files was incorrect.
1980 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1981 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1984 4. Give more explanation in the error message when the command for a transport
1985 filter fails to execute.
1987 5. There are several places where Exim runs a non-Exim command in a
1988 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1989 was being done only for the command run by the queryprogram router. It is
1990 now done for all such subprocesses. The other cases are: ${run, transport
1991 filters, and the commands run by the lmtp and pipe transports.
1993 6. Added CONFIGURE_GROUP build-time option.
1995 7. Some older OS have a limit of 256 on the maximum number of file
1996 descriptors. Exim was using setrlimit() to set 1000 as a large value
1997 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1998 systems. I've change it so that if it can't get 1000, it tries for 256.
2000 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2001 was an oversight, and furthermore, ever since the addition of extra
2002 controls (e.g. 4.43/32), the checks on when to allow different forms of
2003 "control" were broken. There should now be diagnostics for all cases when a
2004 control that does not make sense is encountered.
2006 9. Added the /retain_sender option to "control=submission".
2008 10. $recipients is now available in the predata ACL (oversight).
2010 11. Tidy the search cache before the fork to do a delivery from a message
2011 received from the command line. Otherwise the child will trigger a lookup
2012 failure and thereby defer the delivery if it tries to use (for example) a
2013 cached ldap connection that the parent has called unbind on.
2015 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2016 of $address_data from the recipient verification was clobbered by the
2017 sender verification.
2019 13. The value of address_data from a sender verification is now available in
2020 $sender_address_data in subsequent conditions in the ACL statement.
2022 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2024 15. Added a new option "connect=<time>" to callout options, to set a different
2027 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2028 was its contents. (It was OK if the option was not defined at all.)
2030 17. A "Completed" log line is now written for messages that are removed from
2031 the spool by the -Mrm option.
2033 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2034 information about exactly what failed.
2036 19. Added -dd to debug only the daemon process.
2038 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2039 handles timeouts, both on the server side and network timeouts. Renamed the
2040 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2042 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2043 It is now set to "smtps".
2045 22. $host_address is now set to the target address during the checking of
2046 ignore_target_hosts.
2048 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2049 being passed; this would have caused $sender_host_name to have been used if
2050 matching the list had actually called for a host name (not very likely,
2051 since this list is usually IP addresses). A host name is now passed as
2054 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2055 code that specifies a non-restarting handler (typically sigaction() in
2056 modern systems) in an attempt to fix a rare and obscure crash bug.
2058 25. Narrowed the window for a race in the daemon that could cause it to ignore
2059 SIGCHLD signals. This is not a major problem, because they are used only to
2060 wake it up if nothing else does.
2062 26. A malformed maildirsize file could cause Exim to calculate negative values
2063 for the mailbox size or file count. Odd effects could occur as a result.
2064 The maildirsize information is now recalculated if the size or filecount
2067 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2068 support for a long time. Removed HAVE_SYS_VFS_H.
2070 28. Installed the latest version of exipick from John Jetmore.
2072 29. In an address list, if the pattern was not a regular expression, an empty
2073 subject address (from a bounce message) matched only if the pattern was an
2074 empty string. Non-empty patterns were not even tested. This was the wrong
2075 because it is perfectly reasonable to use an empty address as part of a
2076 database query. An empty address is now tested by patterns that are
2077 lookups. However, all the other forms of pattern expect the subject to
2078 contain a local part and a domain, and therefore, for them, an empty
2079 address still always fails if the pattern is not itself empty.
2081 30. Exim went into a mad DNS loop when attempting to do a callout where the
2082 host was specified on an smtp transport, and looking it up yielded more
2083 than one IP address.
2085 31. Re-factored the code for checking spool and log partition space into a
2086 function that finds that data and another that does the check. The former
2087 is then used to implement four new variables: $spool_space, $log_space,
2088 $spool_inodes, and $log_inodes.
2090 32. The RFC2047 encoding function was originally intended for short strings
2091 such as real names; it was not keeping to the 75-character limit for
2092 encoded words that the RFC imposes. It now respects the limit, and
2093 generates multiple encoded words if necessary. To be on the safe side, I
2094 have increased the buffer size for the ${rfc2047: expansion operator from
2097 33. It is now permitted to omit both strings after an "if" condition; if the
2098 condition is true, the result is "true". As before, when the second string
2099 is omitted, a false condition yields an empty string. This makes it less
2100 cumbersome to write custom ACL and router conditions.
2102 34. Failure to deliver a bounce message always caused it to be frozen, even if
2103 there was an errors_to setting on the router. The errors_to setting is now
2106 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2107 canonical form (fully expanded) before being placed in
2108 $sender_host_address.
2110 36. The table in the code that translates DNS record types into text (T_A to
2111 "A" for instance) was missing entries for NS and CNAME. It is just possible
2112 that this could have caused confusion if both these types were looked up
2113 for the same domain, because the text type is used as part of Exim's
2114 per-process caching. But the chance of anyone hitting this buglet seems
2117 37. The dnsdb lookup has been extended in a number of ways.
2119 (1) There is a new type, "zns", which walks up the domain tree until it
2120 finds some nameserver records. It should be used with care.
2122 (2) There is a new type, "mxh", which is like "mx" except that it returns
2123 just the host names, not the priorities.
2125 (3) It is now possible to give a list of domains (or IP addresses) to be
2126 looked up. The behaviour when one of the lookups defers can be
2127 controlled by a keyword.
2129 (4) It is now possible to specify the separator character for use when
2130 multiple records are returned.
2132 38. The dnslists ACL condition has been extended: it is now possible to supply
2133 a list of IP addresses and/or domains to be looked up in a particular DNS
2136 39. Added log_selector=+queue_time_overall.
2138 40. When running the queue in the test harness, wait just a tad after forking a
2139 delivery process, to get repeatability of debugging output.
2141 41. Include certificate and key file names in error message when GnuTLS fails
2142 to set them up, because the GnuTLS error message doesn't include the name
2143 of the failing file when there is a problem reading it.
2145 42. Allow both -bf and -bF in the same test run.
2147 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2149 44. The "Exiscan patch" is now merged into the mainline Exim source.
2151 45. Sometimes the final signoff response after QUIT could fail to get
2152 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2153 < 0 before doing a fflush(). This bug looks as though it goes back to the
2154 introduction of TLS in release 3.20, but "sometimes" must have been rare
2155 because the tests only now provoked it.
2157 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2158 (this can affect the format of dates).
2160 47. exim_tidydb, when checking for the continued existence of a message for
2161 which it has found a message-specific retry record, was not finding
2162 messages that were in split spool directories. Consequently, it was
2163 deleting retry records that should have stayed in existence.
2165 48. Steve fixed some bugs in eximstats.
2167 49. The SPA authentication driver was not abandoning authentication and moving
2168 on to the next authenticator when an expansion was forced to fail,
2169 contradicting the general specification for all authenticators. Instead it
2170 was generating a temporary error. It now behaves as specified.
2172 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2173 (the order specifies the preference for clients). The order is now AES256,
2174 AES128, 3DES, ARCFOUR128.
2176 51. Small patch to Sieve code - explicitly set From: when generating an
2179 52. Exim crashed if a remote delivery caused a very long error message to be
2180 recorded - for instance if somebody sent an entire SpamAssassin report back
2181 as a large number of 550 error lines. This bug was coincidentally fixed by
2182 increasing the size of one of Exim's internal buffers (big_buffer) that
2183 happened as part of the Exiscan merge. However, to be on the safe side, I
2184 have made the code more robust (and fixed the comments that describe what
2187 53. Now that there can be additional text after "Completed" in log lines (if
2188 the queue_time_overall log selector is set), a one-byte patch to exigrep
2189 was needed to allow it to recognize "Completed" as not the last thing in
2192 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2193 patch that reportedly fixes this has been added. I am not expert enough to
2194 create a test for it. This is what the patch creator wrote:
2196 "I found a little strange behaviour of ldap code when working with
2197 Windows 2003 AD Domain, where users was placed in more than one
2198 Organization Units. When I tried to give exim partial DN, the exit code
2199 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2200 But simultaneously result of request was absolutely normal ldap result,
2201 so I produce this patch..."
2203 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2204 REFERENCE, so I have modified the code to exclude the patch when that macro
2207 55. Some experimental protocols are using DNS PTR records for new purposes. The
2208 keys for these records are domain names, not reversed IP addresses. The
2209 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2210 leaves it alone. Component reversal etc. now happens only for IP addresses.
2213 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2215 57. Double the size of the debug message buffer (to 2048) so that more of very
2216 long debug lines gets shown.
2218 58. The exicyclog utility now does better if the number of log files to keep
2219 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2221 59. Two changes related to the smtp_active_hostname option:
2223 (1) $smtp_active_hostname is now available as a variable.
2224 (2) The default for smtp_banner uses $smtp_active_hostname instead
2225 of $primary_hostname.
2227 60. The host_aton() function is supposed to be passed a string that is known
2228 to be a valid IP address. However, in the case of IPv6 addresses, it was
2229 not checking this. This is a hostage to fortune. Exim now panics and dies
2230 if the condition is not met. A case was found where this could be provoked
2231 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2232 components; fortuitously, this particular loophole had already been fixed
2233 by change 4.50/55 above.
2235 If there are any other similar loopholes, the new check in host_aton()
2236 itself should stop them being exploited. The report I received stated that
2237 data on the command line could provoke the exploit when Exim was running as
2238 exim, but did not say which command line option was involved. All I could
2239 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2240 running as the user.
2243 61. There was a buffer overflow vulnerability in the SPA authentication code
2244 (which came originally from the Samba project). I have added a test to the
2245 spa_base64_to_bits() function which I hope fixes it.
2248 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2249 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2251 63. The daemon start-up calls getloadavg() while still root for those OS that
2252 need the first call to be done as root, but it missed one case: when
2253 deliver_queue_load_max is set with deliver_drop_privilege. This is
2254 necessary for the benefit of the queue runner, because there is no re-exec
2255 when deliver_drop_privilege is set.
2257 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2258 This has been fixed.
2260 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2261 was in use, was not putting the data itself into the right store pool;
2262 consequently, it could be overwritten for a subsequent message in the same
2263 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2266 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2268 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2269 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2270 the function were treating the return as a boolean value, which happened to
2271 work because 0=false and not-0=true, but is not correct code.
2273 68. The host_aton() function was not handling scoped IPv6 addresses (those
2274 with, for example, "%eth0" on the end) correctly.
2276 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2277 negated items (that is, ~something) in unsigned ints. Some compilers
2278 apparently mutter when there is no cast.
2280 70. If an address verification called from an ACL failed, and did not produce a
2281 user-specific message (i.e. there was only a "system" message), nothing was
2282 put in $acl_verify_message. In this situation, it now puts the system
2285 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2286 session; change 4.31/43 added the unwanted input to the log line - except
2287 that it did not do this in the start of session case. It now does.
2289 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2290 This is wrong and can cause the other end to generate a synchronization
2291 error if it is another Exim or anything else that does the synchronization
2292 check. A QUIT command is no longer sent after a timeout.
2294 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2295 during host lookups.
2297 74. The defer_ok option of callout verification was not working if it was used
2298 when verifying addresses in header lines, that is, for this case:
2300 verify = header_sender/callout=defer_ok
2302 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2303 those file descriptors could be used for SMTP connections. If anything
2304 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2305 could be sent to the SMTP client, causing chaos. The daemon now opens
2306 stdin, stdout, and stderr to /dev/null when it puts itself into the
2309 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2310 log by default. The user can override this with suitable Perl magic.
