1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.461 2007/01/23 14:34:02 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.
59 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
60 fixed by 4.65/MH/01 (is this a record?) are fixed:
62 (i) An empty string was always treated as zero by the numeric comparison
63 operators. This behaviour has been restored.
65 (ii) It is documented that the numeric comparison operators always treat
66 their arguments as decimal numbers. This was broken in that numbers
67 starting with 0 were being interpreted as octal.
69 While fixing these problems I realized that there was another issue that
70 hadn't been noticed. Values of message_size_limit (both the global option
71 and the transport option) were treated as octal if they started with 0.
72 The documentation was vague. These values are now always treated as
73 decimal, and I will make that clear in the documentation.
79 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
80 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
83 MH/01 Don't check that the operands of numeric comparison operators are
84 integers when their expansion is in "skipping" mode (fixes bug
85 introduced by 4.64-PH/07).
87 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
88 child addresses, Exim now panics and dies. Previously, because the count
89 is held in a short int, deliveries were likely to be lost. As such a
90 large number of recipients for a single message is ridiculous
91 (performance will be very, very poor), I have chosen to impose a limit
92 rather than extend the field.
98 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
99 leftover -K file (the existence of which was triggered by #402).
100 While we were at it, introduced process PID as part of the -K
101 filename. This should rule out race conditions when creating
104 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
105 processing considerably. Previous code took too long for large mails,
106 triggering a timeout which in turn triggers #401.
108 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
109 in the DK code in transports.c. sendfile() is not really portable,
110 hence the _LINUX specificness.
112 TF/01 In the add_headers option to the mail command in an Exim filter,
113 there was a bug that Exim would claim a syntax error in any
114 header after the first one which had an odd number of characters
117 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
118 callout verification, Exim cached a "reject" for the entire domain. This
119 is correct for most verifications, but it is not correct for a recipient
120 verification with use_sender or use_postmaster set, because in that case
121 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
122 case of MAIL FROM:<> rejection from other early rejections (e.g.
123 rejection of HELO). When verifying a recipient using a non-null MAIL
124 address, the cache is ignored if it shows MAIL FROM:<> rejection.
125 Whatever the result of the callout, the value of the domain cache is
126 left unchanged (for any other kind of callout, getting as far as trying
127 RCPT means that the domain itself is ok).
129 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
132 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
133 manifest itself as EPIPE rather than ECONNECT. When tidying away a
134 session, the daemon ignores ECONNECT errors and logs others; it now
135 ignores EPIPE as well.
137 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
138 (quoted-printable decoding).
140 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
141 later the small subsequent patch to fix an introduced bug.
143 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
145 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
147 PH/08 An error is now given if message_size_limit is specified negative.
149 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
150 to be given (somewhat) arbitrary names.
152 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
155 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
156 miscellaneous code fixes
158 PH/10 Added the log_reject_target ACL modifier to specify where to log
161 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
162 hostname. This is wrong, because it relates to the incoming message (and
163 probably the interface on which it is arriving) and not to the outgoing
164 callout (which could be using a different interface). This has been
165 changed to use the value of the helo_data option from the smtp transport
166 instead - this is what is used when a message is actually being sent. If
167 there is no remote transport (possible with a router that sets up host
168 addresses), $smtp_active_hostname is used.
170 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
171 tweaks were necessary in order to get it to work (see also 21 below):
172 (a) The code assumed that strncpy() returns a negative number on buffer
173 overflow, which isn't the case. Replaced with Exim's string_format()
175 (b) There were several signed/unsigned issues. I just did the minimum
176 hacking in of casts. There is scope for a larger refactoring.
177 (c) The code used strcasecmp() which is not a standard C function.
178 Replaced with Exim's strcmpic() function.
179 (d) The code set only $1; it now sets $auth1 as well.
180 (e) A simple test gave the error "authentication client didn't specify
181 service in request". It would seem that Dovecot has changed its
182 interface. Fortunately there's a specification; I followed it and
183 changed what the client sends and it appears to be working now.
185 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
188 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
189 address A is aliased to B and C, where B exists and C does not. Without
190 -v the output is "A verified" because verification stops after a
191 successful redirection if more than one address is generated. However,
192 with -v the child addresses are also verified. Exim was outputting "A
193 failed to verify" and then showing the successful verification for C,
194 with its parentage. It now outputs "B failed to verify", showing B's
195 parentage before showing the successful verification of C.
197 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
198 look up a TXT record in a specific list after matching in a combined
201 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
202 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
203 they consult the DNS. I had assumed they would set it the way they
204 wanted; and indeed my experiments on Linux seem to show that in some
205 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
206 To be on the safe side, however, I have now made the interface to
207 host_find_byname() similar to host_find_bydns(), with an argument
208 containing the DNS resolver options. The host_find_byname() function now
209 sets these options at its start, just as host_find_bydns() does. The smtp
210 transport options dns_qualify_single and dns_search_parents are passed to
211 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
212 of host_find_byname() use the default settings of RES_DEFNAMES
213 (qualify_single) but not RES_DNSRCH (search_parents).
215 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
216 spool_read_header() do less string testing, by means of a preliminary
217 switch on the second character of optional "-foo" lines. (This is
218 overdue, caused by the large number of possibilities that now exist.
219 Originally there were few.) While I was there, I also converted the
220 str(n)cmp tests so they don't re-test the leading "-" and the first
221 character, in the hope this might squeeze out yet more improvement.
223 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
224 flag allowing group syntax was set by the header_syntax check but not
225 turned off, possible causing trouble later; (2) The flag was not being
226 set at all for the header_verify test, causing "group"-style headers to
227 be rejected. I have now set it in this case, and also caused header_
228 verify to ignore an empty address taken from a group. While doing this, I
229 came across some other cases where the code for allowing group syntax
230 while scanning a header line wasn't quite right (mostly, not resetting
231 the flag correctly in the right place). These bugs could have caused
232 trouble for malformed header lines. I hope it is now all correct.
234 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
235 with the "reply" argument non-NULL. The code, however (which originally
236 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
237 but it didn't always do it. This confused somebody who was copying the
238 code for some other use. I have removed all the tests.
240 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
241 feature that was used to support insecure browsers during the U.S. crypto
242 embargo. It requires special client support, and Exim is probably the
243 only MTA that supported it -- and would never use it because real RSA is
244 always available. This code has been removed, because it had the bad
245 effect of slowing Exim down by computing (never used) parameters for the
246 RSA_EXPORT functionality.
248 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
249 authenticator to fail if there's a tab character in the incoming data
250 (there should never be unless someone is messing about, as it's supposed
251 to be base64-encoded). Also added, on Timo's advice, the "secured" option
252 if the connection is using TLS or if the remote IP is the same as the
253 local IP, and the "valid-client-cert option" if a client certificate has
256 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
257 authenticators. This can be used for authorization after authentication
258 succeeds. (In the case of plaintext, it servers for both authentication
261 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
262 if any retry times were supplied.
264 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
265 connection, where there is no client IP address to check. In this
266 situation, the verify now always succeeds.
268 PH/25 Applied John Jetmore's -Mset patch.
270 PH/26 Added -bem to be like -Mset, but loading a message from a file.
272 PH/27 In a string expansion for a processed (not raw) header when multiple
273 headers of the same name were present, leading whitespace was being
274 removed from all of them, but trailing whitespace was being removed only
275 from the last one. Now trailing whitespace is removed from each header
276 before concatenation. Completely empty headers in a concatenation (as
279 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
280 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
282 PH/29 [Removed. This was a change that I later backed out, and forgot to
283 correct the ChangeLog entry (that I had efficiently created) before
284 committing the later change.]
286 PH/30 Exim was sometimes attempting to deliver messages that had suffered
287 address errors (4xx response to RCPT) over the same connection as other
288 messages routed to the same hosts. Such deliveries are always "forced",
289 so retry times are not inspected. This resulted in far too many retries
290 for the affected addresses. The effect occurred only when there were more
291 hosts than the hosts_max_try setting in the smtp transport when it had
292 the 4xx errors. Those hosts that it had tried were not added to the list
293 of hosts for which the message was waiting, so if all were tried, there
294 was no problem. Two fixes have been applied:
296 (i) If there are any address or message errors in an SMTP delivery, none
297 of the hosts (tried or untried) are now added to the list of hosts
298 for which the message is waiting, so the message should not be a
299 candidate for sending over the same connection that was used for a
300 successful delivery of some other message. This seems entirely
301 reasonable: after all the message is NOT "waiting for some host".
302 This is so "obvious" that I'm not sure why it wasn't done
303 previously. Hope I haven't missed anything, but it can't do any
304 harm, as the worst effect is to miss an optimization.
306 (ii) If, despite (i), such a delivery is accidentally attempted, the
307 routing retry time is respected, so at least it doesn't keep
308 hammering the server.
310 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
311 in ${readsocket because some servers need this prod.
313 PH/32 Added some extra debug output when updating a wait-xxx database.
315 PH/33 The hint "could be header name not terminated by colon", which has been
316 given for certain expansion errors for a long time, was not being given
317 for the ${if def:h_colon_omitted{... case.
319 PH/34 The spec says: "With one important exception, whenever a domain list is
320 being scanned, $domain contains the subject domain." There was at least
321 one case where this was not true.
323 PH/35 The error "getsockname() failed: connection reset by peer" was being
324 written to the panic log as well as the main log, but it isn't really
325 panic-worthy as it just means the connection died rather early on. I have
326 removed the panic log writing for the ECONNRESET error when getsockname()
329 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
330 runs only) independently of the message's sender address. This meant
331 that, if the 4xx error was in fact related to the sender, a different
332 message to the same recipient with a different sender could confuse
333 things. In particualar, this can happen when sending to a greylisting
334 server, but other circumstances could also provoke similar problems.
335 I have changed the default so that the retry time for these errors is now
336 based a combination of the sender and recipient addresses. This change
337 can be overridden by setting address_retry_include_sender=false in the
340 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
341 remote server are returned as part of bounce messages. This was not
342 happening for LMTP over a pipe (the lmtp transport), but now it is the
343 same for both kinds of LMTP.
345 PH/38 Despite being documented as not happening, Exim was rewriting addresses
346 in header lines that were in fact CNAMEs. This is no longer the case.
348 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
349 and queue runs started by the daemon processed all messages. This has
350 been fixed so that -R and -S can now usefully be given with -q<time>.
352 PH/40 Import PCRE release 6.7 (fixes some bugs).
354 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
356 PH/42 Give an error if -q is specified more than once.
358 PH/43 Renamed the variables $interface_address and $interface_port as
359 $received_ip_address and $received_port, to make it clear that these
360 values apply to message reception, and not to the outgoing interface when
361 a message is delivered. (The old names remain recognized, of course.)
363 PH/44 There was no timeout on the connect() call when using a Unix domain
364 socket in the ${readsocket expansion. There now is.
366 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
367 be meaningful with "accept".
369 SC/01 Eximstats V1.43
370 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
372 SC/02 Eximstats V1.44
373 Use a glob alias rather than an array ref in the generated
374 parser. This improves both readability and performance.
376 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
377 Collect SpamAssassin and rejection statistics.
378 Don't display local sender or destination tables unless
379 there is data to show.
380 Added average volumes into the top table text output.
382 SC/04 Eximstats V1.46
383 Collect data on the number of addresses (recipients)
384 as well as the number of messages.
386 SC/05 Eximstats V1.47
387 Added 'Message too big' to the list of mail rejection
388 reasons (thanks to Marco Gaiarin).
390 SC/06 Eximstats V1.48
391 Mainlog lines which have GMT offsets and are too short to
392 have a flag are now skipped.
394 SC/07 Eximstats V1.49 (Alain Williams)
395 Added the -emptyok flag.
397 SC/08 Eximstats V1.50
398 Fixes for obtaining the IP address from reject messages.
400 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
401 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
402 whitesspace changes from 4.64-PH/27
404 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
407 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
408 are found, allow negative numbers in numeric criteria)
410 JJ/06 exipick.20061117.2, added new $message_body_missing variable
412 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
413 to match changes made in 4.64-PH/43
415 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
417 PH/47 Put in an explicit test for a DNS lookup of an address record where the
418 "domain" is actually an IP address, and force a failure. This locks out
419 those revolvers/nameservers that support "A-for-A" lookups, in
420 contravention of the specifications.
422 PH/48 When a host name was looked up from an IP address, and the subsequent
423 forward lookup of the name timed out, the host name was left in
424 $sender_host_name, contrary to the specification.
426 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
427 restricted to single-key lookups, Exim was not diagnosing an error if
428 * or *@ was used with a query-style lookup.
430 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
432 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
433 long ago, but noone interested enough thought of it. Let's just say that
434 the "1.1" means that there are some new functions that weren't there at
435 some point in the past.
437 PH/51 Error processing for expansion failure of helo_data from an smtp
438 transport during callout processing was broken.
440 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
441 tested/used via the -bh/-bhc/-bs options.
443 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
444 bug, fixed in subsequent PCRE releases).
446 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
447 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
449 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
455 SC/01 Use a glob alias rather than an array ref in eximstats generated
456 parser. This improves both readability and performance.
458 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
459 Don't display local sender or destination tables in eximstats unless
460 there is data to show.
461 Added average volumes into the eximstats top table text output.
463 SC/03 Collect data on the number of addresses (recipients) as well
464 as the number of messages in eximstats.
466 TF/01 Correct an error in the documentation for the redirect router. Exim
467 does (usually) call initgroups() when daemonizing.
469 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
470 with consistent privilege compared to when running as a daemon.
472 TF/03 Note in the spec that $authenticated_id is not set for local
473 submissions from trusted users.
475 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
476 Thanks to Dean Brooks <dean@iglou.com> for the patch.
478 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
479 by adding some example configuration directives to the default
480 configuration file. A little bit of work is required to uncomment the
481 directives and define how usernames and passwords are checked, but
482 there is now a framework to start from.
484 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
485 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
486 without this. I don't know how relevant this is to other LDAP libraries.
488 PH/02 Add the verb name to the "unknown ACL verb" error.
490 PH/03 Magnus Holmgren's patch for filter_prepend_home.
492 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
494 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
495 directory not expanded when it should be if an expanded home directory
496 was set for the address (which is overridden by the transport).
498 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
501 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
502 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
503 because it is too late at that time, and has no effect.
505 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
506 security issue with \' (bugzilla #107). I could not use the
507 PQescapeStringConn() function, because it needs a PGconn value as one of
510 PH/08 When testing addresses using -bt, indicate those final addresses that
511 are duplicates that would not cause an additional delivery. At least one
512 person was confused, thinking that -bt output corresponded to deliveries.
513 (Suppressing duplicates isn't a good idea as you lose the information
514 about possibly different redirections that led to the duplicates.)
516 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
517 systems where poll() doesn't work, in particular OS X.
519 PH/10 Added more information to debugging output for retry time not reached.
521 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
522 operations in malware.c.
524 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
527 PH/13 If write_rejectlog was set false when logging was sent to syslog with
528 syslog_duplication set false, log lines that would normally be written
529 both the the main log and to the reject log were not written to syslog at
532 PH/14 In the default configuration, change the use of "message" in ACL warn
533 statements to "add_header".
535 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
536 not followed by a command (e.g. "seen endif").
538 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
539 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
542 PH/17 Added extra conditions to the default value of delay_warning_condition
546 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
547 { match{$h_precedence:}{(?i)bulk|list|junk} } \
548 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
551 The Auto-Submitted: and various List- headers are standardised, whereas I
552 don't think Precedence: ever was.
554 PH/18 Refactored debugging code in route_finduser() to show more information,
555 in particular, the error code if getpwnam() issues one.
557 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
558 This is apparently needed in addition to the PH/07 change above to avoid
559 any possible encoding problems.
561 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
562 but not after initializing Perl.
564 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
565 output them only if debugging. By default they are written stderr,
566 apparently, which is not desirable.
568 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
571 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
574 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
576 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
577 authenticated or an ident call has been made. Suppress the default
578 values for $authenticated_id and $authenticated_sender (but permit -oMai
579 and -oMas) when testing with -bh.
581 PH/24 Re-jigged the order of the tests in the default configuration so that the
582 tests for valid domains and recipients precede the DNS black list and CSA
583 tests, on the grounds that those ones are more expensive.
585 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
586 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
587 This bug exists in every version of Exim that I still have, right back to
590 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
591 However, an attempt to turn on case-sensitivity in a regex key by
592 including (?-i) didn't work because the subject string was already
593 lowercased, and the effects were non-intuitive. It turns out that a
594 one-line patch can be used to allow (?-i) to work as expected.
600 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
601 other effects) broke the use of negated acl sub-conditions.
603 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
606 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
607 "Deny" causes Exim to reject the incoming connection with a 554 error.
608 Unfortunately, if there is a major crisis, such as a disk failure,
609 tcp-wrappers gives "deny", whereas what one would like would be some
610 kind of temporary error. A kludge has been added to help with this.
611 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
612 554 error is used if errno is still zero or contains ENOENT (which occurs
613 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
616 PH/03 Add -lutil to the default FreeBSD LIBS setting.
618 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
619 errors. Otherwise a message that provokes a temporary error (when other
620 messages do not) can cause a whole host to time out.
622 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
623 addresses were routed directly to files or pipes from a redirect router.
624 File deliveries just didn't batch; pipe deliveries might have suffered
627 PH/06 A failure to get a lock for a hints database would erroneously always say
628 "Failed to get write lock", even when it was really a read lock.
630 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
631 of 0600. This has been changed to use the value of the lockfile_mode
632 option (which defaults to 0600).
634 PH/08 Applied small patch from the Sieve maintainer.
636 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
637 folder from quota calculations, a direct delivery into this folder messed
638 up the contents of the maildirsize file. This was because the regex was
639 used only to exclude .Trash (or whatever) when the size of the mailbox
640 was calculated. There was no check that a delivery was happening into an
641 excluded directory. This bug has been fixed by ignoring all quota
642 processing for deliveries into excluded directories.
644 PH/10 Added the maildirfolder_create_regex option to appendfile.
650 PH/01 The code for finding all the local interface addresses on a FreeBSD
651 system running IPv6 was broken. This may well have applied to all BSD
652 systems, as well as to others that have similar system calls. The broken
653 code found IPv4 interfaces correctly, but gave incorrect values for the
654 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
655 that it would not match correctly against @[] and not recognize the IPv6
658 PH/02 The ipliteral router was not recognizing addresses of the form user@
659 [ipv6:....] because it didn't know about the "ipv6:" prefix.
661 PH/03 Added disable_ipv6.
663 PH/04 Changed $reply_address to use the raw form of the headers instead of the
664 decoded form, because it is most often used to construct To: headers
665 lines in autoreplies, and the decoded form may well be syntactically
666 invalid. However, $reply_address has leading white space removed, and all
667 newlines turned into spaces so that the autoreply transport does not
670 PH/05 If group was specified without a user on a router, and no group or user
671 was specified on a transport, the group from the router was ignored.
673 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
674 for visible compile-time settings that can be used to change these
675 numbers, for those that want even more. Backwards compatibility with old
676 spool files has been maintained. However, going back to a previous Exim
677 release will lost any variables that are in spool files.
679 PH/07 Two small changes when running in the test harness: increase delay when
680 passing a TCP/IP connection to a new process, in case the original
681 process has to generate a bounce, and remove special handling of
682 127.0.0.2 (sic), which is no longer necessary.
684 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
685 be the same on different OS.
687 PH/09 Moved a debug statement in filter processing to avoid a race problem when
690 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
691 whether --show-vars was specified or not
693 JJ/02 exipick: Added support for new ACL variable spool format introduced
696 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
697 syntactically invalid From: or Reply-to: line, and a filter used this to
698 generate an autoreply, and therefore failed to obtain an address for the
699 autoreply, Exim could try to deliver to a non-existent relative file
700 name, causing unrelated and misleading errors. What now happens is that
701 it logs this as a hard delivery error, but does not attempt to create a
704 PH/11 The exinext utility has a -C option for testing purposes, but although
705 the given file was scanned by exinext itself; it wasn't being passed on
706 when Exim was called.
708 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
709 an end-of-file indication when reading a command response.
711 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
712 compiled. In many other places in Exim, IPv6 addresses are always
713 recognized, so I have changed this. It also means that IPv4 domain
714 literals of the form [IPV4:n.n.n.n] are now always recognized.
716 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
717 used if the router is not running as root, for example, when verifying at
718 ACL time, or when using -bh. The debugging output from this situation was
719 non-existent - all you got was a failure to exec. I have made two
722 (a) Failures to set uid/gid, the current directory, or a process leader
723 in a subprocess such as that created by queryprogram now generate
724 suitable debugging ouput when -d is set.
726 (b) The queryprogram router detects when it is not running as root,
727 outputs suitable debugging information if -d is set, and then runs
728 the subprocess without attempting to change uid/gid.
730 PH/15 Minor change to Makefile for building test_host (undocumented testing
733 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
734 additional section of a DNS packet that returns MX or SRV records.
735 Instead, it always explicitly searches for A/AAAA records. This avoids
736 major problems that occur when a DNS server includes only records of one
737 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
738 fixed another bug: if SRV records were looked up and the corresponding
739 address records were *not* found in the additional section, the port
740 values from the SRV records were lost.
742 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
743 using the correct key (the original address) when searching the retry
744 rules in order to find which one to use for generating the retry hint.
746 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
747 adding the default one. Similarly, if it contains a Reply-To: header, the
748 errors_reply_to option, if set, is not used.
750 PH/19 When calculating a retry time, Exim used to measure the "time since
751 failure" by looking at the "first failed" field in the retry record. Now
752 it does not use this if it is later than than the arrival time of the
753 message. Instead it uses the arrival time. This makes for better
754 behaviour in cases where some deliveries succeed, thus re-setting the
755 "first failed" field. An example is a quota failure for a huge message
756 when small messages continue to be delivered. Without this change, the
757 "time since failure" will always be short, possible causing more frequent
758 delivery attempts for the huge message than are intended.
759 [Note: This change was subsequently modified - see PH/04 for 4.62.]
761 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
762 $1, $2, $3) because the numerical variables can be reset during some
763 expansion items (e.g. "match"), thereby losing the authentication data.
765 PH/21 Make -bV show the size of off_t variables so that the test suite can
766 decide whether to run tests for quotas > 2G.
768 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
769 mailbox_size, and mailbox_filecount in the appendfile transport. If a
770 filecount value is greater than 2G or if a quota value is greater than 2G
771 on a system where the size of off_t is not greater than 4, a panic error
774 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
775 never match. The debug and -bh output now contains an explicit error
776 message indicating a malformed IPv4 address or mask.
778 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
779 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
782 PH/25 Do not write to syslog when running in the test harness. The only
783 occasion when this arises is a failure to open the main or panic logs
784 (for which there is an explicit test).
786 PH/26 Added the /no_tell option to "control=freeze".
788 PH/27 If a host name lookup failed very early in a connection, for example, if
789 the IP address matched host_lookup and the reverse lookup yielded a name
790 that did not have a forward lookup, an error message of the form "no IP
791 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
792 could be logged. Now it outputs the IP address instead of "NULL".