2312 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2313 the log message when discard triggers, was not working for the DATA ACL or
2314 for the non-SMTP ACL.
2316 78. Error message wording change in sieve.c.
2318 79. If smtp_accept_max_per_host was set, the number of connections could be
2319 restricted to fewer than expected, because the daemon was trying to set up
2320 a new connection before checking whether the processes handling previous
2321 connections had finished. The check for completed processes is now done
2322 earlier. On busy systems, this bug wouldn't be noticed because something
2323 else would have woken the daemon, and it would have reaped the completed
2326 80. If a message was submitted locally by a user whose login name contained one
2327 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2328 It caused a spool format error. I have fixed the spool reading code. A
2329 related problem was that the "from" clause in the Received: line became
2330 illegal because of the space(s). It is now covered by ${quote_local_part.
2332 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2335 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2337 83. Patch from Sieve maintainer to fix unterminated string problem in
2338 "vacation" handling.
2340 84. Some minor changes to the Linux configuration files to help with other
2341 OS variants using glibc.
2343 85. One more patch for Sieve to update vacation handling to latest spec.
2346 ----------------------------------------------------
2347 See the note above about the 4.44 and 4.50 releases.
2348 ----------------------------------------------------
2354 1. Change 4.43/35 introduced a bug that caused file counts to be
2355 incorrectly computed when quota_filecount was set in an appendfile
2358 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2359 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2362 3. Give more explanation in the error message when the command for a transport
2363 filter fails to execute.
2365 4. There are several places where Exim runs a non-Exim command in a
2366 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2367 was being done only for the command run by the queryprogram router. It is
2368 now done for all such subprocesses. The other cases are: ${run, transport
2369 filters, and the commands run by the lmtp and pipe transports.
2371 5. Some older OS have a limit of 256 on the maximum number of file
2372 descriptors. Exim was using setrlimit() to set 1000 as a large value
2373 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2374 systems. I've change it so that if it can't get 1000, it tries for 256.
2376 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2377 was an oversight, and furthermore, ever since the addition of extra
2378 controls (e.g. 4.43/32), the checks on when to allow different forms of
2379 "control" were broken. There should now be diagnostics for all cases when a
2380 control that does not make sense is encountered.
2382 7. $recipients is now available in the predata ACL (oversight).
2384 8. Tidy the search cache before the fork to do a delivery from a message
2385 received from the command line. Otherwise the child will trigger a lookup
2386 failure and thereby defer the delivery if it tries to use (for example) a
2387 cached ldap connection that the parent has called unbind on.
2389 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2390 of $address_data from the recipient verification was clobbered by the
2391 sender verification.
2393 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2394 was its contents. (It was OK if the option was not defined at all.)
2396 11. A "Completed" log line is now written for messages that are removed from
2397 the spool by the -Mrm option.
2399 12. $host_address is now set to the target address during the checking of
2400 ignore_target_hosts.
2402 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2403 being passed; this would have caused $sender_host_name to have been used if
2404 matching the list had actually called for a host name (not very likely,
2405 since this list is usually IP addresses). A host name is now passed as
2408 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2409 code that specifies a non-restarting handler (typically sigaction() in
2410 modern systems) in an attempt to fix a rare and obscure crash bug.
2412 15. Narrowed the window for a race in the daemon that could cause it to ignore
2413 SIGCHLD signals. This is not a major problem, because they are used only to
2414 wake it up if nothing else does.
2416 16. A malformed maildirsize file could cause Exim to calculate negative values
2417 for the mailbox size or file count. Odd effects could occur as a result.
2418 The maildirsize information is now recalculated if the size or filecount
2421 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2422 support for a long time. Removed HAVE_SYS_VFS_H.
2424 18. Updated exipick to current release from John Jetmore.
2426 19. Allow an empty sender to be matched against a lookup in an address list.
2427 Previously the only cases considered were a regular expression, or an
2430 20. Exim went into a mad DNS lookup loop when doing a callout where the
2431 host was specified on the transport, if the DNS lookup yielded more than
2434 21. The RFC2047 encoding function was originally intended for short strings
2435 such as real names; it was not keeping to the 75-character limit for
2436 encoded words that the RFC imposes. It now respects the limit, and
2437 generates multiple encoded words if necessary. To be on the safe side, I
2438 have increased the buffer size for the ${rfc2047: expansion operator from
2441 22. Failure to deliver a bounce message always caused it to be frozen, even if
2442 there was an errors_to setting on the router. The errors_to setting is now
2445 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2446 canonical form (fully expanded) before being placed in
2447 $sender_host_address.
2449 24. Updated eximstats to version 1.33
2451 25. Include certificate and key file names in error message when GnuTLS fails
2452 to set them up, because the GnuTLS error message doesn't include the name
2453 of the failing file when there is a problem reading it.
2455 26. Expand error message when OpenSSL has problems setting up cert/key files.
2458 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2459 (this can affect the format of dates).
2461 28. exim_tidydb, when checking for the continued existence of a message for
2462 which it has found a message-specific retry record, was not finding
2463 messages that were in split spool directories. Consequently, it was
2464 deleting retry records that should have stayed in existence.
2466 29. eximstats updated to version 1.35
2467 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2468 1.35 - bugfix such that pie charts by volume are generated correctly
2470 30. The SPA authentication driver was not abandoning authentication and moving
2471 on to the next authenticator when an expansion was forced to fail,
2472 contradicting the general specification for all authenticators. Instead it
2473 was generating a temporary error. It now behaves as specified.
2475 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2476 (the order specifies the preference for clients). The order is now AES256,
2477 AES128, 3DES, ARCFOUR128.
2479 31. Small patch to Sieve code - explicitly set From: when generating an
2482 32. Exim crashed if a remote delivery caused a very long error message to be
2483 recorded - for instance if somebody sent an entire SpamAssassin report back
2484 as a large number of 550 error lines. This bug was coincidentally fixed by
2485 increasing the size of one of Exim's internal buffers (big_buffer) that
2486 happened as part of the Exiscan merge. However, to be on the safe side, I
2487 have made the code more robust (and fixed the comments that describe what
2490 33. Some experimental protocols are using DNS PTR records for new purposes. The
2491 keys for these records are domain names, not reversed IP addresses. The
2492 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2493 leaves it alone. Component reversal etc. now happens only for IP addresses.
2496 34. The host_aton() function is supposed to be passed a string that is known
2497 to be a valid IP address. However, in the case of IPv6 addresses, it was
2498 not checking this. This is a hostage to fortune. Exim now panics and dies
2499 if the condition is not met. A case was found where this could be provoked
2500 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2501 components; fortuitously, this particular loophole had already been fixed
2502 by change 4.50/55 or 4.44/33 above.
2504 If there are any other similar loopholes, the new check in host_aton()
2505 itself should stop them being exploited. The report I received stated that
2506 data on the command line could provoke the exploit when Exim was running as
2507 exim, but did not say which command line option was involved. All I could
2508 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2509 running as the user.
2512 35. There was a buffer overflow vulnerability in the SPA authentication code
2513 (which came originally from the Samba project). I have added a test to the
2514 spa_base64_to_bits() function which I hope fixes it.
2517 36. The daemon start-up calls getloadavg() while still root for those OS that
2518 need the first call to be done as root, but it missed one case: when
2519 deliver_queue_load_max is set with deliver_drop_privilege. This is
2520 necessary for the benefit of the queue runner, because there is no re-exec
2521 when deliver_drop_privilege is set.
2523 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2524 was in use, was not putting the data itself into the right store pool;
2525 consequently, it could be overwritten for a subsequent message in the same
2526 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2529 38. Sometimes the final signoff response after QUIT could fail to get
2530 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2531 < 0 before doing a fflush(). This bug looks as though it goes back to the
2532 introduction of TLS in release 3.20, but "sometimes" must have been rare
2533 because the tests only now provoked it.
2539 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2540 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2541 Now it returns an integer. A number of calls were still expecting a T/F
2542 return. Fortuitously, in all cases, the tests worked in OK situations,
2543 which is the norm. However, things would have gone wrong on any write
2544 failures on the smtp file descriptor. This function is used when sending
2545 messages over SMTP and also when doing verify callouts.
2547 2. When Exim is called to do synchronous delivery of a locally submitted
2548 message (the -odf or -odi options), it no longer closes stderr before doing
2551 3. Implemented the mua_wrapper option.
2553 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2555 5. Implemented the functions header_remove(), header_testname(),
2556 header_add_at_position(), and receive_remove_recipient(), and exported them
2559 6. If an ACL "warn" statement specified the addition of headers, Exim already
2560 inserted X-ACL-Warn: at the start if there was no header name. However, it
2561 was not making this test for the second and subsequent header lines if
2562 there were newlines in the string. This meant that an invalid header could
2563 be inserted if Exim was badly configured.
2565 7. Allow an ACL "warn" statement to add header lines at the start or after all
2566 the Received: headers, as well as at the end.
2568 8. Added the rcpt_4xx retry error code.
2570 9. Added postmaster_mailfrom=xxx to callout verification option.
2572 10. Added mailfrom=xxxx to the callout verification option, for verify=
2575 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2576 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2578 12. Inserted some casts to stop certain compilers warning when using pointer
2579 differences as field lengths or precisions in printf-type calls (mostly
2580 affecting debugging statements).
2582 13. Added optional readline() support for -be (dynamically loaded).
2584 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2585 same clock tick as a message's arrival, so that its received time was the
2586 same as the "first fail" time on the retry record, and that message
2587 remained on the queue past the ultimate address timeout, every queue runner
2588 would try a delivery (because it was past the ultimate address timeout) but
2589 after another failure, the ultimate address timeout, which should have then
2590 bounced the address, did not kick in. This was a "< instead of <=" error;
2591 in most cases the first failure would have been in the next clock tick
2592 after the received time, and all would be well.
2594 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2595 being recognized when the domain list was tested by the match_domain
2596 condition in an expansion string.
2598 16. Added the ${str2b64: operator.
2600 17. Exim was always calling setrlimit() to set a large limit for the number of
2601 processes, without checking whether the existing limit was already
2602 adequate. (It did check for the limit on file descriptors.) Furthermore,
2603 errors from getrlimit() and setrlimit() were being ignored. Now they are
2604 logged to the main and panic logs, but Exim does carry on, to try to do its
2605 job under whatever limits there are.
2607 18. Imported PCRE 5.0.
2609 19. Trivial typo in log message " temporarily refused connection" (the leading
2612 20. If the log selector return_path_on_delivery was set and an address was
2613 redirected to /dev/null, the delivery process crashed because it assumed
2614 that a return path would always be set for a "successful" delivery. In this
2615 case, the whole delivery is bypassed as an optimization, and therefore no
2618 21. Internal re-arrangement: the function for sending a challenge and reading
2619 a response while authentication was assuming a zero-terminated challenge
2620 string. It's now changed to take a pointer and a length, to allow for
2621 binary data in such strings.
2623 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2625 23. Exim was not respecting finduser_retries when seeking the login of the
2626 uid under which it was called; it was always trying 10 times. (The default
2627 setting of finduser_retries is zero.) Also, it was sleeping after the final
2628 failure, which is pointless.
2630 24. Implemented tls_on_connect_ports.
2632 25. Implemented acl_smtp_predata.
2634 26. If the domain in control=submission is set empty, Exim assumes that the
2635 authenticated id is a complete email address when it generates From: or
2636 Sender: header lines.
2638 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2639 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2640 chown and chgrp in /bin and hostname in /usr/bin.
2642 28. Exim was keeping the "process log" file open after each use, just as it
2643 does for the main log. This opens the possibility of it remaining open for
2644 long periods when the USR1 signal hits a daemon. Occasional processlog
2645 errors were reported, that could have been caused by this. Anyway, it seems
2646 much more sensible not to leave this file open at all, so that is what now
2649 29. The long-running daemon process does not normally write to the log once it
2650 has entered its main loop, and it closes the log before doing so. This is
2651 so that log files can straightforwardly be renamed and moved. However,
2652 there are a couple of unusual error situations where the daemon does write
2653 log entries, and I had neglected to close the log afterwards.
2655 30. The text of an SMTP error response that was received during a remote
2656 delivery was being truncated at 512 bytes. This is too short for some of
2657 the long messages that one sometimes sees. I've increased the limit to
2660 31. It is now possible to make retry rules that apply only when a message has a
2661 specific sender, in particular, an empty sender.
2663 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2664 it possible to be selective about when SMTP synchronization is enforced.
2666 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2668 32. Implemented hosts_connection_nolog.
2670 33. Added an ACL for QUIT.
2672 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2675 35. Added mailbox_size and mailbox_filecount to appendfile.
2677 36. Added control = no_multiline_responses to ACLs.
2679 37. There was a bug in the logic of the code that waits for the clock to tick
2680 in the case where the clock went backwards by a substantial amount such
2681 that the microsecond fraction of "now" was more than the microsecond
2682 fraction of "then" (but the whole seconds number was less).
2684 38. Added support for the libradius Radius client library this is found on
2685 FreeBSD (previously only the radiusclient library was supported).
2691 1. When certain lookups returned multiple values in the form name=value, the
2692 quoting of the values was not always being done properly. Specifically:
2693 (a) If the value started with a double quote, but contained no whitespace,
2695 (b) If the value contained whitespace other than a space character (i.e.
2696 tabs or newlines or carriage returns) it was not quoted.
2697 This fix has been applied to the mysql and pgsql lookups by writing a
2698 separate quoting function and calling it from the lookup code. The fix
2699 should probably also be applied to nisplus, ibase and oracle lookups, but
2700 since I cannot test any of those, I have not disturbed their existing code.
2702 2. A hit in the callout cache for a specific address caused a log line with no
2703 reason for rejecting RCPT. Now it says "Previous (cached) callout
2704 verification failure".
2706 3. There was an off-by-one bug in the queryprogram router. An over-long
2707 return line was truncated at 256 instead of 255 characters, thereby
2708 overflowing its buffer with the terminating zero. As well as fixing this, I
2709 have increased the buffer size to 1024 (and made a note to document this).
2711 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2712 when Exim is waiting for an SMTP response from a remote server, Exim
2713 restarts its select() call on the socket, thereby resetting its timeout.
2714 This is not a problem when such interrupts are rare. Somebody set up a cron
2715 job to run exiwhat every 2 minutes, which is less than the normal select()
2716 timeout (5 or 10 minutes). This meant that the select() timeout never
2717 kicked in because it was always reset. I have fixed this by comparing the
2718 time when an interrupt arrives with the time at the start of the first call
2719 to select(). If more time than the timeout has elapsed, the interrupt is
2720 treated as a timeout.
2722 5. Some internal re-factoring in preparation for the addition of Sieve
2723 extensions (by MH). In particular, the "personal" test is moved to a
2724 separate function, and given an option for scanning Cc: and Bcc: (which is
2725 not set for Exim filters).
2727 6. When Exim created an email address using the login of the caller as the
2728 local part (e.g. when creating a From: or Sender: header line), it was not
2729 quoting the local part when it contained special characters such as @.
2731 7. Installed new OpenBSD configuration files.
2733 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2734 try to make them clearer.
2736 9. Callout options, other than the timeout value, were being ignored when
2737 verifying sender addresses in header lines. For example, when using
2739 verify = header_sender/callout=no_cache
2741 the cache was (incorrectly) being used.
2743 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2744 only the Cygwin environment.
2746 11. When return_path_on_delivery was set as a log selector, if different remote
2747 addresses in the same message used different return paths and parallel
2748 remote delivery occurred, the wrong values would sometimes be logged.
2749 (Whenever a remote delivery process finished, the return path value from
2750 the most recently started remote delivery process was logged.)
2752 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2753 lines when AUTH and/or TLS are in use. This is the "received protocol"
2754 field. Exim used to use "asmtp" for authenticated SMTP, without any
2755 indication (in the protocol name) for TLS use. Now it follows the RFC and
2756 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2757 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2758 names appear in log lines as well as in Received: header lines.
2760 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2761 extensions, and comparison tests, and to fix some bugs.
2763 14. Changes to the "personal" filter test:
2765 (1) The test was buggy in that it was just doing the equivalent of
2766 "contains" tests on header lines. For example, if a user's address was
2767 anne@some.where, the "personal" test would incorrectly be true for
2769 To: susanne@some.where
2771 This test is now done by extracting each address from the header in turn,
2772 and checking the entire address. Other tests that are part of "personal"
2773 are now done using regular expressions (for example, to check local parts
2774 of addresses in From: header lines).
2776 (2) The list of non-personal local parts in From: addresses has been
2777 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2778 taken from the Sieve specification recommendations.
2780 (3) If the message contains any header line starting with "List-" it is
2781 treated as non-personal.
2783 (4) The test for "circular" in the Subject: header line has been removed
2784 because it now seems ill-conceived.
2786 15. Minor typos in src/EDITME comments corrected.
2788 16. Installed latest exipick from John Jetmore.
2790 17. If headers_add on a router specified a text string that was too long for
2791 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2792 of string_sprintf() is now avoided.
2794 18. $message_body_size was not set (it was always zero) when running the DATA
2795 ACL and the local_scan() function.
2797 19. For the "mail" command in an Exim filter, no default was being set for
2798 the once_repeat time, causing a random time value to be used if "once" was
2799 specified. (If the value happened to be <= 0, no repeat happened.) The
2800 default is now 0s, meaning "never repeat". The "vacation" command was OK
2801 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2802 (I found it when inspecting the code).
2804 20. There is now an overall timeout for performing a callout verification. It
2805 defaults to 4 times the callout timeout, which applies to individual SMTP
2806 commands during the callout. The overall timeout applies when there is more
2807 than one host that can be tried. The timeout is checked before trying the
2808 next host. This prevents very long delays if there are a large number of
2809 hosts and all are timing out (e.g. when the network connections are timing
2810 out). The value of the overall timeout can be changed by specifying an
2811 additional sub-option for "callout", called "maxwait". For example:
2813 verify = sender/callout=5s,maxwait=20s
2815 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2816 to the end before writing, but this should make it even safer).
2818 22. Exim was forgetting that it had advertised PIPELINING for the second and
2819 subsequent messages on an SMTP connection. It was also not resetting its
2820 memory on STARTTLS and an internal HELO.
2822 23. When Exim logs an SMTP synchronization error within a session, it now
2823 records whether PIPELINING has been advertised or not.
2825 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2826 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2827 rather than long int.
2829 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2831 26. Added the never_mail option to autoreply.
2837 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2838 crash if the getsockname() call failed; this can happen if a connection is
2839 closed very soon after it is established. The problem was simply in the
2840 order in which certain operations were done, causing Exim to try to write
2841 to the SMTP stream before it had set up the file descriptor. The bug has
2842 been fixed by making things happen in the correct order.
2848 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2849 before the connection was closed, thus losing the rejection response.
2851 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2852 some early Solaris releases, but causes trouble in current releases where
2853 socklen_t is defined.
2855 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2858 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2861 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2862 that releasing the top of it at the end releases what was used for sub-
2863 expansions (unless the block got too big). However, discard this block if
2864 the first thing is a variable or header, so that we can use its block when
2865 it is dynamic (useful for very large $message_headers, for example).
2867 6. Lookups now cache *every* query, not just the most recent. A new, separate
2868 store pool is used for this. It can be recovered when all lookup caches are
2869 flushed. Lookups now release memory at the end of their result strings.
2870 This has involved some general refactoring of the lookup sources.
2872 7. Some code has been added to the store_xxx() functions to reduce the amount
2873 of flapping under certain conditions.
2875 8. log_incoming_interface used to affect only the <= reception log lines. Now
2876 it causes the local interface and port to be added to several more SMTP log
2877 lines, for example "SMTP connection from", and rejection lines.
2879 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2881 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2883 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2884 could be overwritten at the end of the current message (or the start of a
2885 new message if it was set in a HELO ACL). The value is now preserved for
2886 the duration of the SMTP connection.
2888 12. If a transport had a headers_rewrite setting, and a matching header line
2889 contained an unqualified address, that address was qualified, even if it
2890 did not match any rewriting rules. The underlying bug was that the values
2891 of the flags that permit the existence of unqualified sender and recipient
2892 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2893 non-local messages, and by -bnq for local messages) were not being
2894 preserved with the message after it was received.
2896 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2897 "next input=" as part of the text comprising the host identity instead of
2898 the correct text. The code was using the same buffer for two different
2899 strings. However, depending on which order the printing function evaluated
2900 its arguments, the bug did not always show up. Under Linux, for example, my
2901 test suite worked just fine.
2903 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2904 doesn't work with some older versions of Perl. It has been changed to "my",
2905 which in any case is probably the better facility to use.
2907 15. A really picky compiler found some instances of statements for creating
2908 error messages that either had too many or two few arguments for the format
2911 16. The size of the buffer for calls to the DNS resolver has been increased
2912 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2913 for addresses that have a lot of PTR records. This alleviates a problem; it
2914 does not fully solve it.
2916 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2917 buffer now truncates the list and logs the incident, which is the same
2918 action as happens when Exim is looking up a host name and its aliases.
2919 Previously in this situation something unpredictable would happen;
2920 sometimes it was "internal error: store_reset failed".
2922 18. If a server dropped the connection unexpectedly when an Exim client was
2923 using GnuTLS and trying to read a response, the client delivery process
2924 crashed while trying to generate an error log message.
2926 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2927 string, for example:
2929 warn message = H1: something\nH2: something
2931 the text was added as a single header line from Exim's point of view
2932 though it ended up OK in the delivered message. However, searching for the
2933 second and subsequent header lines using $h_h2: did not work. This has been
2934 fixed. Similarly, if a system filter added multiple headers in this way,
2935 the routers could not see them.
2937 20. Expanded the error message when iplsearch is called with an invalid key to
2938 suggest using net-iplsearch in a host list.
2940 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2941 ACLs are no longer actually imposed (and a message to that effect is
2944 22. If a "gecos" field in a passwd entry contained escaped characters, in
2945 particular, if it contained a \" sequence, Exim got it wrong when building
2946 a From: or a Sender: header from that name. A second bug also caused
2947 incorrect handling when an unquoted " was present following a character
2948 that needed quoting.
2950 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2951 was not being matched caselessly.