794 PH/28 An enabling patch from MH: add new function child_open_exim2() which
795 allows the sender and the authenticated sender to be set when
796 submitting a message from within Exim. Since child_open_exim() is
797 documented for local_scan(), the new function should be too.
799 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
800 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
801 results in an empty string is now treated as unset.
803 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
805 PH/31 Added sender_verify_fail logging option.
807 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
808 needed by RFC 822 but not by RFC 2822 was commented out. I have now
809 tidied the source and removed it altogether.
811 PH/33 When a queue run was abandoned because the load average was too high, a
812 log line was always written; now it is written only if the queue_run log
813 selector is set. In addition, the log line for abandonment now contains
814 information about the queue run such as the pid. This is always present
815 in "start" and "stop" lines but was omitted from the "abandon" line.
817 PH/34 Omit spaces between a header name and the colon in the error message that
818 is given when verify = headers_syntax fails (if there are lots of them,
819 the message gets confusing).
821 PH/35 Change the default for dns_check_names_pattern to allow slashes within
822 names, as there are now some PTR records that contain slashes. This check
823 is only to protect against broken name servers that fall over on strange
824 characters, so the fact that it applies to all lookups doesn't matter.
826 PH/36 Now that the new test suite is complete, we can remove some of the
827 special code in Exim that was needed for the old test suite. For example,
828 sorting DNS records because real resolvers return them in an arbitrary
829 order. The new test suite's fake resolver always returns records in the
832 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
833 bounces) except when queue_only is set, to avoid logging races between
834 the different processes.
836 PH/38 Panic-die if .include specifies a non-absolute path.
838 PH/39 A tweak to the "H" retry rule from its user.
840 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
841 a label. They prevented compilation on older perls.
843 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
844 a warning to be raised on newish perls.
846 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
847 on queue. Changes to match documented behaviour of showing count of
848 messages matching specified criteria.
850 PH/40 Changed the default ident timeout from 30s to 5s.
852 PH/41 Added support for the use of login_cap features, on those BSD systems
853 that have them, for controlling the resources used by pipe deliveries.
855 PH/42 The content-scanning code uses fopen() to create files in which to put
856 message data. Previously it was not paying any attention to the mode of
857 the files. Exim runs with umask(0) because the rest of the code creates
858 files with open(), and sets the required mode explicitly. Thus, these
859 files were ending up world-writeable. This was not a big issue, because,
860 being within the spool directory, they were not world-accessible. I have
861 created a function called modefopen, which takes an additional mode
862 argument. It sets umask(777), creates the file, chmods it to the required
863 mode, then resets the umask. All the relevant calls to fopen() in the
864 content scanning code have been changed to use this function.
866 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
867 to 24 hours. This avoids potential overflow problems when processing G
868 and H retry rules. I suspect nobody ever tinkers with this value.
870 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
872 PH/45 When the plaintext authenticator is running as a client, the server's
873 challenges are checked to ensure they are valid base64 strings. By
874 default, the authentication attempt is cancelled if an invalid string is
875 received. Setting client_ignore_invalid_base64 true ignores these errors.
876 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
877 they are received. Thus, the responses can be made to depend on the
878 challenges. If an invalid string is ignored, an empty string is placed in
881 PH/46 Messages that are created by the autoreply transport now contains a
882 References: header, in accordance with RFCs 2822 and 3834.
884 PH/47 Added authenticated_sender_force to the smtp transport.
886 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
888 PH/49 Installed latest patch from the Sieve maintainer.
890 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
891 was also set, the appendfile transport was unnecessarily scanning a
892 directory of message files (e.g. for maildir delivery) to find the count
893 of files (along with the size), even though it did not need this
894 information. It now does the scan only if it needs to find either the
895 size of the count of files.
897 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
899 PH/52 Two bugs concerned with error handling when the smtp transport is
902 (i) Exim was not creating retry information for temporary errors given
903 for individual recipients after the DATA command when the smtp transport
904 was used in LMTP mode. This meant that they could be retried too
905 frequently, and not timed out correctly.
907 (ii) Exim was setting the flag that allows error details to be returned
908 for LMTP errors on RCPT commands, but not for LMTP errors for individual
909 recipients that were returned after the DATA command.
911 PH/53 This is related to PH/52, but is more general: for any failing address,
912 when detailed error information was permitted to be returned to the
913 sender, but the error was temporary, then after the final timeout, only
914 "retry timeout exceeded" was returned. Now it returns the full error as
915 well as "retry timeout exceeded".
917 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
918 do this, and (what is worse) MTAs that accept it.
920 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
921 will now be deprecated.
923 PH/56 New os.c-cygwin from the Cygwin maintainer.
925 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
926 formats (previously only available in exim formats via -bpr, -bpru,
927 and -bpra. Now also available in native and exiqgrep formats)
929 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
930 with very large, slow to parse queues
932 JJ/08 exipick: added ! as generic prefix to negate any criteria format
934 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
936 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
937 responses to authentication challenges, though it was showing the
938 challenges; (ii) I've removed the CR characters from the debug output for
941 PH/58 Allow for the insertion of a newline as well as a space when a string
942 is turned into more than one encoded-word during RFC 2047 encoding. The
943 Sieve code now uses this.
945 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
946 data_4xx, lost_connection, tls_required.
948 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
949 message was being sent as an SMTP response.
951 PH/61 Add -l and -k options to exicyclog.
953 PH/62 When verifying, if an address was redirected to one new address, so that
954 verification continued, and the new address failed or deferred after
955 having set something in $address_data, the value of $address_data was not
956 passed back to the ACL. This was different to the case when no
957 redirection occurred. The value is now passed back in both cases.
959 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
960 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
961 use login_cap.h, so on its own it isn't the distinguishing feature. The
962 new name refers directly to the setclassresources() function.
964 PH/65 Added configuration files for NetBSD3.
966 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
968 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
969 is preferred over IPv4.
971 PH/68 The bounce_return_message and bounce_return_body options were not being
972 honoured for bounces generated during the reception of non-SMTP messages.
973 In particular, this applied to messages rejected by the ACL. This bug has
974 been fixed. However, if bounce_return_message is true and bounce_return_
975 body is false, the headers that are returned for a non-SMTP message
976 include only those that have been read before the error was detected.
977 (In the case of an ACL rejection, they have all been read.)
979 PH/69 The HTML version of the specification is now built in a directory called
980 spec_html instead of spec.html, because the latter looks like a path with
981 a MIME-type, and this confuses some software.
983 PH/70 Catch two compiler warnings in sieve.c.
985 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
986 function verify_get_ident() calls ip_connect() to connect a socket, but
987 if the "connect()" function timed out, ip_connect() used to close the
988 socket. However, verify_get_ident() also closes the socket later, and in
989 between Exim writes to the log, which may get opened at this point. When
990 the socket was closed in ip_connect(), the log could get the same file
991 descriptor number as the socket. This naturally causes chaos. The fix is
992 not to close the socket in ip_connect(); the socket should be closed by
993 the function that creates it. There was only one place in the code where
994 this was missing, in the iplookup router, which I don't think anybody now
995 uses, but I've fixed it anyway.
997 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
998 well as to direct DNS lookups. Otherwise the handling of names in host
999 lists is inconsistent and therefore confusing.
1005 PH/01 Two changes to the default runtime configuration:
1007 (1) Move the checks for relay_from_hosts and authenticated clients from
1008 after to before the (commented out) DNS black list checks.
1010 (2) Add control=submission to the relay_from_hosts and authenticated
1011 clients checks, on the grounds that messages accepted by these
1012 statements are most likely to be submissions.
1014 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1016 (1) Generate an error if the third argument for the ${prvs expansion is
1019 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1022 (3) Reset the variables that are obtained from the first argument of
1023 ${prvscheck and used in the second argument before leaving the code,
1024 because their memory is reclaimed, so using them afterwards may do
1027 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1028 one (it's much easier than Tom thought :-).
1030 (5) Because of (4), we can now allow for the use of $prvscheck_result
1031 inside the third argument.
1033 PH/03 For some reason, the default setting of PATH when running a command from
1034 a pipe transport was just "/usr/bin". I have changed it to
1037 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1038 anything to be listed in the output from -bV.
1040 PH/05 When a filter generated an autoreply, the entire To: header line was
1041 quoted in the delivery log line, like this:
1043 => >A.N.Other <ano@some.domain> <original@ddress> ...
1045 This has been changed so that it extracts the operative address. There
1046 may be more than one such address. If so, they are comma-separated, like
1049 => >ano@some.domain,ona@other.domain <original@ddress> ...
1051 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1052 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1053 not being looked up in the rDNS to get a host name, Exim was showing the
1054 IP address twice in Received: lines, even though the IP addresses were
1055 identical. For example:
1057 Received: from [1.2.3.4] (helo=[1.2.3.4])
1059 However, if the real host name was known, it was omitting the HELO data
1060 if it matched the actual IP address. This has been tidied up so that it
1061 doesn't show the same IP address twice.
1063 PH/07 When both +timestamp and +memory debugging was on, the value given by
1064 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1065 called by the debug printing, thereby overwriting the timestamp buffer.
1066 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1068 PH/08 When the original message was included in an autoreply transport, it
1069 always said "this is a copy of the message, including all the headers",
1070 even if body_only or headers_only was set. It now gives an appropriate
1073 PH/09 Applied a patch from the Sieve maintainer which:
1075 o fixes some comments
1076 o adds the (disabled) notify extension core
1077 o adds some debug output for the result of if/elsif tests
1078 o points to the current vacation draft in the documentation
1079 and documents the missing references header update
1083 o fixes a bug in processing the envelope test (when testing
1084 multiple envelope elements, the last element determinted the
1087 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1088 Electronic Mail") by including:
1090 Auto-submitted: auto-generated
1092 in the messages that it generates (bounce messages and others, such as
1093 warnings). In the case of bounce messages for non-SMTP mesages, there was
1094 also a typo: it was using "Auto_submitted" (underscore instead of
1095 hyphen). Since every message generated by Exim is necessarily in response
1096 to another message, thes have all been changed to:
1098 Auto-Submitted: auto-replied
1100 in accordance with these statements in the RFC:
1102 The auto-replied keyword:
1104 - SHOULD be used on messages sent in direct response to another
1105 message by an automatic process,
1107 - MUST NOT be used on manually-generated messages,
1109 - MAY be used on Delivery Status Notifications (DSNs) and Message
1110 Disposition Notifications (MDNs),
1112 - MUST NOT be used on messages generated by automatic or periodic
1113 processes, except for messages which are automatic responses to
1116 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1117 to the default Received: header definition.
1119 PH/12 Added log selector acl_warn_skipped (default on).
1121 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1122 variables because (a) they are in the wrong storage pool and (b) even if
1123 they were copied, it wouldn't work properly because of the caching.
1125 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1126 checking when decoding. Apparently there are clients that generate
1127 overlong encoded strings. Why am I not surprised?
1129 PH/15 If the first argument of "${if match_address" was not empty, but did not
1130 contain an "@" character, Exim crashed. Now it writes a panic log message
1131 and treats the condition as false.
1133 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1135 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1136 extension "envelope-auth". The code is finished and in agreement with
1137 other implementations, but there is no documentation so far and in fact,
1138 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1139 not changing the active code.
1141 Print executed "if" and "elsif" statements when debugging is used. This
1142 helps a great deal to understand what a filter does.
1144 Document more things not specified clearly in RFC3028. I had all this
1145 sorted out, when out of a sudden new issues came to my mind. Oops."
1147 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1150 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1151 canonical form (as documented). However, after a host name lookup from
1152 the IP address, check_host() was doing a simple string comparison with
1153 addresses acquired from the DNS when checking that the found name did
1154 have the original IP as one of its addresses. Since any found IPv6
1155 addresses are likely to be in abbreviated form, the comparison could
1156 fail. Luckily, there already exists a function for doing the comparison
1157 by converting both addresses to binary, so now that is used instead of
1158 the text comparison.
1160 PH/20 There was another similar case to PH/19, when a complete host name was
1161 given in a host list; looking up its IP address could give an abbreviated
1162 form, whereas the current host's name might or might not be abbreviated.
1163 The same fix has been applied.
1169 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1170 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1173 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1174 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1176 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1178 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1179 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1180 Also, for "Auto-Submitted", treat anything other than "no" as
1181 non-personal, in accordance with RFC 3834. (Previously it treated
1182 anything starting "auto-" as non-personal.)
1184 TF/01 The control=submission/name=... option had a problem with syntax
1185 errors if the name included a slash character. The /name= option
1186 now slurps the rest of the string, so it can include any characters
1187 but it must come last in the list of options (after /sender_retain
1190 PH/05 Some modifications to the interface to the fake nameserver for the new
1198 TK/01 Added the "success_on_redirect" address verification option. See
1199 NewStuff for rationale and an example.
1201 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1203 PH/02 Patch to exigrep to allow it to work on syslog lines.
1205 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1206 fread() to skip over the body file's header line, because in Cygwin the
1207 header line is locked and is inaccessible.
1209 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1210 co-exist for some time) to make it clear that it is the Exim ID that is
1211 referenced, not the Message-ID: header line.
1213 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1214 string_format() function, because snprintf() does not exist on all
1217 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1218 ${stat: expansion item.
1220 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1221 protocol synchronization error", to keep the pedants happy.
1223 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1224 well as for IRIX systems, when gcc is being used. See the host.c source
1227 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1229 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1232 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1233 transport and to the smtp transport in LMTP mode.
1235 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1237 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1238 run a filter in a subprocess. This could lead to confusion in subsequent
1239 lookups in the parent process. There should also be a search_tidyup() at
1240 the end of the subprocess.
1242 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1243 only if the host matched helo_try_verify_hosts, which caused the
1244 verification to occur when the EHLO/HELO command was issued. The ACL just
1245 tested the remembered result. Now, if a previous verification attempt has
1246 not happened, "verify = helo" does it there and then.
1248 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1250 TK/03 Fix log output including CR from clamd.
1252 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1253 exist provoked a memory error which could cause a segfault.
1255 PH/15 Installed PCRE 6.2
1257 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1259 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1260 of the problem. Specifically, suggested +O2 rather than +O1 for the
1263 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1265 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1266 record, the port was not correctly logged when the outgoing_port log
1267 selector was set (it logged the transort's default port).
1269 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1270 fallback_hosts, and "hosts" in the smtp transport.
1272 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1273 host errors such as "Connection refused".
1275 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1276 authentication with radiusclient 0.4.9:
1278 - Error returned from rc_read_config was caught wrongly
1279 - Username/password not passed on to radius server due to wrong length.
1281 The presumption is that some radiusclient API changes for 4.51/PH/17
1282 were not taken care of correctly. The code is still untested by me (my
1283 Linux distribution still has 0.3.2 of radiusclient), but it was
1284 contributed by a Radius user.
1286 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1287 expanding the "port" option of the smtp transport.
1289 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1290 while reading a MIME header. Thanks to Tom Hughes for a patch.
1292 PH/24 Include config.h inside local_scan.h so that configuration settings are
1295 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1296 that in an ACL for RCPT (for example), you can examine exactly what was
1299 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1300 commands, but it was not correctly comparing the address with the actual
1301 client host address. Thus, it would show the EHLO address in Received:
1302 header lines when this was not necessary.
1304 PH/27 Added the % operator to ${eval:}.
1306 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1307 it should be ignoring failures (because with -C, for example, it has lost
1308 privilege). It wasn't ignoring creation failures other than "already
1311 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1314 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1315 because an input line was too long, either on its own, or by virtue of
1316 too many continuations, the temporary file was not being removed, and the
1317 return code was incorrect.
1319 PH/31 Missing "BOOL" in function definition in filtertest.c.
1321 PH/32 Applied Sieve patches from the maintainer.
1323 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1325 PH/33 Added "verify = not_blind".
1327 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1328 Local/Makefile (with some defaults set). These are used in built scripts
1329 such as exicyclog, but they have never been used in the exim_install
1330 script (though there are many overriding facilities there). I have
1331 arranged that the exim_install script now takes note of these two
1334 PH/35 Installed configuration files for Dragonfly.
1336 PH/36 When a locally submitted message by a trusted user did not contain a
1337 From: header, and the sender address was obtained from -f or from an SMTP
1338 MAIL command, and the trusted user did not use -F to supply a sender
1339 name, $originator_name was incorrectly used when constructing a From:
1340 header. Furthermore, $originator_name was used for submission mode
1341 messages from external hosts without From: headers in a similar way,
1342 which is clearly wrong.
1344 PH/37 Added control=suppress_local_fixups.
1346 PH/38 When log_selector = +received_sender was set, and the addition of the
1347 sender made the log line's construction buffer exactly full, or one byte
1348 less than full, an overflow happened when the terminating "\n" was
1351 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1352 when the result of a list match is failure because a DNS lookup failed.
1354 PH/40 RM_COMMAND is now used in the building process.
1356 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1357 the "build-* directories that it finds.
1359 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1360 address in a domain literal was a prefix of an interface address.
1362 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1363 when verifying a sender address, unless rewrite_headers is false.
1365 PH/44 Wrote a long comment about why errors_to addresses are verified as
1366 recipients, not senders.
1368 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1369 the ratelimit ACL was added.
1371 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1373 PH/46 Added extra information about PostgreSQL errors to the error string.
1375 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1376 avoiding the need to install special zones in a real server. This is
1377 backwards compatible; if it can't find the fake resolver, it drops back.
1378 Thus, both old and new test suites can be run.
1380 TF/02 Added util/ratelimit.pl
1382 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1383 clock is set back in time.
1385 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1386 Candler <B.Candler@pobox.com>.
1388 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1389 OK for addresses that are the result of redirections.
1391 PH/48 A number of further additions for the benefit of the new test suite,
1392 including a fake gethostbyname() that interfaces to the fake DNS resolver
1395 TF/06 The fix for widen_domains has also been applied to qualify_single and
1396 search_parents which are the other dnslookup options that can cause
1399 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1402 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1404 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1405 service" if it encountered an MX record with an empty target hostname.
1406 The message is now "an MX or SRV record indicated no SMTP service".
1408 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1409 if the DNS of the sending site is misconfigured. This is quite a
1410 common situation. This change restores the behaviour of treating a
1411 helo verification defer as a failure.
1413 PH/51 If self=fail was set on a router, the bounce message did not include the
1414 actual error message.
1420 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1422 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1423 command itself ended in error, the underlying message about the transport
1424 filter timeout was being overwritten with the pipe command error. Now the
1425 underlying error message should be appended to the second error message.
1427 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1429 PH/02 Reduce the amount of output that "make" produces by default. Full output
1430 can still be requested.
1432 PH/03 The warning log line about a condition test deferring for a "warn" verb
1433 was being output only once per connection, rather than after each
1434 occurrence (because it was using the same function as for successful
1435 "warn" verbs). This seems wrong, so I have changed it.
1437 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1438 it should not have, which might have caused a crash in the right
1439 circumstances, but probably never did.
1441 PH/04 Installed a modified version of Tony Finch's patch to make submission
1442 mode fix the return path as well as the Sender: header line, and to
1443 add a /name= option so that you can make the user's friendly name appear
1446 TF/03 Added the control = fakedefer ACL modifier.
1448 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1449 Mark Lowes for thorough testing.
1451 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1453 TK/03 Merged latest SRS patch from Miles Wilton.
1455 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1456 with the definition in sysexits.h (which is #included earlier).
1457 Fortunately, Exim does not actually use EX_OK. The code used to try to
1458 preserve the sysexits.h value, by assumimg that macro definitions were
1459 scanned for macro replacements. I have been disabused of this notion,
1460 so now the code just undefines EX_OK before #including unistd.h.
1462 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1463 in the smtp transport. When a block could not be written in a single
1464 write() function, the timeout was being re-applied to each part-write.
1465 This seems wrong - if the receiver was accepting one byte at a time it
1466 would take for ever. The timeout is now adjusted when this happens. It
1467 doesn't have to be particularly precise.
1469 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1470 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1472 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1473 without a domain if the check to <postmaster@domain> fails.
1475 SC/01 Eximstats: added -xls and the ability to specify output files
1476 (patch written by Frank Heydlauf).
1478 SC/02 Eximstats: use FileHandles for outputing results.
1480 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1482 SC/04 Eximstats: fixed display of large numbers with -nvr option
1484 SC/05 Eximstats: fixed merging of reports with empty tables.
1486 SC/06 Eximstats: added the -include_original_destination flag
1488 SC/07 Eximstats: removed tabs and trailing whitespace.
1490 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1492 TK/06 MBOX spool code: Add real "From " MBOX separator line
1493 so the .eml file is really in mbox format (even though
1494 most programs do not really care). Patch from Alex Miller.
1496 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1497 The latter is generated from $received_to and is only set if the
1498 message has one envelope recipient. SA can use these headers,
1499 obviously out-of-the-box. Patch from Alex Miller.
1501 PH/08 The ${def test on a variable was returning false if the variable's
1502 value was "0", contrary to what the specification has always said!
1503 The result should be true unless the variable is empty.
1505 PH/09 The syntax error of a character other than { following "${if
1506 def:variable_name" (after optional whitespace) was not being diagnosed.
1507 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1508 accidental colon was present, for example, could give incorrect results.
1510 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1511 result is used (not including appendfile, where other changes are about
1514 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1515 This involved changing a lot of size variables from int to off_t. It
1516 should work with maildirs and everything.
1518 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1519 spamd dying while we are connected to it.
1521 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1524 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1525 function for BDB 4.3.
1527 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1529 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1530 the sources for PCRE have been re-arranged and more files are now
1533 PH/15 The code I had for printing potentially long long variables in PH/11
1534 above was not the best (it lost precision). The length of off_t variables
1535 is now inspected at build time, and an appropriate printing format (%ld
1536 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1537 to be "long long int" or "long int". This is needed for the internal
1538 formatting function string_vformat().
1540 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1541 the configuration file to be ":syslog", then the script "guesses" where
1542 the logs files are, rather than using the compiled in default. In our
1543 case the guess is not the same as the compiled default, so the script
1544 suddenly stopped working when I started to use syslog. The patch checks
1545 to see if log_file_path is "". If so, it attempts to read it from exim
1546 with no configuration file to get the compiled in version, before it
1547 falls back to the previous guessing code."
1549 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1550 implementing BATV in an Exim configuration. See NewStuff for the gory
1553 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1554 Makefile that are specific to HP-UX.
1556 PH/18 If the "use_postmaster" option was set for a recipient callout together
1557 with the "random" option, the postmaster address was used as the MAIL
1558 FROM address for the random test, but not for the subsequent recipient
1559 test. It is now used for both.
1561 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1562 patch removes a few documentation additions to RFC 3028, because the
1563 latest draft now contains them. It adds the new en;ascii-case comparator
1564 and a new error check for 8bit text in MIME parts. Comparator and
1565 require names are now matched exactly. I enabled the subaddress
1566 extension, but it is not well tested yet (read: it works for me)."