2953 24. Arranged for all hyphens in the exim.8 source to be escaped with
2956 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2957 properly. Recipient callout cache records were still being keyed to include
2958 the sender, even when use_sender was set false. This led to far more
2959 callouts that were necessary. The sender is no longer included in the key
2960 when use_sender is false.
2962 26. Added "control = submission" modifier to ACLs.
2964 27. Added the ${base62d: operator to decode base 62 numbers.
2966 28. dnsdb lookups can now access SRV records.
2968 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2969 the configuration file.
2971 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2972 (-v) mode. This makes the output for a verbose queue run more intelligible.
2974 31. Added a use_postmaster feature to recipient callouts.
2976 32. Added the $body_zerocount variable, containing the number of binary zero
2977 bytes in the message body.
2979 33. The time of last modification of the "new" subdirectory is now used as the
2980 "mailbox time last read" when there is a quota error for a maildir
2983 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2985 35. Added +ignore_unknown as a special item in host lists.
2987 36. Code for decoding IPv6 addresses in host lists is now included, even if
2988 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2989 address was recognized as an IP address, but was then not correctly decoded
2990 into binary, causing unexpected and incorrect effects when compared with
2997 1. Very minor rewording of debugging text in manualroute to say "list of
2998 hosts" instead of "hostlist".
3000 2. If verify=header_syntax was set, and a header line with an unqualified
3001 address (no domain) and a large number of spaces between the end of the
3002 name and the colon was received, the reception process suffered a buffer
3003 overflow, and (when I tested it) crashed. This was caused by some obsolete
3004 code that should have been removed. The fix is to remove it!
3006 3. When running in the test harness, delay a bit after writing a bounce
3007 message to get a bit more predictability in the log output.
3009 4. Added a call to search_tidyup() just before forking a reception process. In
3010 theory, someone could use a lookup in the expansion of smtp_accept_max_
3011 per_host which, without the tidyup, could leave open a database connection.
3013 5. Added the variables $recipient_data and $sender_data which get set from a
3014 lookup success in an ACL "recipients" or "senders" condition, or a router
3015 "senders" option, similar to $domain_data and $local_part_data.
3017 6. Moved the writing of debug_print from before to after the "senders" test
3020 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3021 problems for message scanning, either using a data ACL, or using
3022 local_scan() because the Received: header was not generated till after they
3023 were called (in order to set the time as the time of reception completion).
3024 I have revised the way this works. The header is now generated after the
3025 body is received, but before the ACL or local_scan() are called. After they
3026 are run, the timestamp in the header is updated.
3032 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3033 before starting a queue runner without re-exec. This happened only when
3034 deliver_drop_privilege was set or when the Exim user was set to root. The
3035 effect of the bug was that timeouts during subsequent deliveries caused
3036 crashes instead of being properly handled. The handler is now left at its
3037 default (and expected) setting.
3039 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3040 message, again when deliver_drop_privilege is set or Exim is run as root.
3041 The bug described in (1) was not present in this case, but the tidying up
3042 of the other signals was missing. I have made the two cases consistent.
3044 3. The ignore_target_hosts setting on a manualroute router was being ignored
3045 for hosts that were looked up using the /MX notation.
3047 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3050 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3051 operated on the sender address. After changing the $sender_address to <>
3052 for the sender address verify, Exim was re-instated it as the original
3053 (before rewriting) address, but remembering that it had rewritten it, so it
3054 wasn't rewriting it again. This bug also had the effect of breaking the
3055 sender address verification caching when the sender address was rewritten.
3057 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3058 This has been changed so that if the ip literal address matches
3059 ignore_target_hosts, the router declines.
3061 7. Added expansion conditions match_domain, match_address, and match_local_
3062 part (NOT match_host).
3064 8. The placeholder for the Received: header didn't have a length field set.
3066 9. Added code to Exim itself and to exim_lock to test for a specific race
3067 condition that could lead to file corruption when using MBX delivery. The
3068 issue is with the lockfile that is created in /tmp. If this file is removed
3069 after a process has opened it but before that process has acquired a lock,
3070 there is the potential for a second process to recreate the file and also
3071 acquire a lock. This could lead to two Exim processes writing to the file
3072 at the same time. The added code performs the same test as UW imapd; it
3073 checks after acquiring the lock that its file descriptor still refers to
3074 the same named file.
3076 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3077 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3080 11. Added the smtp_active_hostname option. If used, this will typically be made
3081 to depend on the incoming interface address. Because $interface_address is
3082 not set up until the daemon has forked a reception process, error responses
3083 that can happen earlier (such as "too many connections") no longer contain
3086 12. If an expansion in a condition on a "warn" statement fails because a lookup
3087 defers, the "warn" statement is abandoned, and the next ACL statement is
3088 processed. Previously this caused the whole ACL to be aborted.
3090 13. Added the iplsearch lookup type.
3092 14. Added ident_timeout as a log selector.
3094 15. Added tls_certificate_verified as a log selector.
3096 16. Added a global option tls_require_ciphers (compare the smtp transport
3097 option of the same name). This controls incoming TLS connections.
3099 17. I finally figured out how to make tls_require_ciphers do a similar thing
3100 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3101 before starting the TLS session.
3103 18. Tabs are now shown as \t in -bP output.
3105 19. If the log selector return_path_on_delivery was set, Exim crashed when
3106 bouncing a message because it had too many Received: header lines.
3108 20. If two routers both had headers_remove settings, and the first one included
3109 a superfluous trailing colon, the final name in the first list and the
3110 first name in the second list were incorrectly joined into one item (with a
3111 colon in the middle).
3117 1. Added -C and -D options to the exinext utility, mainly to make it easier
3118 to include in the automated testing, but these could be helpful when
3119 multiple configurations are in use.
3121 2. The exinext utility was not formatting the output nicely when there was
3122 an alternate port involved in the retry record key, nor when there was a
3123 message id as well (for retries that were specific to a specific message
3124 and a specific host). It was also confused by IPv6 addresses, because of
3125 the additional colons they contain. I have fixed the IPv4 problem, and
3126 patched it up to do a reasonable job for IPv6.
3128 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3129 delivery, the log line now contains "pipelined" if PIPELINING was used.
3131 4. An SMTP transport process used to panic and die if the bind() call to set
3132 an explicit outgoing interface failed. This has been changed; it is now
3133 treated in the same way as a connect() failure.
3135 5. A reference to $sender_host_name in the part of a conditional expansion
3136 that was being skipped was still causing a DNS lookup. This no longer
3139 6. The def: expansion condition was not recognizing references to header lines
3140 that used bh_ and bheader_.
3142 7. Added the _cache feature to named lists.
3144 8. The code for checking quota_filecount in the appendfile transport was
3145 allowing one more file than it should have been.
3147 9. For compatibility with Sendmail, the command line option
3155 and sets the incoming protocol and host name (for trusted callers). The
3156 host name and its colon can be omitted when only the protocol is to be set.
3157 Note the Exim already has two private options, -pd and -ps, that refer to
3158 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3159 "s", but I don't think that's a major issue.
3161 10. A number of refactoring changes to the code, none of which should affect
3164 (a) The number of logging options was getting close to filling up the
3165 32-bit word that was used as a bit map. I have split them into two classes:
3166 those that are passed in the argument to log_write(), and those that are
3167 only ever tested independently outside of that function. These are now in
3168 separate 32-bit words, so there is plenty of room for expansion again.
3169 There is no change in the user interface or the logging behaviour.
3171 (b) When building, for example, log lines, the code previously used a
3172 macro that called string_cat() twice, in order to add two strings. This is
3173 not really sufficiently general. Furthermore, there was one instance where
3174 it was actually wrong because one of the argument was used twice, and in
3175 one call a function was used. (As it happened, calling the function twice
3176 did not affect the overall behaviour.) The macro has been replaced by a
3177 function that can join an arbitrary number of extra strings onto a growing
3180 (c) The code for expansion conditions now uses a table and a binary chop
3181 instead of a serial search (which was left over from when there were very
3182 few conditions). Also, it now recognizes conditions like "pam" even when
3183 the relevant support is not compiled in: a suitably worded error message is
3184 given if an attempt is made to use such a condition.
3186 11. Added ${time_interval:xxxxx}.
3188 12. A bug was causing one of the ddress fields not to be passed back correctly
3189 from remote delivery subprocesses. The field in question was not being
3190 subsequently used, so this caused to problems in practice.
3192 13. Added new log selectors queue_time and deliver_time.
3194 14. Might have fixed a bug in maildirsizefile handling that threw up
3195 "unexpected character" debug warnings, and recalculated the data
3196 unnecessarily. In any case, I expanded the warning message to give more
3199 15. Added the message "Restricted characters in address" to the statements in
3200 the default ACL that block characters like @ and % in local parts.
3202 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3203 Three changes have been made:
3205 (a) There was a serious bug; a negative response to MAIL caused the whole
3206 recipient domain to be cached as invalid, thereby blocking all messages
3207 to all local parts at the same domain, from all senders. This bug has
3208 been fixed. The domain is no longer cached after a negative response to
3209 MAIL if the sender used is not empty.
3211 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3214 (c) A new callout option, "use_sender" has been added for people who want
3215 the modified behaviour.
3221 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3224 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3225 indeed breaks things for older releases.
3227 3. Added additional logging to the case where there is a problem reading data
3228 from a filter that is running in a subprocess using a pipe, in order to
3229 try to track down a specific problem.
3231 4. Testing facility fudge: when running in the test harness and attempting
3232 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3233 getting "No route to host". Convert this to a timeout.
3235 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3238 6. Some OS don't have socklen_t but use size_t instead. This affects the
3239 fifth argument of getsockopt() amongst other things. This is now
3240 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3241 can be set for individual OS. I have set it for SunOS5, OSF1, and
3242 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3243 some earlier ones do not.
3245 7. Change 4.30/15 was not doing the test caselessly.
3247 8. The standard form for an IPv6 address literal was being rejected by address
3248 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3249 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3250 this, as well as the form without the "IPv6" on the front (but only when
3251 address literals are enabled, of course).
3253 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3255 10. Exim crashed if a message with an empty sender address specified by -f
3256 encountered a router with an errors_to setting. This could be provoked only
3257 by a command such as
3261 where an empty string was supplied; "<>" did not hit this bug.
3263 11. Installed PCRE release 4.5.
3265 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3266 remained set. It is now erased.
3268 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3269 times from message ids (which are base 36 rather than the normal 62).
3271 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3272 were being counted as actual protocol errors, and logged if the log
3273 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3274 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3275 and DATA following a set of rejected RCPTs do not count as protocol errors.
3276 In other words, Exim assumes they were pipelined, though this may not
3277 actually be the case. Of course, in all cases the client gets an
3278 appropriate error code.
3280 15. If a lookup fails in an ACL condition, a message about the failure may
3281 be available; it is used if testing the ACL cannot continue, because most
3282 such messages specify what the cause of the deferral is. However, some
3283 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3284 that caused an old message to be retained and used if a later statement
3285 caused a defer, replacing the real cause of the deferral.
3287 16. If an IP address had so many PTR records that the DNS lookup buffer
3288 was not large enough to hold them, Exim could crash while trying to process
3289 the truncated data. It now detects and logs this case.
3291 17. Further to 4.21/58, another change has been made: if (and only if) the
3292 first line of a message (the first header line) ends with CRLF, a bare LF
3293 in a subsequent header line has a space inserted after it, so as not to
3294 terminate the header.