1568 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1569 rework some of the code of TK/09 above to avoid the hardwired use of
1570 "%lld" and "long long". Replaced the call to snprintf() with a call to
1573 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1574 records point to non-existent hosts", "retry timeout exceeded", and
1575 "retry time not reached for any host after a long failure period".
1577 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1578 experimental DomainKeys support:
1580 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1581 (2) On an error such as an illegally used "control", the wrong name for
1582 the control was given.
1584 These problems did NOT occur unless DomainKeys support was compiled.
1586 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1588 PH/24 Added ${if match_ip condition.
1590 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1591 databases so that it will be absolutely obvious if a crash occurs in the
1592 DB library. This is a regular occurrence (often caused by mis-matched
1595 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1596 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1597 picked up on a user's system that detects such things. There doesn't seem
1598 to be a gcc warning option for this - only an attribute that has to be
1599 put on the function's prototype. It seems that in Fedora Core 4 they have
1600 set this on a number of new functions. No doubt there will be more in due
1603 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1604 specify a transport. However, if an address that was verified by such a
1605 router was the subject of a callout, Exim crashed because it tried to
1606 read the rcpt_include_affixes from the non-existent transport. Now it
1607 just assumes that the setting of that option is false. This bug was
1608 introduced by 4.51/PH/31.
1610 PH/28 Changed -d+all to exclude +memory, because that information is very
1611 rarely of interest, but it makes the output a lot bigger. People tend to
1612 do -d+all out of habit.
1614 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1615 code in os-type was giving problems when libc.so lives in lib64, like on
1618 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1619 aren't the modern standard, and it seems that some systems' include files
1620 don't always have them. Exim was already checking for some of the newer
1621 ones like T_AAAA, and defining it itself. I've added checks for all the
1622 record types that Exim uses.
1624 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1625 not automatically generating a new one, as it is supposed to. This
1626 prevented TLS from working. If the file did exist, but contained invalid
1627 data, a new version was generated, as expected. It was only the case of a
1628 non-existent file that was broken.
1630 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1631 with a change in libdomainkeys > 0.64.
1633 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1634 from DNS. If the selector record carries the flag, it now has
1635 precedence over the domain-wide flag.
1637 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1639 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1640 the use of a port name that isn't defined in /etc/services) occurred, the
1641 message was deferred as in a normal delivery, and thus remained on the
1642 spool, instead of being failed because of the mua_wrapper setting. This
1643 is now fixed, and I tidied up some of the mua_wrapper messages at the
1646 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1647 the messages in a hash of arrays rather than using individual hashes.
1648 This is a bit cleaner and results in dramatic memory savings, albeit
1649 at a slight CPU cost.
1651 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1652 as requested by Marc Sherman.
1654 SC/10 Eximstats: added histograms for user specified patterns as requested
1657 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1659 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1660 fopen() in the content-scanning modules that did not already have it.
1666 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1667 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1669 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1671 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1673 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1674 to test Sieve filters that use "vacation".
1676 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1677 that changes the way the GnuTLS parameters are stored in the cache file.
1678 The new format can be generated externally. For backward compatibility,
1679 if the data in the cache doesn't make sense, Exim assumes it has read an
1680 old-format file, and it generates new data and writes a new file. This
1681 means that you can't go back to an older release without removing the
1684 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1685 work if there are any delivery delays because "one_time" forces the
1686 parent to be marked "delivered", so its unseen clone is never tried
1687 again. For this reason, Exim now forbids the simultaneous setting of
1690 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1691 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1692 entry if you want to know the details. The fix, however, neglected to
1693 consider the case when local delivery batching is involved. The test for
1694 "previously delivered" was not happening when checking to see if an
1695 address could be batched with a previous (undelivered) one; under
1696 certain circumstances this could lead to multiple deliveries to the same
1699 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1700 in its include files, and this causes problems building Exim.
1702 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1703 header_syntax) but Exim was just ignoring anything given after a slash.
1704 In particular, this caused confusion with an attempt to use "verify =
1705 reverse_host_lookup/defer_ok". An error is now given when options are
1706 supplied for verify items that do not have them. (Maybe reverse_host_
1707 lookup should have a defer_ok option, but that's a different point.)
1709 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1710 defined by RFC 821) to 2048, because there were problems with some AUTH
1711 commands, and RFC 1869 says the size should be increased for extended
1712 SMTP commands that take arguments.
1714 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1717 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1718 "unknown" error; now it says that the functionality isn't in the binary.
1720 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1721 an address' error message when a string expansion fails (syntax or
1722 whatever). Otherwise the password may appear in the log. Following change
1723 PH/42 below, there is no longer a chance of it appearing in a bounce
1726 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1728 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1729 of its addresses was ever tried. (Bugzilla bug #2.)
1731 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1732 the result incorrectly in the debug output. (It correctly added a newline
1733 to what was transported.)
1735 TF/01 Added $received_time.
1737 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1738 commented out examples of how to interface to a virus scanner and to
1739 SpamAssassin. Also added commented examples of av_scanner and
1740 spamd_address settings.
1742 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1743 and controls are allowed in which ACLs. There were a couple of minor
1744 errors. Some of the entries in the conditions table (which is a table of
1745 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1746 negation of where the condition IS allowed.
1748 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1750 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1751 header file does not have a version number, so I've had to invent a new
1752 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1753 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1754 radiusclient), but it was contributed by a Radius user.
1756 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1757 files or directories, for OpenSSL.
1759 PH/19 When an Exim process that is running as root has to create an Exim log
1760 file, it does so in a subprocess that runs as exim:exim so as to get the
1761 ownership right at creation (otherwise, other Exim processes might see
1762 the file with the wrong ownership). There was no test for failure of this
1763 fork() call, which would lead to the process getting stuck as it waited
1764 for a non-existent subprocess. Forks do occasionally fail when resources
1765 run out. I reviewed all the other calls to fork(); they all seem to check
1768 PH/20 When checking for unexpected SMTP input at connect time (before writing
1769 the banner), Exim was not dealing correctly with a non-positive return
1770 from the read() function. If the client had disconnected by this time,
1771 the result was a log entry for a synchronization error with an empty
1772 string after "input=" when read() returned zero. If read() returned -1
1773 (an event I could not check), uninitialized data bytes were printed.
1774 There were reports of junk text (parts of files, etc) appearing after
1777 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1779 PH/22 Added support for macro redefinition, and (re)definition in between
1780 driver and ACL definitions.
1782 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1783 forgetting to use the resulting value; it was using the unexpanded value.
1785 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1786 hadn't been configured. The fix is from Juergen Kreileder, who
1787 understands it better than I do:
1789 "Here's what I see happening with three configured cyrus_sasl
1790 authenticators configured (plain, login, cram-md5):
1792 On startup auth_cyrus_sasl_init() gets called for each of these.
1793 This means three calls to sasl_listmech() without a specified mech_list.
1794 => SASL tests which mechs of all available mechs actually work
1795 => three warnings about OTP not working
1796 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1798 With the patch, sasl_listmech() also gets called three times. But now
1799 SASL's mech_list option is set to the server_mech specified in the the
1800 authenticator. Or in other words, the answer from sasl_listmech()
1801 gets limited to just the mech you're testing for (which is different
1803 => the return list contains just 'plain' or 'login', 'cram-md5' or
1804 nothing depending on the value of ob->server_mech.
1806 I've just tested the patch: Authentication still works fine,
1807 unavailable mechs specified in the exim configuration are still
1808 caught, and the auth.log warnings about OTP are gone."
1810 PH/25 When debugging is enabled, the contents of the command line are added
1811 to the debugging output, even when log_selector=+arguments is not
1814 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1815 answer is "GNU", and only if the return is "GNU/something" is the answer
1818 PH/27 $acl_verify_message is now set immediately after the failure of a
1819 verification in an ACL, and so is available in subsequent modifiers. In
1820 particular, the message can be preserved by coding like this:
1822 warn !verify = sender
1823 set acl_m0 = $acl_verify_message
1825 Previously, $acl_verify_message was set only while expanding "message"
1826 and "log_message" when a very denied access.
1828 PH/28 Modified OS/os.c-Linux with
1830 -#ifndef OS_LOAD_AVERAGE
1831 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1833 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1834 nomenclature these days.)
1836 PH/29 Installed patch from the Sieve maintainer that adds the options
1837 sieve_useraddress and sieve_subaddress to the redirect router.
1839 PH/30 In these circumstances:
1840 . Two addresses routed to the same list of hosts;
1841 . First host does not offer TLS;
1842 . First host accepts first address;
1843 . First host gives temporary error to second address;
1844 . Second host offers TLS and a TLS session is established;
1845 . Second host accepts second address.
1846 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1847 and peerdn, if requested) that were in fact used only for the second
1850 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1851 attention to any local part prefix or suffix that was matched by the
1852 router that accepted the address. It now behaves in the same way as it
1853 does for delivery: the affixes are removed from the local part unless
1854 rcpt_include_affixes is set on the transport.
1856 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1857 timeout during the DATA phase of an incoming message.
1859 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1860 applied a patch sanctioned by the Sieve maintainer.
1862 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1863 the uid or gid is negative. A case of a negative gid caused this to be
1864 noticed. The fix allows for either to be negative.
1866 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1867 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1870 PH/36 The change PH/12 above was broken. Fixed it.
1872 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1873 the grounds that routing the same address twice would always produce the
1874 same answer. This might have been true once, but it is certainly no
1875 longer true now. Routing a child address may depend on the previous
1876 routing that produced that child. Some complicated redirection strategies
1877 went wrong when messages had multiple recipients, and made Exim's
1878 behaviour dependent on the order in which the addresses were given.
1880 I have moved the duplicate checking until after the routing is complete.
1881 Exim scans the addresses that are assigned to local and remote
1882 transports, and removes any duplicates. This means that more work will be
1883 done, as duplicates will always all be routed, but duplicates are
1884 presumably rare, so I don't expect this is of any significance.
1886 For deliveries to pipes, files, and autoreplies, the duplicate checking
1887 still happens during the routing process, since they are not going to be
1890 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1891 It corrects a timeout issue with spamd. This is Ian's comment: "The
1892 background is that sometimes spamd either never reads data from a
1893 connection it has accepted, or it never writes response data. The exiscan
1894 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1895 blindly assumes that writes won't block so it may never time out."
1897 PH/39 Allow G after quota size as well as K and M.
1899 PH/40 The value set for $authenticated_id in an authenticator may not contain
1900 binary zeroes or newlines because the value is written to log lines and
1901 to spool files. There was no check on this. Now the value is run through
1902 the string_printing() function so that such characters are converted to
1903 printable escape sequences.
1905 PH/41 $message_linecount is a new variable that contains the total number of
1906 lines in the message. Compare $body_linecount, which is the count for the
1909 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1910 bounce and delay warning messages, except in certain special cases, which
1913 (a) An SMTP error message from a remote host;
1914 (b) A message specified in a :fail: redirection;
1915 (c) A message specified in a "fail" command in a system filter;
1916 (d) A message specified in a FAIL return from the queryprogram router;
1917 (e) A message specified by the cannot_route_message router option.
1919 In these cases only, Exim does include the error details in bounce and
1920 warning messages. There are also a few cases where bland messages such
1921 as "unrouteable address" or "local delivery error" are given.