3296 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3297 unnecessarily, used atoi() instead of strtol(), and didn't check the
3298 termination when getting file sizes from file names by regex.
3300 19. Completely re-implemented the support for maildirsize files, in the light
3301 of a number of problems with the previous contributed implementation
3302 (4.30/29). In particular:
3304 . If the quota is zero, the maildirsize file is maintained, but no quota is
3307 . If the maildir directory does not exist, it is created before any attempt
3308 to write a maildirsize file.
3310 . The quota value in the file is just a cache; if the quota is changed in
3311 the transport, the new value overrides.
3313 . A regular expression is available for excluding directories from the
3316 20. The autoreply transport checks the characters in options that define the
3317 message's headers; it allows continued headers, but it was checking with
3318 isspace() after an embedded newline instead of explicitly looking for a
3321 21. If all the "regular" hosts to which an address was routed had passed their
3322 expiry times, and had not reached their retry times, the address was
3323 bounced, even if fallback hosts were defined. Now Exim should go on to try
3326 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3327 equivalent code in the SMTP transport. Some hosts send humungous responses
3328 to HELO/EHLO, more than 1024 it seems.
3330 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3331 gives compiler warnings in some environments. I've now done it "properly",
3334 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3335 (because of problems with the built-in one) was declared to return uschar *
3336 instead of char *, causing compiler failure.
3338 25. Fixed a file descriptor leak when processing alias/forward files.
3340 26. Fixed a minor format string issue in dbfn.c.
3342 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3344 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3345 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3348 29. When a sender address is verified, it is cached, to save repeating the test
3349 when there is more than one recipient in a message. However, when the
3350 verification involves a callout, it is possible for different callout
3351 options to be set for different recipients. It is too complicated to keep
3352 track of this in the cache, so now Exim always runs a verification when a
3353 callout is required, relying on the callout cache for the optimization.
3354 The overhead is duplication of the address routing, but this should not be
3357 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3358 to be verified with callout=postmaster, and the main callout worked but the
3359 postmaster check failed, the verification correctly failed. However, if a
3360 subsequent RCPT command asked for sender verification *without* the
3361 postmaster check, incorrect caching caused this verification also to fail,
3364 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3365 it was not caching the DNS options (qualify_single, search_parents) that
3366 were used when the lookup failed. A subsequent lookup with different
3367 options therefore always gave the same answer, though there were cases
3368 where it should not have. (Example: a "domains = !$mx_any" option on a
3369 dnslookup router: the "domains" option is always processed without any
3370 widening, but the router might have qualify_single set.) Now Exim uses the
3371 cached value only when the same options are set.
3373 32. Added John Jetmore's "exipick" utility to the distribution.
3375 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3376 than a timeout (e.g. a certificate is required, and is not provided), an
3377 Exim server now closes the connection immediately. Previously it waited for
3378 the client to close - but if the client is SSL, it seems that they each
3379 wait for each other, leading to a delay before one of them times out.
3381 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3382 maintained 0.8.x compatibility because I don't think many are using it, and
3383 it is clearly obsolete.
3385 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3388 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3389 client certificate was expired. A simple patch fixes this, though I don't
3390 understand the full logic of why the verify callback is called multiple
3393 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3394 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3395 which causes problems with some clients (such as the Certicom SSL Plus
3396 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3397 disables the coutermeasure allowing Eudora to connect."
3399 38. Exim was not checking that a write() to a log file succeeded. This could
3400 lead to Bad Things if a log got too big, in particular if it hit a file
3401 size limit. Exim now panics and dies if it cannot write to a log file, just
3402 as it does if it cannot open a log file.
3404 39. Modified OS/Makefile-Linux so that it now contains
3406 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3408 The two -D definitions ensure that Exim is compiled with large file
3409 support, which makes it possible to handle log files that are bigger than
3412 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3413 instance) a domain was checked against a named list that involved a lookup,
3414 causing $domain_data to be set, then another domain was checked against the
3415 same list, then the first domain was re-checked, the value of $domain_data
3416 after the final check could be wrong. In particular, if the second check
3417 failed, it could be set empty. This bug probably also applied to
3420 41. The strip_trailing_dot option was not being applied to the address given
3421 with the -f command-line option.
3423 42. The code for reading a message's header from the spool was incrementing
3424 $received_count, but never initializing it. This meant that the value was
3425 incorrect (doubled) while delivering a message in the same process in which
3426 it was received. In the most common configuration of Exim, this never
3427 happens - a fresh exec is done - but it can happen when
3428 deliver_drop_privilege is set.
3430 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3431 it now includes up to 150 characters of the unexpected data in the log
3434 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3435 and building data strings. The size of both of these buffers was 10 000
3436 bytes - far larger than anybody would *ever* want, thought I. Needless to
3437 say, somebody hit the limit. I have increased the maximum line length to
3438 20 000 and the maximum data length of concatenated lines to 100 000. I have
3439 also fixed two bugs, because there was no checking on these buffers. Tsk,
3440 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3441 buffer is too small.
3443 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3444 lsearch lookups. Now it does.
3446 46. When parsing a route_list item in a manualroute router, a fixed-length
3447 buffer was used for the list of hosts. I made this 1024 bytes long,
3448 thinking that nobody would ever have a list of hosts that long. Wrong.
3449 Somebody had a whole pile of complicated expansion conditions, and the
3450 string was silently truncated, leading to an expansion error. It turns out
3451 that it is easier to change to an unlimited length (owing to other changes
3452 that have happened since this code was originally written) than to build
3453 structure for giving a limitation error. The length of the item that
3454 expands into the list of hosts is now unlimited.
3456 47. The lsearch lookup could not handle data where the length of text line was
3457 more than 4095 characters. Such lines were truncated, leading to shortened
3458 data being returned. It should now handle lines of any length.
3460 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3461 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3464 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3465 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3466 no longer happens in comment lines. A list of replacements is now placed
3467 at the head of all of the source files, except those whose only change is
3468 to replace PERL_COMMAND in the very first #! line.
3470 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3471 messages on the queue, with a bottom-up merge sort, using code contributed
3472 by Michael Haardt. This should make operations like -bp somewhat faster on
3473 large queues. It won't affect queue runners, except when queue_run_in_order
3476 51. Installed eximstats 1.31 in the distribution.
3478 52. Added support for SRV lookups to the dnslookup router.
3480 53. If an ACL referred to $message_body or $message_body_end, the value was not
3481 reset for any messages that followed in the same SMTP session.
3483 54. The store-handling optimization for building very long strings was not
3484 differentiating between the different store pools. I don't think this
3485 actually made any difference in practice, but I've tidied it.
3487 55. While running the routers to verify a sender address, $sender_address
3488 was still set to the sender address. This is wrong, because when routing to
3489 send a bounce to the sender, it would be empty. Therefore, I have changed
3490 it so that, while verifying a sender address, $sender_address is set to <>.
3491 (There is no change to what happens when verifying a recipient address.)
3493 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3494 target A or AAAA records (if not already returned) without resetting the
3495 qualify_single or search_parents options of the DNS resolver. These are
3496 inappropriate in this case because the targets of MX and SRV records must
3497 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3498 target that, when qualified, matched something in the local domain. These
3499 two options are now turned off when doing these lookups.
3501 57. It seems that at least some releases of Reiserfs (which does not have the
3502 concept of a fixed number of inodes) returns zero and not -1 for the
3503 number of available inodes. This interacted badly with check_spool_inodes,
3504 which assumed that -1 was the "no such thing" setting. What I have done is
3505 to check that the total number of inodes is greater than zero before doing
3506 the test of how many are available.
3508 58. When a "warn" ACL statement has a log_message modifier, the message is
3509 remembered, and not repeated. This is to avoid a lot of repetition when a
3510 message has many recipients that cause the same warning to be written.
3511 Howewer, Exim was preserving the list of already written lines for an
3512 entire SMTP session, which doesn't seem right. The memory is now reset if a
3513 new message is started.
3515 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3516 debugging output unless log_rewrite was also set.
3518 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3519 of (int)(handle) when we know that handle contains (void *)(-1).
3521 61. The Exim daemon panic-logs an error return when it closes the incoming
3522 connection. However "connection reset by peer" seems to be common, and
3523 isn't really an error worthy of noting specially, so that particular error
3526 62. When Exim is trying to find all the local interfaces, it used to panic and
3527 die if the ioctl to get the interface flags failed. However, it seems that
3528 on at least one OS (Solaris 9) it is possible to have an interface that is
3529 included in the list of interfaces, but for which you get a failure error
3530 for this call. This happens when the interface is not "plumbed" into a
3531 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3532 failure of the "get flags" call assumes that the interface is down.
3534 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3535 makes life easier for people who are doing arithmetic on fields extracted
3536 from dates, where you often get leading zeros that should not be
3537 interpreted as octal.
3539 64. Added qualify_domain to the redirect router, to override the global
3542 65. If a pathologically long header line contained very many addresses (the
3543 report of this problem mentioned 10 000) and each of them was rewritten,
3544 Exim could use up a very large amount of memory. (It kept on making new
3545 copies of the header line as it rewrote, and never released the old ones.)
3546 At the expense of a bit more processing, the header rewriting function has
3547 been changed so that it no longer eats memory in this way.
3549 66. The generation of the Received: header has been moved from the time that a
3550 message starts to be received, to the time that it finishes. The timestamp
3551 in the Received: header should now be very close to that of the <= log
3552 line. There are two side-effects of this change:
3554 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3555 logged header lines no longer include the local Received: line, because
3556 it has not yet been created. The same applies to a copy of the message
3557 that is returned to a non-SMTP sender when a message is rejected.
3559 (b) When a filter file is tested using -bf, no additional Received: header
3560 is added to the test message. After some thought, I decided that this
3563 This change does not affect the value of $received_for. It is still set
3564 after address rewriting, but before local_scan() is called.
3566 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3568 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3569 gave an unhelpful panic error message, and a defer error. I have managed to
3570 change this behaviour so that it now rejects any supplied certificate,
3571 which seems right, as the list of acceptable certificates is empty.
3573 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3574 gave an unhelpful defer error. I have not managed to make this reject any
3575 supplied certificates, but the error message it gives is "no certificate
3576 supplied", which is not helpful.
3578 70. exigrep's output now also includes lines that are not associated with any
3579 message, but which match the given pattern. Implemented by a patch from
3580 Martin Sluka, which also tidied up the Perl a bit.
3582 71. Recipient callout verification, like sender verification, was using <> in
3583 the MAIL FROM command. This isn't really the right thing, since the actual
3584 sender may affect whether the remote host accepts the recipient or not. I
3585 have changed it to use the actual sender in the callout; this means that
3586 the cache record is now keyed on a recipient/sender pair, not just the
3587 recipient address. There doesn't seem to be a real danger of callout loops,
3588 since a callout by the remote host to check the sender would use <>.
3589 [SEE ABOVE: changed after hitting problems.]
3591 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3592 temporary errors. However, in the case of such a code being given after
3593 the end of a data transmission (i.e. after ".") Exim was failing to write
3594 a retry record for the message. (Yes, there was some broken host that was
3595 actually sending 8xx at this point.)