1923 PH/43 $value is now also set for the "else" part of a ${run expansion.
1925 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1926 being worked on, but at least Exim now implements the latest version to
1929 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1930 process to complete was treated as a delivery failure, a timeout while
1931 writing the message to the pipe was logged, but erroneously treated as a
1932 successful delivery. Such timeouts include transport filter timeouts. For
1933 consistency with the overall process timeout, these timeouts are now
1934 treated as errors, giving rise to delivery failures by default. However,
1935 there is now a new Boolean option for the pipe transport called
1936 timeout_defer, which, if set TRUE, converts the failures into defers for
1937 both kinds of timeout. A transport filter timeout is now identified in
1940 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1941 systems where "make" and "gmake" are different, calling "gmake" at top
1942 level broke things. I've arranged for the value of $(MAKE) to be passed
1943 from the Makefile to this script so that it can call the same version of
1947 A note about Exim versions 4.44 and 4.50
1948 ----------------------------------------
1950 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1951 changes of various kinds. As a consequence, a big documentation update was
1952 needed. This delayed the release for rather longer than seemed good, especially
1953 in the light of a couple of (minor) security issues. Therefore, the changes
1954 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1955 release. So 4.44 and 4.50 are in effect two different branches that both start
1958 I have left the 4.50 change log unchanged; it contains all the changes since
1959 4.43. The change log for 4.44 is below; many of its items are identical to
1960 those for 4.50. This seems to be the most sensible way to preserve the
1961 historical information.
1967 1. Minor wording change to the doc/README.SIEVE file.
1969 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1970 computation of the current number of files was incorrect.
1972 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1973 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1976 4. Give more explanation in the error message when the command for a transport
1977 filter fails to execute.
1979 5. There are several places where Exim runs a non-Exim command in a
1980 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1981 was being done only for the command run by the queryprogram router. It is
1982 now done for all such subprocesses. The other cases are: ${run, transport
1983 filters, and the commands run by the lmtp and pipe transports.
1985 6. Added CONFIGURE_GROUP build-time option.
1987 7. Some older OS have a limit of 256 on the maximum number of file
1988 descriptors. Exim was using setrlimit() to set 1000 as a large value
1989 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1990 systems. I've change it so that if it can't get 1000, it tries for 256.
1992 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1993 was an oversight, and furthermore, ever since the addition of extra
1994 controls (e.g. 4.43/32), the checks on when to allow different forms of
1995 "control" were broken. There should now be diagnostics for all cases when a
1996 control that does not make sense is encountered.
1998 9. Added the /retain_sender option to "control=submission".
2000 10. $recipients is now available in the predata ACL (oversight).
2002 11. Tidy the search cache before the fork to do a delivery from a message
2003 received from the command line. Otherwise the child will trigger a lookup
2004 failure and thereby defer the delivery if it tries to use (for example) a
2005 cached ldap connection that the parent has called unbind on.
2007 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2008 of $address_data from the recipient verification was clobbered by the
2009 sender verification.
2011 13. The value of address_data from a sender verification is now available in
2012 $sender_address_data in subsequent conditions in the ACL statement.
2014 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2016 15. Added a new option "connect=<time>" to callout options, to set a different
2019 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2020 was its contents. (It was OK if the option was not defined at all.)
2022 17. A "Completed" log line is now written for messages that are removed from
2023 the spool by the -Mrm option.
2025 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2026 information about exactly what failed.
2028 19. Added -dd to debug only the daemon process.
2030 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2031 handles timeouts, both on the server side and network timeouts. Renamed the
2032 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2034 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2035 It is now set to "smtps".
2037 22. $host_address is now set to the target address during the checking of
2038 ignore_target_hosts.
2040 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2041 being passed; this would have caused $sender_host_name to have been used if
2042 matching the list had actually called for a host name (not very likely,
2043 since this list is usually IP addresses). A host name is now passed as
2046 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2047 code that specifies a non-restarting handler (typically sigaction() in
2048 modern systems) in an attempt to fix a rare and obscure crash bug.
2050 25. Narrowed the window for a race in the daemon that could cause it to ignore
2051 SIGCHLD signals. This is not a major problem, because they are used only to
2052 wake it up if nothing else does.
2054 26. A malformed maildirsize file could cause Exim to calculate negative values
2055 for the mailbox size or file count. Odd effects could occur as a result.
2056 The maildirsize information is now recalculated if the size or filecount
2059 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2060 support for a long time. Removed HAVE_SYS_VFS_H.
2062 28. Installed the latest version of exipick from John Jetmore.
2064 29. In an address list, if the pattern was not a regular expression, an empty
2065 subject address (from a bounce message) matched only if the pattern was an
2066 empty string. Non-empty patterns were not even tested. This was the wrong
2067 because it is perfectly reasonable to use an empty address as part of a
2068 database query. An empty address is now tested by patterns that are
2069 lookups. However, all the other forms of pattern expect the subject to
2070 contain a local part and a domain, and therefore, for them, an empty
2071 address still always fails if the pattern is not itself empty.
2073 30. Exim went into a mad DNS loop when attempting to do a callout where the
2074 host was specified on an smtp transport, and looking it up yielded more
2075 than one IP address.
2077 31. Re-factored the code for checking spool and log partition space into a
2078 function that finds that data and another that does the check. The former
2079 is then used to implement four new variables: $spool_space, $log_space,
2080 $spool_inodes, and $log_inodes.
2082 32. The RFC2047 encoding function was originally intended for short strings
2083 such as real names; it was not keeping to the 75-character limit for
2084 encoded words that the RFC imposes. It now respects the limit, and
2085 generates multiple encoded words if necessary. To be on the safe side, I
2086 have increased the buffer size for the ${rfc2047: expansion operator from
2089 33. It is now permitted to omit both strings after an "if" condition; if the
2090 condition is true, the result is "true". As before, when the second string
2091 is omitted, a false condition yields an empty string. This makes it less
2092 cumbersome to write custom ACL and router conditions.
2094 34. Failure to deliver a bounce message always caused it to be frozen, even if
2095 there was an errors_to setting on the router. The errors_to setting is now
2098 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2099 canonical form (fully expanded) before being placed in
2100 $sender_host_address.
2102 36. The table in the code that translates DNS record types into text (T_A to
2103 "A" for instance) was missing entries for NS and CNAME. It is just possible
2104 that this could have caused confusion if both these types were looked up
2105 for the same domain, because the text type is used as part of Exim's
2106 per-process caching. But the chance of anyone hitting this buglet seems
2109 37. The dnsdb lookup has been extended in a number of ways.
2111 (1) There is a new type, "zns", which walks up the domain tree until it
2112 finds some nameserver records. It should be used with care.
2114 (2) There is a new type, "mxh", which is like "mx" except that it returns
2115 just the host names, not the priorities.
2117 (3) It is now possible to give a list of domains (or IP addresses) to be
2118 looked up. The behaviour when one of the lookups defers can be
2119 controlled by a keyword.
2121 (4) It is now possible to specify the separator character for use when
2122 multiple records are returned.
2124 38. The dnslists ACL condition has been extended: it is now possible to supply
2125 a list of IP addresses and/or domains to be looked up in a particular DNS
2128 39. Added log_selector=+queue_time_overall.
2130 40. When running the queue in the test harness, wait just a tad after forking a
2131 delivery process, to get repeatability of debugging output.
2133 41. Include certificate and key file names in error message when GnuTLS fails
2134 to set them up, because the GnuTLS error message doesn't include the name
2135 of the failing file when there is a problem reading it.
2137 42. Allow both -bf and -bF in the same test run.
2139 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2141 44. The "Exiscan patch" is now merged into the mainline Exim source.
2143 45. Sometimes the final signoff response after QUIT could fail to get
2144 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2145 < 0 before doing a fflush(). This bug looks as though it goes back to the
2146 introduction of TLS in release 3.20, but "sometimes" must have been rare
2147 because the tests only now provoked it.
2149 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2150 (this can affect the format of dates).
2152 47. exim_tidydb, when checking for the continued existence of a message for
2153 which it has found a message-specific retry record, was not finding
2154 messages that were in split spool directories. Consequently, it was
2155 deleting retry records that should have stayed in existence.
2157 48. Steve fixed some bugs in eximstats.
2159 49. The SPA authentication driver was not abandoning authentication and moving
2160 on to the next authenticator when an expansion was forced to fail,
2161 contradicting the general specification for all authenticators. Instead it
2162 was generating a temporary error. It now behaves as specified.
2164 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2165 (the order specifies the preference for clients). The order is now AES256,
2166 AES128, 3DES, ARCFOUR128.
2168 51. Small patch to Sieve code - explicitly set From: when generating an
2171 52. Exim crashed if a remote delivery caused a very long error message to be
2172 recorded - for instance if somebody sent an entire SpamAssassin report back
2173 as a large number of 550 error lines. This bug was coincidentally fixed by
2174 increasing the size of one of Exim's internal buffers (big_buffer) that
2175 happened as part of the Exiscan merge. However, to be on the safe side, I
2176 have made the code more robust (and fixed the comments that describe what
2179 53. Now that there can be additional text after "Completed" in log lines (if
2180 the queue_time_overall log selector is set), a one-byte patch to exigrep
2181 was needed to allow it to recognize "Completed" as not the last thing in
2184 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2185 patch that reportedly fixes this has been added. I am not expert enough to
2186 create a test for it. This is what the patch creator wrote:
2188 "I found a little strange behaviour of ldap code when working with
2189 Windows 2003 AD Domain, where users was placed in more than one
2190 Organization Units. When I tried to give exim partial DN, the exit code
2191 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2192 But simultaneously result of request was absolutely normal ldap result,
2193 so I produce this patch..."
2195 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2196 REFERENCE, so I have modified the code to exclude the patch when that macro
2199 55. Some experimental protocols are using DNS PTR records for new purposes. The
2200 keys for these records are domain names, not reversed IP addresses. The
2201 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2202 leaves it alone. Component reversal etc. now happens only for IP addresses.
2205 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2207 57. Double the size of the debug message buffer (to 2048) so that more of very
2208 long debug lines gets shown.
2210 58. The exicyclog utility now does better if the number of log files to keep
2211 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2213 59. Two changes related to the smtp_active_hostname option:
2215 (1) $smtp_active_hostname is now available as a variable.
2216 (2) The default for smtp_banner uses $smtp_active_hostname instead
2217 of $primary_hostname.
2219 60. The host_aton() function is supposed to be passed a string that is known
2220 to be a valid IP address. However, in the case of IPv6 addresses, it was
2221 not checking this. This is a hostage to fortune. Exim now panics and dies
2222 if the condition is not met. A case was found where this could be provoked
2223 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2224 components; fortuitously, this particular loophole had already been fixed
2225 by change 4.50/55 above.
2227 If there are any other similar loopholes, the new check in host_aton()
2228 itself should stop them being exploited. The report I received stated that
2229 data on the command line could provoke the exploit when Exim was running as
2230 exim, but did not say which command line option was involved. All I could
2231 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2232 running as the user.
2235 61. There was a buffer overflow vulnerability in the SPA authentication code
2236 (which came originally from the Samba project). I have added a test to the
2237 spa_base64_to_bits() function which I hope fixes it.
2240 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2241 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2243 63. The daemon start-up calls getloadavg() while still root for those OS that
2244 need the first call to be done as root, but it missed one case: when
2245 deliver_queue_load_max is set with deliver_drop_privilege. This is
2246 necessary for the benefit of the queue runner, because there is no re-exec
2247 when deliver_drop_privilege is set.
2249 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2250 This has been fixed.
2252 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2253 was in use, was not putting the data itself into the right store pool;
2254 consequently, it could be overwritten for a subsequent message in the same
2255 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2258 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2260 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2261 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2262 the function were treating the return as a boolean value, which happened to
2263 work because 0=false and not-0=true, but is not correct code.
2265 68. The host_aton() function was not handling scoped IPv6 addresses (those
2266 with, for example, "%eth0" on the end) correctly.