3597 73. An unknown lookup type in a host list could cause Exim to panic-die when
3598 the list was checked. (An example that provoked this was putting <; in the
3599 middle of a list instead of at the start.) If this happened during a DATA
3600 ACL check, a -D file could be left lying around. This kind of configuration
3601 error no longer causes Exim to die; instead it causes a defer errror. The
3602 incident is still logged to the main and panic logs.
3604 74. Buglet left over from Exim 3 conversion. The message "too many messages
3605 in one connection" was written to the rejectlog but not the mainlog, except
3606 when address rewriting (yes!) was being logged.
3608 75. Added write_rejectlog option.
3610 76. When a system filter was run not as root (that is, when system_filter_user
3611 was set), the values of the $n variables were not being returned to the
3612 main process; thus, they were not subsequently available in the $sn
3615 77. Added +return_path_on_delivery log selector.
3617 78. A connection timeout was being treated differently from recipients deferred
3618 when testing hosts_max_try with a message that was older than the host's
3619 retry timeout. (The host should not be counted, thus allowing all hosts to
3620 be tried at least once before bouncing.) This may have been the cause of an
3621 occasionally reported bug whereby a message would remain on the queue
3622 longer than the retry timeout, but would be bounced if a delivery was
3623 forced. I say "may" because I never totally pinned down the problem;
3624 setting up timeout/retry tests is difficult. See also the next item.
3626 79. The ultimate address timeout was not being applied to errors that involved
3627 a combination of host plus message (for example, a timeout on a MAIL
3628 command). When an address resolved to a number of possible hosts, and they
3629 were not all tried for each delivery (e.g. because of hosts_max_try), a
3630 message could remain on the queue longer than the retry timeout.
3632 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3635 81. Fixed an obscure SMTP outgoing bug which required at least the following
3636 conditions: (a) there was another message waiting for the same server;
3637 (b) the server returned 5xx to all RCPT commands in the first message so
3638 that the message was not completed; (c) the server dropped the connection
3639 or gave a negative response to the RSET that Exim sends to abort the
3640 transaction. The observed case was a dropped connection after DATA that had
3641 been sent in pipelining mode. That is, the server had advertised PIPELINING
3642 but was not implementing it correctly. The effect of the bug was incorrect
3643 behaviour, such as trying another host, and this could lead to a crash.
3649 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3650 and daemon.c were passed as pointers to ints; they should have been
3651 pointers to socklen_t variables (which are typically unsigned ints).
3653 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3656 3. Fixed a really odd bug that affected only the testing scheme; patching a
3657 certain fixed string in the binary changed the value of another string that
3658 happened to be identical to the end of the original first string.
3660 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3661 name", it returns that address as the IP address. On some operating
3662 systems (e.g. Solaris), it also passes back the IP address string as the
3663 "host name". However, on others (e.g. Linux), it passes back an empty
3664 string. Exim wasn't checking for this, and was changing the host name to an
3665 empty string, assuming it had been canonicized.
3667 5. Although rare, it is permitted to have more than one PTR record for a given
3668 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3669 all the names associated with an address, because they do in Solaris.
3670 However, it seems that they do not in Linux for data that comes from the
3671 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3672 I found this out when I moved to a new Linux workstation and tried to run
3673 the Exim test suite.
3675 To get round this problem I have changed the code so that it now does its
3676 own call to the DNS to look up PTR records when searching for a host name.
3677 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3678 addresses that are only in /etc/hosts are still found.
3680 This behaviour is, however, controlled by an option called host_lookup_
3681 order, which defaults to "bydns:byaddr". If people want to use the other
3682 order, or indeed, just use one or the other means of lookup, they can
3683 specify it in this variable.
3685 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3686 some operating systems, this comes back from gethostbyaddr() as an empty
3687 string, and this is what Exim used to test for. However, it seems that in
3688 other systems, "." is yielded. Exim now tests for this case too.
3690 7. The values of check_spool_space and check_log_space are now held internally
3691 as a number of kilobytes instead of an absolute number of bytes. If a
3692 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3693 kilobyte. This means that much larger values can be stored.
3695 8. Exim monitor: an attempt to get the action menu when not actually pointing
3696 at a message produces an empty menu entitled "No message selected". This
3697 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3698 no entries in it ("Shell widget menu has zero width and/or height"). So I
3699 have added a single, blank menu entry in this case.
3701 9. Added ${quote_local_part.
3703 10. MIME decoding is now applied to the contents of Subject: header lines when
3706 11. Now that a reference to $sender_host_address automatically causes a reverse
3707 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3708 host lookup before query-style lookups in lists that might use this
3709 variable. This has therefore been abolished, and the "net-" prefix is no
3710 longer necessary for query-style lookups.
3712 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3713 have been a typo for LFLAGS, so it has been changed.
3715 13. The install script calls Exim with "-C /dev/null" in order to find the
3716 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3717 to be output. Howeve, since Exim outputs its version number before the
3718 error, it didn't break the script. It just looked ugly. I fixed this by
3719 always allowing "-C /dev/null" if the caller is root.
3721 14. Ignore overlarge ACL variable number when reading spool file - insurance
3722 against a later release with more variables having written the file.
3724 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3725 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3726 this, as well as the form without the "IPv6" on the front.
3728 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3729 OS/Makefile-Darwin file.
3731 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3732 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3734 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3735 variable. However, some OS use ints for inodes. I've added cast to long int
3736 to get rid of the compiler warning.
3738 19. I had forgotten to lock out "/../" in configuration file names when
3739 ALT_CONFIG_PREFIX was set.
3741 20. Routers used for verification do not need to specify transports. However,
3742 if such a router generated a host list, and callout was configured, Exim
3743 crashed, because it could not find a port number from the (non-existent)
3744 transport. It now assumes port 25 in this circumstance.
3746 21. Added the -t option to exigrep.
3748 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3749 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3750 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3751 them is set, LOOKUP_LSEARCH is forced.
3753 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3755 24. Added sender and host information to the "rejected by local_scan()" log
3756 line; previously there was no indication of these.
3758 25. Added .include_if_exists.
3760 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3761 Linux. It turns out that not all file systems support this. Apparently some
3762 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3763 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3764 not supported on the file descriptor, is now ignored when Exim is trying to
3765 sync a directory. This applies only to Linux.
3767 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3769 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3770 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3773 29. Added support for maildirsize files from supplied patch (modified a bit).
3775 30. The use of :fail: followed by an empty string could lead Exim to respond to
3776 sender verification failures with (e.g.):
3778 550 Verification failed for <xxx>
3779 550 Sender verify failed
3781 where the first response line was missing the '-' that indicates it is not
3782 the final line of the response.
3784 31. The loop for finding the name of the user that called Exim had a hardwired
3785 limit of 10; it now uses the value of finduser_retries, which is used for
3786 all other user lookups.
3788 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3791 33. Exim was neglecting to zero errno before one call of strtol() when
3792 expanding a string and expecting an integer value. On some systems this
3793 resulted in spurious "integer overflow" errors. Also, it was casting the
3794 result into an int without checking.
3796 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3797 did not work. The code looks as if it has *never* worked, though it appears
3798 to have been documented since at least releast 1.62. I have made it work.
3800 35. The "timeout_DNS" error in retry rules, also documented since at least
3801 1.62, also never worked. As it isn't clear exactly what this means, and
3802 clearly it isn't a major issue, I have abolished the feature by treating it
3803 as "timeout", and writing a warning to the main and panic logs.
3805 36. The display of retry rules for -brt wasn't always showing the error code
3808 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3809 timeout_connect_A, timeout_connect_MX.
3811 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3812 to the empty sender.
3814 39. The daemon was not analysing the content of -oX till after it had closed
3815 stderr and disconnected from the controlling terminal. This meant that any
3816 syntax errors were only noted on the panic log, and the return code from
3817 the command was 0. By re-arranging the code a little, I've made the
3818 decoding happen first, so such errors now appear on stderr, and the return
3819 code is 1. However, the actual setting up of the sockets still happens in
3820 the disconnected process, so errors there are still only recorded on the
3823 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3824 connections (as happens on some IP stacks) was logged at start up time as
3825 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3826 it from "IPv6 and IPv4", which means that two separate sockets are being
3829 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3830 says whether AF_INET or AF_INET6 was passed as an argument.
3832 42. Exiwhat output was messed up when time zones were included in log
3835 43. Exiwhat now gives more information about the daemon's listening ports,
3836 and whether -tls-on-connect was used.
3838 44. The "port" option of the smtp transport is now expanded.
3840 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3841 silently ignored. Now an error message is written to the main and panic
3844 46. There's a new ACL modifier called "logwrite" which writes to a log file
3845 as soon as it is encountered.
3847 47. Added $local_user_uid and $local_user_gid at routing time.
3849 48. Exim crashed when trying to verify a sender address that was being
3852 49. Exim was recognizing only a space character after ".include". It now also
3853 recognizes a tab character.
3855 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3856 extracting the relevant information from the specification. The man page no
3857 longer contains scrambled data for the -d option, and I've added a section
3858 at the front about calling Exim under different names.
3860 51. Added "extra_headers" argument to the "mail" command in filter files.
3862 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3865 53. Installed eximstats 1.29.
3867 54. Added transport_filter_timeout as a generic transport option.
3869 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3870 Cc: header lines. This was required by RFC 822, but it not required by RFC
3873 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3874 incoming messages that did not have them. Now it does so only if the
3875 message originates locally, that is, if there is no associated remote host
3876 address. When Resent- header lines are present, this applies to the Resent-
3877 lines rather than the non-Resent- lines.
3879 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3880 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3882 58. Messages for configuration errors now include the name of the main
3883 configuration file - useful now that there may be more than one file in a
3884 list (.included file names were always shown).
3886 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3887 for those rare installations that do not start the daemon as root or run it
3888 setuid root. I've cut out the call to initgroups() if the daemon is not
3891 60. The Exim user and group can now be bound into the binary as text strings
3892 that are looked up at the start of Exim's processing.
3894 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3896 62. Added $mailstore_basename variable.
3898 63. Installed patch to sieve.c from Michael Haardt.
3900 64. When Exim failed to open the panic log after failing to open the main log,
3901 the original message it was trying to log was written to stderr and debug
3902 output, but if they were not available (the usual case in production), it
3903 was lost. Now it is written to syslog before the two lines that record the
3904 failures to open the logs.
3906 65. Users' Exim filters run in subprocesses under the user's uid. It is
3907 possible for a "deliver" command or an alias in a "personal" command to
3908 provoke an address rewrite. If logging of address rewriting is configured,
3909 this fails because the process is not running as root or exim. There may be
3910 a better way of dealing with this, but for the moment (because 4.30 needs
3911 to be released), I have disabled address rewrite logging when running a
3912 filter in a non-root, non-exim process.
3918 1. The buildconfig auxiliary program wasn't quoting the value set for
3919 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3920 not defined. This bug was masked in 4.22 by the effect that was fixed in
3923 2. Some messages that were rejected after a message id was allocated were
3924 shown as "incomplete" by exigrep. It no longer does this for messages that
3925 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3927 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3928 have allow_domain_literals set, the ID did not get logged in the <= line.
3929 Domain literals are now always recognized in Message-ID: header lines.
3931 4. The first argument for a ${extract expansion item is the key name or field
3932 number. Leading and trailing spaces in this item were not being ignored,
3933 causing some misleading effects.
3935 5. When deliver_drop_privilege was set, single queue runner processes started
3936 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3937 same command in the process it spins off) were not dropping privilege.