2268 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2269 negated items (that is, ~something) in unsigned ints. Some compilers
2270 apparently mutter when there is no cast.
2272 70. If an address verification called from an ACL failed, and did not produce a
2273 user-specific message (i.e. there was only a "system" message), nothing was
2274 put in $acl_verify_message. In this situation, it now puts the system
2277 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2278 session; change 4.31/43 added the unwanted input to the log line - except
2279 that it did not do this in the start of session case. It now does.
2281 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2282 This is wrong and can cause the other end to generate a synchronization
2283 error if it is another Exim or anything else that does the synchronization
2284 check. A QUIT command is no longer sent after a timeout.
2286 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2287 during host lookups.
2289 74. The defer_ok option of callout verification was not working if it was used
2290 when verifying addresses in header lines, that is, for this case:
2292 verify = header_sender/callout=defer_ok
2294 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2295 those file descriptors could be used for SMTP connections. If anything
2296 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2297 could be sent to the SMTP client, causing chaos. The daemon now opens
2298 stdin, stdout, and stderr to /dev/null when it puts itself into the
2301 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2302 log by default. The user can override this with suitable Perl magic.
2304 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2305 the log message when discard triggers, was not working for the DATA ACL or
2306 for the non-SMTP ACL.
2308 78. Error message wording change in sieve.c.
2310 79. If smtp_accept_max_per_host was set, the number of connections could be
2311 restricted to fewer than expected, because the daemon was trying to set up
2312 a new connection before checking whether the processes handling previous
2313 connections had finished. The check for completed processes is now done
2314 earlier. On busy systems, this bug wouldn't be noticed because something
2315 else would have woken the daemon, and it would have reaped the completed
2318 80. If a message was submitted locally by a user whose login name contained one
2319 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2320 It caused a spool format error. I have fixed the spool reading code. A
2321 related problem was that the "from" clause in the Received: line became
2322 illegal because of the space(s). It is now covered by ${quote_local_part.
2324 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2327 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2329 83. Patch from Sieve maintainer to fix unterminated string problem in
2330 "vacation" handling.
2332 84. Some minor changes to the Linux configuration files to help with other
2333 OS variants using glibc.
2335 85. One more patch for Sieve to update vacation handling to latest spec.
2338 ----------------------------------------------------
2339 See the note above about the 4.44 and 4.50 releases.
2340 ----------------------------------------------------
2346 1. Change 4.43/35 introduced a bug that caused file counts to be
2347 incorrectly computed when quota_filecount was set in an appendfile
2350 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2351 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2354 3. Give more explanation in the error message when the command for a transport
2355 filter fails to execute.
2357 4. There are several places where Exim runs a non-Exim command in a
2358 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2359 was being done only for the command run by the queryprogram router. It is
2360 now done for all such subprocesses. The other cases are: ${run, transport
2361 filters, and the commands run by the lmtp and pipe transports.
2363 5. Some older OS have a limit of 256 on the maximum number of file
2364 descriptors. Exim was using setrlimit() to set 1000 as a large value
2365 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2366 systems. I've change it so that if it can't get 1000, it tries for 256.
2368 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2369 was an oversight, and furthermore, ever since the addition of extra
2370 controls (e.g. 4.43/32), the checks on when to allow different forms of
2371 "control" were broken. There should now be diagnostics for all cases when a
2372 control that does not make sense is encountered.
2374 7. $recipients is now available in the predata ACL (oversight).
2376 8. Tidy the search cache before the fork to do a delivery from a message
2377 received from the command line. Otherwise the child will trigger a lookup
2378 failure and thereby defer the delivery if it tries to use (for example) a
2379 cached ldap connection that the parent has called unbind on.
2381 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2382 of $address_data from the recipient verification was clobbered by the
2383 sender verification.
2385 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2386 was its contents. (It was OK if the option was not defined at all.)
2388 11. A "Completed" log line is now written for messages that are removed from
2389 the spool by the -Mrm option.
2391 12. $host_address is now set to the target address during the checking of
2392 ignore_target_hosts.
2394 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2395 being passed; this would have caused $sender_host_name to have been used if
2396 matching the list had actually called for a host name (not very likely,
2397 since this list is usually IP addresses). A host name is now passed as
2400 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2401 code that specifies a non-restarting handler (typically sigaction() in
2402 modern systems) in an attempt to fix a rare and obscure crash bug.
2404 15. Narrowed the window for a race in the daemon that could cause it to ignore
2405 SIGCHLD signals. This is not a major problem, because they are used only to
2406 wake it up if nothing else does.
2408 16. A malformed maildirsize file could cause Exim to calculate negative values
2409 for the mailbox size or file count. Odd effects could occur as a result.
2410 The maildirsize information is now recalculated if the size or filecount
2413 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2414 support for a long time. Removed HAVE_SYS_VFS_H.
2416 18. Updated exipick to current release from John Jetmore.
2418 19. Allow an empty sender to be matched against a lookup in an address list.
2419 Previously the only cases considered were a regular expression, or an
2422 20. Exim went into a mad DNS lookup loop when doing a callout where the
2423 host was specified on the transport, if the DNS lookup yielded more than
2426 21. The RFC2047 encoding function was originally intended for short strings
2427 such as real names; it was not keeping to the 75-character limit for
2428 encoded words that the RFC imposes. It now respects the limit, and
2429 generates multiple encoded words if necessary. To be on the safe side, I
2430 have increased the buffer size for the ${rfc2047: expansion operator from
2433 22. Failure to deliver a bounce message always caused it to be frozen, even if
2434 there was an errors_to setting on the router. The errors_to setting is now
2437 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2438 canonical form (fully expanded) before being placed in
2439 $sender_host_address.
2441 24. Updated eximstats to version 1.33
2443 25. Include certificate and key file names in error message when GnuTLS fails
2444 to set them up, because the GnuTLS error message doesn't include the name
2445 of the failing file when there is a problem reading it.
2447 26. Expand error message when OpenSSL has problems setting up cert/key files.
2450 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2451 (this can affect the format of dates).
2453 28. exim_tidydb, when checking for the continued existence of a message for
2454 which it has found a message-specific retry record, was not finding
2455 messages that were in split spool directories. Consequently, it was
2456 deleting retry records that should have stayed in existence.
2458 29. eximstats updated to version 1.35
2459 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2460 1.35 - bugfix such that pie charts by volume are generated correctly
2462 30. The SPA authentication driver was not abandoning authentication and moving
2463 on to the next authenticator when an expansion was forced to fail,
2464 contradicting the general specification for all authenticators. Instead it
2465 was generating a temporary error. It now behaves as specified.
2467 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2468 (the order specifies the preference for clients). The order is now AES256,
2469 AES128, 3DES, ARCFOUR128.
2471 31. Small patch to Sieve code - explicitly set From: when generating an
2474 32. Exim crashed if a remote delivery caused a very long error message to be
2475 recorded - for instance if somebody sent an entire SpamAssassin report back
2476 as a large number of 550 error lines. This bug was coincidentally fixed by
2477 increasing the size of one of Exim's internal buffers (big_buffer) that
2478 happened as part of the Exiscan merge. However, to be on the safe side, I
2479 have made the code more robust (and fixed the comments that describe what
2482 33. Some experimental protocols are using DNS PTR records for new purposes. The
2483 keys for these records are domain names, not reversed IP addresses. The
2484 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2485 leaves it alone. Component reversal etc. now happens only for IP addresses.
2488 34. The host_aton() function is supposed to be passed a string that is known
2489 to be a valid IP address. However, in the case of IPv6 addresses, it was
2490 not checking this. This is a hostage to fortune. Exim now panics and dies
2491 if the condition is not met. A case was found where this could be provoked
2492 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2493 components; fortuitously, this particular loophole had already been fixed
2494 by change 4.50/55 or 4.44/33 above.
2496 If there are any other similar loopholes, the new check in host_aton()
2497 itself should stop them being exploited. The report I received stated that
2498 data on the command line could provoke the exploit when Exim was running as
2499 exim, but did not say which command line option was involved. All I could
2500 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2501 running as the user.
2504 35. There was a buffer overflow vulnerability in the SPA authentication code
2505 (which came originally from the Samba project). I have added a test to the
2506 spa_base64_to_bits() function which I hope fixes it.
2509 36. The daemon start-up calls getloadavg() while still root for those OS that
2510 need the first call to be done as root, but it missed one case: when
2511 deliver_queue_load_max is set with deliver_drop_privilege. This is
2512 necessary for the benefit of the queue runner, because there is no re-exec
2513 when deliver_drop_privilege is set.
2515 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2516 was in use, was not putting the data itself into the right store pool;
2517 consequently, it could be overwritten for a subsequent message in the same
2518 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2521 38. Sometimes the final signoff response after QUIT could fail to get
2522 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2523 < 0 before doing a fflush(). This bug looks as though it goes back to the
2524 introduction of TLS in release 3.20, but "sometimes" must have been rare
2525 because the tests only now provoked it.
2531 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2532 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2533 Now it returns an integer. A number of calls were still expecting a T/F
2534 return. Fortuitously, in all cases, the tests worked in OK situations,
2535 which is the norm. However, things would have gone wrong on any write
2536 failures on the smtp file descriptor. This function is used when sending
2537 messages over SMTP and also when doing verify callouts.
2539 2. When Exim is called to do synchronous delivery of a locally submitted
2540 message (the -odf or -odi options), it no longer closes stderr before doing
2543 3. Implemented the mua_wrapper option.
2545 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2547 5. Implemented the functions header_remove(), header_testname(),
2548 header_add_at_position(), and receive_remove_recipient(), and exported them
2551 6. If an ACL "warn" statement specified the addition of headers, Exim already
2552 inserted X-ACL-Warn: at the start if there was no header name. However, it
2553 was not making this test for the second and subsequent header lines if
2554 there were newlines in the string. This meant that an invalid header could
2555 be inserted if Exim was badly configured.
2557 7. Allow an ACL "warn" statement to add header lines at the start or after all
2558 the Received: headers, as well as at the end.
2560 8. Added the rcpt_4xx retry error code.
2562 9. Added postmaster_mailfrom=xxx to callout verification option.
2564 10. Added mailfrom=xxxx to the callout verification option, for verify=
2567 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2568 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2570 12. Inserted some casts to stop certain compilers warning when using pointer
2571 differences as field lengths or precisions in printf-type calls (mostly
2572 affecting debugging statements).
2574 13. Added optional readline() support for -be (dynamically loaded).
2576 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2577 same clock tick as a message's arrival, so that its received time was the
2578 same as the "first fail" time on the retry record, and that message
2579 remained on the queue past the ultimate address timeout, every queue runner
2580 would try a delivery (because it was past the ultimate address timeout) but
2581 after another failure, the ultimate address timeout, which should have then
2582 bounced the address, did not kick in. This was a "< instead of <=" error;
2583 in most cases the first failure would have been in the next clock tick
2584 after the received time, and all would be well.
2586 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2587 being recognized when the domain list was tested by the match_domain
2588 condition in an expansion string.
2590 16. Added the ${str2b64: operator.
2592 17. Exim was always calling setrlimit() to set a large limit for the number of
2593 processes, without checking whether the existing limit was already
2594 adequate. (It did check for the limit on file descriptors.) Furthermore,
2595 errors from getrlimit() and setrlimit() were being ignored. Now they are
2596 logged to the main and panic logs, but Exim does carry on, to try to do its
2597 job under whatever limits there are.
2599 18. Imported PCRE 5.0.
2601 19. Trivial typo in log message " temporarily refused connection" (the leading
2604 20. If the log selector return_path_on_delivery was set and an address was
2605 redirected to /dev/null, the delivery process crashed because it assumed
2606 that a return path would always be set for a "successful" delivery. In this
2607 case, the whole delivery is bypassed as an optimization, and therefore no
2610 21. Internal re-arrangement: the function for sending a challenge and reading
2611 a response while authentication was assuming a zero-terminated challenge
2612 string. It's now changed to take a pointer and a length, to allow for
2613 binary data in such strings.
2615 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2617 23. Exim was not respecting finduser_retries when seeking the login of the
2618 uid under which it was called; it was always trying 10 times. (The default
2619 setting of finduser_retries is zero.) Also, it was sleeping after the final
2620 failure, which is pointless.
2622 24. Implemented tls_on_connect_ports.
2624 25. Implemented acl_smtp_predata.
2626 26. If the domain in control=submission is set empty, Exim assumes that the
2627 authenticated id is a complete email address when it generates From: or
2628 Sender: header lines.
2630 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2631 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2632 chown and chgrp in /bin and hostname in /usr/bin.
2634 28. Exim was keeping the "process log" file open after each use, just as it
2635 does for the main log. This opens the possibility of it remaining open for
2636 long periods when the USR1 signal hits a daemon. Occasional processlog
2637 errors were reported, that could have been caused by this. Anyway, it seems
2638 much more sensible not to leave this file open at all, so that is what now
2641 29. The long-running daemon process does not normally write to the log once it
2642 has entered its main loop, and it closes the log before doing so. This is
2643 so that log files can straightforwardly be renamed and moved. However,
2644 there are a couple of unusual error situations where the daemon does write
2645 log entries, and I had neglected to close the log afterwards.
2647 30. The text of an SMTP error response that was received during a remote
2648 delivery was being truncated at 512 bytes. This is too short for some of
2649 the long messages that one sometimes sees. I've increased the limit to
2652 31. It is now possible to make retry rules that apply only when a message has a
2653 specific sender, in particular, an empty sender.
2655 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2656 it possible to be selective about when SMTP synchronization is enforced.
2658 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2660 32. Implemented hosts_connection_nolog.
2662 33. Added an ACL for QUIT.
2664 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2667 35. Added mailbox_size and mailbox_filecount to appendfile.
2669 36. Added control = no_multiline_responses to ACLs.
2671 37. There was a bug in the logic of the code that waits for the clock to tick
2672 in the case where the clock went backwards by a substantial amount such
2673 that the microsecond fraction of "now" was more than the microsecond
2674 fraction of "then" (but the whole seconds number was less).
2676 38. Added support for the libradius Radius client library this is found on
2677 FreeBSD (previously only the radiusclient library was supported).
2683 1. When certain lookups returned multiple values in the form name=value, the
2684 quoting of the values was not always being done properly. Specifically:
2685 (a) If the value started with a double quote, but contained no whitespace,
2687 (b) If the value contained whitespace other than a space character (i.e.
2688 tabs or newlines or carriage returns) it was not quoted.
2689 This fix has been applied to the mysql and pgsql lookups by writing a
2690 separate quoting function and calling it from the lookup code. The fix
2691 should probably also be applied to nisplus, ibase and oracle lookups, but
2692 since I cannot test any of those, I have not disturbed their existing code.
2694 2. A hit in the callout cache for a specific address caused a log line with no
2695 reason for rejecting RCPT. Now it says "Previous (cached) callout
2696 verification failure".
2698 3. There was an off-by-one bug in the queryprogram router. An over-long
2699 return line was truncated at 256 instead of 255 characters, thereby
2700 overflowing its buffer with the terminating zero. As well as fixing this, I
2701 have increased the buffer size to 1024 (and made a note to document this).
2703 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2704 when Exim is waiting for an SMTP response from a remote server, Exim
2705 restarts its select() call on the socket, thereby resetting its timeout.
2706 This is not a problem when such interrupts are rare. Somebody set up a cron
2707 job to run exiwhat every 2 minutes, which is less than the normal select()
2708 timeout (5 or 10 minutes). This meant that the select() timeout never
2709 kicked in because it was always reset. I have fixed this by comparing the
2710 time when an interrupt arrives with the time at the start of the first call
2711 to select(). If more time than the timeout has elapsed, the interrupt is
2712 treated as a timeout.
2714 5. Some internal re-factoring in preparation for the addition of Sieve
2715 extensions (by MH). In particular, the "personal" test is moved to a
2716 separate function, and given an option for scanning Cc: and Bcc: (which is
2717 not set for Exim filters).
2719 6. When Exim created an email address using the login of the caller as the
2720 local part (e.g. when creating a From: or Sender: header line), it was not
2721 quoting the local part when it contained special characters such as @.
2723 7. Installed new OpenBSD configuration files.
2725 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2726 try to make them clearer.
2728 9. Callout options, other than the timeout value, were being ignored when
2729 verifying sender addresses in header lines. For example, when using
2731 verify = header_sender/callout=no_cache
2733 the cache was (incorrectly) being used.
2735 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2736 only the Cygwin environment.
2738 11. When return_path_on_delivery was set as a log selector, if different remote
2739 addresses in the same message used different return paths and parallel
2740 remote delivery occurred, the wrong values would sometimes be logged.
2741 (Whenever a remote delivery process finished, the return path value from
2742 the most recently started remote delivery process was logged.)
2744 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2745 lines when AUTH and/or TLS are in use. This is the "received protocol"
2746 field. Exim used to use "asmtp" for authenticated SMTP, without any
2747 indication (in the protocol name) for TLS use. Now it follows the RFC and
2748 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2749 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2750 names appear in log lines as well as in Received: header lines.
2752 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2753 extensions, and comparison tests, and to fix some bugs.
2755 14. Changes to the "personal" filter test:
2757 (1) The test was buggy in that it was just doing the equivalent of
2758 "contains" tests on header lines. For example, if a user's address was
2759 anne@some.where, the "personal" test would incorrectly be true for
2761 To: susanne@some.where
2763 This test is now done by extracting each address from the header in turn,
2764 and checking the entire address. Other tests that are part of "personal"
2765 are now done using regular expressions (for example, to check local parts
2766 of addresses in From: header lines).
2768 (2) The list of non-personal local parts in From: addresses has been
2769 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2770 taken from the Sieve specification recommendations.
2772 (3) If the message contains any header line starting with "List-" it is
2773 treated as non-personal.
2775 (4) The test for "circular" in the Subject: header line has been removed
2776 because it now seems ill-conceived.
2778 15. Minor typos in src/EDITME comments corrected.
2780 16. Installed latest exipick from John Jetmore.
2782 17. If headers_add on a router specified a text string that was too long for
2783 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2784 of string_sprintf() is now avoided.
2786 18. $message_body_size was not set (it was always zero) when running the DATA
2787 ACL and the local_scan() function.
2789 19. For the "mail" command in an Exim filter, no default was being set for
2790 the once_repeat time, causing a random time value to be used if "once" was
2791 specified. (If the value happened to be <= 0, no repeat happened.) The
2792 default is now 0s, meaning "never repeat". The "vacation" command was OK
2793 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2794 (I found it when inspecting the code).
2796 20. There is now an overall timeout for performing a callout verification. It
2797 defaults to 4 times the callout timeout, which applies to individual SMTP
2798 commands during the callout. The overall timeout applies when there is more
2799 than one host that can be tried. The timeout is checked before trying the
2800 next host. This prevents very long delays if there are a large number of
2801 hosts and all are timing out (e.g. when the network connections are timing
2802 out). The value of the overall timeout can be changed by specifying an
2803 additional sub-option for "callout", called "maxwait". For example:
2805 verify = sender/callout=5s,maxwait=20s
2807 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2808 to the end before writing, but this should make it even safer).
2810 22. Exim was forgetting that it had advertised PIPELINING for the second and
2811 subsequent messages on an SMTP connection. It was also not resetting its
2812 memory on STARTTLS and an internal HELO.
2814 23. When Exim logs an SMTP synchronization error within a session, it now
2815 records whether PIPELINING has been advertised or not.
2817 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2818 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2819 rather than long int.
2821 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2823 26. Added the never_mail option to autoreply.
2829 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2830 crash if the getsockname() call failed; this can happen if a connection is
2831 closed very soon after it is established. The problem was simply in the
2832 order in which certain operations were done, causing Exim to try to write
2833 to the SMTP stream before it had set up the file descriptor. The bug has
2834 been fixed by making things happen in the correct order.
2840 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2841 before the connection was closed, thus losing the rejection response.
2843 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2844 some early Solaris releases, but causes trouble in current releases where
2845 socklen_t is defined.
2847 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2850 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2853 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2854 that releasing the top of it at the end releases what was used for sub-
2855 expansions (unless the block got too big). However, discard this block if
2856 the first thing is a variable or header, so that we can use its block when
2857 it is dynamic (useful for very large $message_headers, for example).
2859 6. Lookups now cache *every* query, not just the most recent. A new, separate
2860 store pool is used for this. It can be recovered when all lookup caches are
2861 flushed. Lookups now release memory at the end of their result strings.
2862 This has involved some general refactoring of the lookup sources.
2864 7. Some code has been added to the store_xxx() functions to reduce the amount
2865 of flapping under certain conditions.
2867 8. log_incoming_interface used to affect only the <= reception log lines. Now
2868 it causes the local interface and port to be added to several more SMTP log
2869 lines, for example "SMTP connection from", and rejection lines.
2871 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2873 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2875 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2876 could be overwritten at the end of the current message (or the start of a
2877 new message if it was set in a HELO ACL). The value is now preserved for
2878 the duration of the SMTP connection.
2880 12. If a transport had a headers_rewrite setting, and a matching header line
2881 contained an unqualified address, that address was qualified, even if it
2882 did not match any rewriting rules. The underlying bug was that the values
2883 of the flags that permit the existence of unqualified sender and recipient
2884 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2885 non-local messages, and by -bnq for local messages) were not being
2886 preserved with the message after it was received.
2888 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2889 "next input=" as part of the text comprising the host identity instead of
2890 the correct text. The code was using the same buffer for two different
2891 strings. However, depending on which order the printing function evaluated
2892 its arguments, the bug did not always show up. Under Linux, for example, my
2893 test suite worked just fine.
2895 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2896 doesn't work with some older versions of Perl. It has been changed to "my",
2897 which in any case is probably the better facility to use.
2899 15. A really picky compiler found some instances of statements for creating
2900 error messages that either had too many or two few arguments for the format
2903 16. The size of the buffer for calls to the DNS resolver has been increased
2904 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2905 for addresses that have a lot of PTR records. This alleviates a problem; it
2906 does not fully solve it.
2908 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2909 buffer now truncates the list and logs the incident, which is the same
2910 action as happens when Exim is looking up a host name and its aliases.
2911 Previously in this situation something unpredictable would happen;
2912 sometimes it was "internal error: store_reset failed".
2914 18. If a server dropped the connection unexpectedly when an Exim client was
2915 using GnuTLS and trying to read a response, the client delivery process
2916 crashed while trying to generate an error log message.
2918 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2919 string, for example:
2921 warn message = H1: something\nH2: something
2923 the text was added as a single header line from Exim's point of view
2924 though it ended up OK in the delivered message. However, searching for the
2925 second and subsequent header lines using $h_h2: did not work. This has been
2926 fixed. Similarly, if a system filter added multiple headers in this way,
2927 the routers could not see them.
2929 20. Expanded the error message when iplsearch is called with an invalid key to
2930 suggest using net-iplsearch in a host list.
2932 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2933 ACLs are no longer actually imposed (and a message to that effect is
2936 22. If a "gecos" field in a passwd entry contained escaped characters, in
2937