3939 6. When the daemon running as "exim" started a queue runner, it always
3940 re-executed Exim in the spun-off process. This is a waste of effort when
3941 deliver_drop_privilege is set. The new process now just calls the
3942 queue-runner function directly.
3948 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3951 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3952 ignored. Though the use of -oP was forcing the writing of a pid file, it
3953 was always written to the default place.
3955 3. If the message "no IP address found for host xxxx" is generated during
3956 incoming verification, it is now followed by identification of the incoming
3957 connection (so you can more easily find what provoked it).
3959 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3961 5. Added some features to "harden" Exim a bit more against certain attacks:
3963 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3964 be put in Local/Makefile. This is like the never_users runtime option,
3965 but it cannot be overridden. The default setting is "root".
3967 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3968 prefix string with which any file named in a -C command line option
3971 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3972 is retained for -C and -D only if the caller of Exim is root. Without
3973 it, the exim user may also use -C and -D and retain privilege.
3975 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3976 command line option is disabled.
3978 6. Macro names set by the -D option must start with an upper case letter, just
3979 like macro names defined in the configuration file.
3981 7. Added "dereference=" facility to LDAP.
3983 8. Two instances of the typo "uknown" in the source files are fixed.
3985 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3986 the Configure-Makefile script screwed up while processing it.
3988 10. Incorporated PCRE 4.4.
3990 11. The SMTP synchronization check was not operating right at the start of an
3991 SMTP session. For example, it could not catch a HELO sent before the client
3992 waited for the greeting. There is now a check for outstanding input at the
3993 point when the greeting is written. Because of the duplex, asynchronous
3994 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3995 way, but not yet received, when the check is performed.
3997 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3998 on TCP/IP sockets, because this apparently causes some broken clients to
4001 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4002 unchanged) from the Cygwin maintainer.
4004 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4005 was supported instead of testing for mbx. Effectively a typo.
4007 15. The spa authenticator server code was not checking that the input it
4008 received was valid base64.
4010 16. The debug output line for the "set" modifier in ACLs was not showing the
4011 name of the variable that was being set.
4013 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4015 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4016 lookup on its own. Something else was needed to trigger the lookup. For
4017 example, a match in host_lookup or the need for a host name in a host list.
4018 Now, if $sender_host_name is referenced and the host name has not yet been
4019 looked up, a lookup is performed. If the lookup fails, the variable remains
4020 empty, and $host_lookup_failed is set to "1".
4022 19. Added "eqi" as a case-independent comparison operator.
4024 20. The saslauthd authentication condition could segfault if neither service
4025 nor realm was specified.
4027 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4028 error message that was logged was misleading, and incoming SMTP
4029 connections were dropped. The message is now more accurate, and temporary
4030 errors are given to SMTP connections.
4032 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4033 (except RSET) with a 421 error, until QUIT is received. However, it was
4034 failing to send a response to QUIT.
4036 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4037 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4038 ACL is now run after the helo_try_verify_hosts code.
4040 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4041 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4042 case-independent, so other case variants are also recognized). Apparently
4043 some systems use these upper case variants.
4045 25. If more than two messages were waiting for the same host, and a transport
4046 filter was specified for the transport, Exim sent two messages over the
4047 same TCP/IP connection, and then failed with "socket operation on non-
4048 socket" when it tried to send the third.
4050 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4052 27. The extern definition of crypt16() in expand.c was not being excluded when
4053 the OS had its own crypt16() function.
4055 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4056 as a preferred synonym for return_size_limit, both as an option and as an
4059 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4061 30. Changed the default configuration ACL to relax the local part checking rule
4062 for addresses that are not in any local domains. For these addresses,
4063 slashes and pipe symbols are allowed within local parts, but the sequence
4064 /../ is explicitly forbidden.
4066 31. SPA server authentication was not clearing the challenge buffer before
4069 32. log_message in a "warn" ACL statement was writing to the reject log as
4070 well as to the main log, which contradicts the documentation and doesn't
4071 seem right (because no rejection is happening). So I have stopped it.
4073 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4074 However, I am unable to do any testing of this.
4076 34. Fixed an infelicity in the appendfile transport. When checking directories
4077 for a mailbox, to see if any needed to be created, it was accidentally
4078 using path names with one or more superfluous leading slashes; tracing
4079 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4081 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4082 contents are added to the log line that is written for every discarded
4083 recipient. (Previously a log_message setting was ignored.)
4085 36. The ${quote: operator now quotes the string if it is empty.
4087 37. The install script runs exim in order to find its version number. If for
4088 some reason other than non-existence or emptiness, which it checks, it
4089 could not run './exim', it was installing it with an empty version number,
4090 i.e. as "exim-". This error state is now caught, and the installation is
4093 38. An argument was missing from the function that creates an error message
4094 when Exim fails to connect to the socket for saslauthd authentication.
4095 This could cause Exim to crash, or give a corrupted message.
4097 39. Added isip, isip4, and isip6 to ${if conditions.
4099 40. The ACL variables $acl_xx are now saved with the message, and can be
4100 accessed later in routers, transports, and filters.
4102 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4103 strings in the file are not string-expanded.
4105 42. If a MAIL command specified a SIZE value that was too large to fit into an
4106 int variable, the check against message_size_limit failed. Such values are
4107 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4108 day this will have to be increased, but I don't think I want to be around
4109 when emails are that large.
4116 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4117 iconv() is not standard in FreeBSD.
4119 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4120 IPv6 enabled. The observed symptom was a segmentation fault on return from
4121 the function os_common_find_running_interfaces() in src/os.c.
4123 3. In the check_special_case() function in daemon.c I had used "errno" as an
4124 argument name, which causes warnings on some systems. This was basically a
4125 typo, since it was named "eno" in the comments!
4127 4. The code that waits for the clock to tick (at a resolution of some fraction
4128 of a second) so as to ensure message-id uniqueness was always waiting for
4129 at least one whole tick, when it could have waited for less. [This is
4130 almost certainly not relevant at current processor speeds, where it is
4131 unlikely to ever wait at all. But we try to future-proof.]
4133 5. The function that sleeps for a time interval that includes fractions of a
4134 second contained a race. It did not block SIGALRM between setting the
4135 timer, and suspending (a couple of lines later). If the interval was short
4136 and the sigsuspend() was delayed until after it had expired, the suspension
4137 never ended. On busy systems this could lead to processes getting stuck for
4140 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4141 process, before it forks any delivery processes. The open lookup caching
4142 mechanism meant that the open file or database connection was passed into
4143 the delivery process. The problem was that delivery processes always tidy
4144 up cached lookup data. This could cause a problem for the next delivery
4145 process started by the queue runner, because the external queue runner
4146 process does not know about the closure. So the next delivery process
4147 still has data in the lookup cache. In the case of a file lookup, there was
4148 no problem because closing a file descriptor in a subprocess doesn't affect
4149 the parent. However, if the lookup was caching a connection to a database,
4150 the connection was closed, and the second delivery process was likely to
4151 see errors such as "PGSQL: query failed: server closed the connection
4152 unexpectedly". The problem has been fixed by closing all cached lookups
4153 in a queue runner before running a delivery process.
4155 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4156 seem to have the "const" qualifier which it has on other OS. I've
4159 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4160 *of the same type* (client or server) with the same public name that an
4161 error should be diagnosed.
4163 9. When Exim looked up a host name for an IP address, but failed to find the
4164 original IP address when looking up the host name (a safety check), it
4165 output the message "<ip address> does not match any IP for NULL", which was
4166 confusing, to say the least. The bug was that the host name should have
4167 appeared instead of "NULL".
4169 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4170 user that is built into the binary, and the -C or -D options is used, root
4171 privilege is dropped before the configuration file is read. In addition,
4172 logging is switched to stderr instead of the normal log files. If the
4173 configuration then re-defines the Exim user, the unprivileged environment
4174 is probably not what is expected, so Exim logs a panic warning message (but
4177 However, if deliver_drop_privilege is set, the unprivileged state may well
4178 be exactly what is intended, so the warning has been cut out in that case,
4179 and Exim is allowed to try to write to its normal log files.
4185 1. smtp_return_error_details was not giving details for temporary sender
4186 or receiver verification errors.
4188 2. Diagnose a configuration error if two authenticators have the same public
4191 3. Exim used not to create the message log file for a message until the first
4192 delivery attempt. This could be confusing when incoming messages were held
4193 for policy or load reasons. The message log file is now created at the time
4194 the message is received, and an initial "Received" line is written to it.
4196 4. The automatically generated man page for command line options had a minor
4197 bug that caused no ill effects; however, a more serious problem was that
4198 the procedure for building the man page automatically didn't always
4199 operate. Consequently, release 4.20 contains an out-of-date version. This
4200 shouldn't happen again.
4202 5. When building Exim with embedded Perl support, the script that builds the
4203 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4204 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4206 6. The freeze_tell option was not being used for messages that were frozen on
4207 arrival, either by an ACL or by local_scan().
4209 7. Added the smtp_incomplete_transaction log selector.
4211 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4212 was accepting AUTH without a new EHLO.
4214 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4215 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4218 10. Logging of TCP/IP connections (when configured) now happens in the main
4219 daemon process instead of the child process, so that the TCP/IP connection
4220 count is more accurate (but it can never be perfect).
4222 11. The use of "drop" in a nested ACL was not being handled correctly in the
4223 outer ACL. Now, if condition failure induced by the nested "drop" causes
4224 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4225 or "require"), the connection is dropped.
4227 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4228 that yield "discard" can now be used with an "accept" or a "discard" verb,
4229 but an error is generated for any others (because I can't see a useful way
4230 to define what should happen).
4232 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4233 are now processed in the same way as lines in the Exim configuration file.
4234 In particular, continuation lines are supported.
4236 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4238 15. Added -ti meaning -t -i.
4240 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4241 domains, and warn by logging if others are found.
4243 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4244 returned from GIFCONF when Exim is trying to find the list of interfaces on
4245 a host. The code in os.c has been modified to copy each ifreq to an aligned
4246 structure in all cases.
4248 Also, in some cases, the returned ifreq's were being copied to a 'struct
4249 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4250 means the last couple of bytes of an IPv6 address could be chopped if the
4251 ifreq contained only a normal sockaddr (14 bytes storage).
4253 18. Named domain lists were not supported in the hosts_treat_as_local option.
4254 An entry such as +xxxx was not recognized, and was treated as a literal
4257 19. Ensure that header lines added by a DATA ACL are included in the reject log
4258 if the ACL subsequently rejects the message.
4260 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4261 MD5 (which is deprecated).
4263 21. When testing a filter file using -bf, Exim was writing a message when it
4264 took the sender from a "From " line in the message, but it was not doing so
4265 when it took $return_path from a Return-Path: header line. It now does.
4267 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4268 with a valid header line field name (a series of printing characters
4269 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4271 23. Changed "disc" in the source to "disk" to conform to the documentation and
4272 the book and for uniformity.
4274 24. Ignore Sendmail's -Ooption=value command line item.
4276 25. When execve() failed while trying to run a command in a pipe transport,
4277 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4278 could be confused with a return value of 69 from the command itself. This
4279 has been changed to 127, the value the shell returns if it is asked to run
4280 a non-existent command. The wording for the related log line suggests a
4281 non-existent command as the problem.
4283 26. If received_header_text expands to an empty string, do not add a Received:
4284 header line to the message. (Well, it adds a token one on the spool, but
4285 marks it "old" so that it doesn't get used or transmitted.)
4287 27. Installed eximstats 1.28 (addition of -nt option).
4289 28. There was no check for failure on the call to getsockname() in the daemon
4290 code. This can fail if there is a shortage of resources on the system, with
4291 ENOMEM, for example. A temporary error is now given on failure.
4293 29. Contrary to the C standard, it seems that in some environments, the
4294 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4295 program. Exim now does this explicitly; it affects the formatting of
4296 timestamps using strftime().
4298 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4299 complaints. I've now initialized all the variables, to avoid this.
4301 32. Header lines added by a system filter were not being "seen" during
4302 transport-time rewrites.
4304 33. The info_callback() function passed to OpenSSL is set up with type void
4305 (*)(SSL *, int, int), as described somewhere. However, when calling the
4306 function (actually a macro) that sets it up, the type void(*)() is
4307 expected. I've put in a cast to prevent warnings from picky compilers.
4309 34. If a DNS black list lookup found a CNAME record, but there were no A
4310 records associated with the domain it pointed at, Exim crashed.
4312 35. If a DNS black list lookup returned more than one A record, Exim ignored
4313 all but the first. It now scans all returned addresses if a particular IP
4314 value is being sought. In this situation, the contents of the
4315 $dnslist_value variable are a list of all the addresses, separated by a
4318 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4319 to accept a host name and all its aliases if the forward lookup for any of
4320 them yielded the IP address of the incoming connection. Now it accepts only
4321 those names whose forward lookup yields the correct IP address. Any other
4322 names are discarded. This closes a loophole whereby a rogue DNS
4323 administrator could create reverse DNS records to break through a
4324 wildcarded host restriction in an ACL.
4326 37. If a user filter or a system filter that ran in a subprocess used any of
4327 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4328 the wrong values were passed to the pipe command ($thisaddress had the
4329 value of $0, $0 had the value of $1, etc). This bug was introduced by
4330 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4332 38. Improved the line breaking for long SMTP error messages from ACLs.
4333 Previously, if there was no break point between 40 and 75 characters, Exim
4334 left the rest of the message alone. Two changes have been made: (a) I've
4335 reduced the minimum length to 35 characters; (b) if it can't find a break
4336 point between 35 and 75 characters, it looks ahead and uses the first one
4337 that it finds. This may give the occasional overlong line, but at least the
4338 remaining text gets split now.
4340 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4341 file descriptors might be low, and that setting 1000 would always raise it.
4342 It turns out that in some environments, the limit is already over 1000 and
4343 that lowering it causes trouble. So now Exim takes care not to decrease it.
4345 40. When delivering a message, the value of $return_path is set to $sender_
4346 address at the start of routing (routers may change the value). By an
4347 oversight, this default was not being set up when an address was tested by
4348 -bt or -bv, which affected the outcome if any router or filter referred to
4351 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4352 log only when "message" or "log_message" are set. However, if one of the
4353 conditions was an address verification, or a call to a nested ACL, the
4354 messages generated by the underlying test were being passed through. This
4355 no longer happens. The underlying message is available in $acl_verify_
4356 message for both "message" and "log_message" expansions, so it can be
4357 passed through if needed.
4359 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4360 new expansion $bh_ to give the encoded byte string without charset
4361 translation. Translation happens only if iconv() is available; HAVE_ICONV
4362 indicates this at build time. HEADERS_CHARSET gives the charset to
4363 translate to; headers_charset can change it in the configuration, and
4364 "headers charset" can change it in an individual filter file.
4366 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4367 that creates RFC 2047 encoded "words" labels them as that charset instead
4368 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4369 expansion operator; (ii) when Exim creates a From: line for a local
4370 message; (iii) when a header line is rewritten to include a "phrase" part.
4372 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4373 buggy, causing it to skip the first lines of messages whose message ID
4374 ended in 'D'. This would not have bitten before Exim release 4.14, because
4375 message IDs were unlikely to end in 'D' before then. The effect was to have
4376 incorrect size information for certain domains.
4378 45. #include "config.h" was missing at the start of the crypt16.c module. This
4379 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4382 46. If there was a timeout during a "random" callout check, Exim treated it as
4383 a failure of the random address, and carried on sending RSET and the real
4384 address. If the delay was just some slowness somewhere, the response to the
4385 original RCPT would be taken as a response to RSET and so on, causing
4386 mayhem of various kinds.
4388 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4389 when I implemented it. It didn't allow for the fact that some option values
4390 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4391 the right test for positive values.
4393 48. Domain names in DNS records are case-independent. Exim always looks them up
4394 in lower case. Some resolvers return domain names in exactly the case they
4395 appear in the zone file, that is, they may contain uppercase letters. Not
4396 all resolvers do this - some return always lower case. Exim was treating a
4397 change of case by a resolver as a change of domain, similar to a widening
4398 of a domain abbreviation. This triggered its re-routing code and so it was
4399 trying to route what was effectively the same domain again. This normally
4400 caused routing to fail (because the router wouldn't handle the domain
4401 twice). Now Exim checks for this case specially, and just changes the
4402 casing of the domain that it ultimately uses when it transmits the message
4405 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4408 50. If a filter generated a file delivery with a non-absolute name (possible if
4409 no home directory exists for the router), the forbid_file option was not
4412 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4413 the existing equality matching.
4415 52. Exim was using ints instead of ino_t variables in some places where it was
4416 dealing with inode numbers.
4418 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4419 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4420 called TMPDIR, and if it finds it is different, it changes its value.
4422 54. The smtp_printf() function is now made available to local_scan() so
4423 additional output lines can be written before returning. There is also an
4424 smtp_fflush() function to enable the detection of a dropped connection.
4425 The variables smtp_input and smtp_batched_input are exported to
4428 55. Changed the default runtime configuration: the message "Unknown user"
4429 has been removed from the ACL, and instead placed on the localuser router,
4430 using the cannot_route_message feature. This means that any verification
4431 failures that generate their own messages won't get overridden. Similarly,
4432 the "Unrouteable address" message that was in the ACL for unverifiable
4433 relay addresses has also been removed.
4435 56. Added hosts_avoid_esmtp to the smtp transport.
4437 57. The exicyclog script was not checking for the esoteric option
4438 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4439 will work only if exicyclog is run under the appropriate euid.
4441 58. Following a discussion on the list, the rules by which Exim recognises line
4442 endings on incoming messages have been changed. The -dropcr and drop_cr
4443 options are now no-ops, retained only for backwards compatibility. The
4444 following line terminators are recognized: LF CRLF CR. However, special
4445 processing applies to CR:
4447 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4448 nor a local message in the state where . is a terminator.
4450 (ii) If a bare CR is encountered in a header line, an extra space is added
4451 after the line terminator so as not to end the header. The reasoning
4452 behind this is that bare CRs in header lines are most likely either
4453 to be mistakes, or people trying to play silly games.
4455 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4456 was being incorrectly given as 18 bytes larger than it should have been.
4457 This is a VOB (very old bug).
4459 60. This may never have affected anything current, but just in case it has:
4460 When the local host is found other than at the start of a list of hosts,
4461 the local host, those with the same MX, and any that follow, are discarded.
4462 When the list in question was part of a longer list of hosts, the following
4463 hosts (not currently being processed) were also being discarded. This no
4464 longer happens. I'm not sure if this situation could ever has previously
4467 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4470 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4471 header. This is something that is recommended in a new Internet Draft, and
4472 is something that is documented as being done by Sendmail. There are two
4473 possible values. For messages generated by the autoreply transport, Exim
4476 Auto-Submitted: auto-replied
4478 whereas for all other generated messages (e.g. bounces) it adds
4480 Auto-Submitted: auto-generated
4482 63. The "personal" condition in filters now includes a test for the
4483 Auto-Submitted: header. If it contains the string "auto-" the message it
4484 not considered personal.
4486 64. Added rcpt_include_affixes as a generic transport option.
4488 65. Added queue_only_override (default true).
4490 66. Added the syslog_duplication option.
4492 67. If what should have been the first header line of a message consisted of
4493 a space followed by a colon, Exim was mis-interpreting it as a header line.
4494 It isn't of course - it is syntactically invalid and should therefore be
4495 treated as the start of the message body. The misbehaviour could have
4496 caused a number of strange effects, including loss of data in subsequent
4497 header lines, and spool format errors.
4499 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4500 client host had authenticated. This control can now be exercised by an ACL
4501 for more flexibility.
4503 69. By default, callouts do not happen when testing with -bh. There is now a
4504 variant, -bhc, which does actually run the callout code, including
4505 consulting and updating the callout cache.
4507 70. Added support for saslauthd authentication, courtesy of Alexander
4510 71. If statvfs() failed on the spool or log directories while checking their
4511 size for availability, Exim confusingly gave the error "space shortage".
4512 Furthermore, in debugging mode it crashed with a floating point exception.
4513 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4514 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4515 serious problem, Exim now writes to the main and panic logs when this
4516 happens, with details of the failure. It then refuses to accept the
4517 incoming message, giving the message "spool directory problem" or "log
4518 directory problem" with a 421 code for SMTP messages.
4520 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4521 0, 1, and 2 exist, because some OS complain for execs without them (see
4522 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4523 descriptors. However, the code omitted to check that the open succeeded,
4524 causing mysterious errors if for some reason the permissions on /dev/null
4525 got screwed. Now Exim writes a message to the main and panic logs, and
4526 bombs out if it can't open /dev/null.
4528 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4529 interact so that it is all more flexible. It is supposed to remain
4530 backwards compatible. Also added extra_local_interfaces.
4532 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4533 to bomb out with an assertion failure - to the client this appears as a
4534 connection drop. This problem occurs in the part of the code that was taken
4535 from the Samba project. Fortunately, the assertion is in a very simple
4536 function, so I have fixed this by reproducing the function inline in the
4537 one place where it is called, and arranging for authentication to fail
4538 instead of killing the process with assert().
4540 75. The SPA client code was not working when the server requested OEM rather
4541 than Unicode encoding.
4543 76. Added code to make require_files with a specific uid setting more usable in
4544 the case where statting the file as root fails - usually a non-root-mounted
4545 NFS file system. When this happens and the failure is EACCES, Exim now
4546 forks a subprocess and does the per-uid checking as the relevant uid.
4548 77. Added process_log_path.
4550 78. If log_file_path was not explicitly set, a setting of check_log_space or
4551 check_log_inodes was ignored.
4553 79. If a space check for the spool or log partitions fails, the incident is now
4554 logged. Of course, in the latter case the data may get lost...
4556 80. Added the %p formatting code to string_format() so that it can be used to
4557 print addresses in debug_print(). Adjusted all the address printing in the
4558 debugging in store.c to use %p rather than %d.
4560 81. There was a concern that a line of code in smtp_in.c could overflow a
4561 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4562 initially expressed, the concern was not well-founded, because trailing
4563 spaces are removed early. However, if the trailing spaces were followed by
4564 a NULL, they did not get removed, so the overflow was possible. Two fixes
4567 (a) I re-wrote the offending code in a cleaner fashion.
4568 (b) If an incoming SMTP command contains a NULL character, it is rejected
4571 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4572 runs initgroups(), so that if the Exim user is in any additional groups,
4573 they will be used during message reception.
4579 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob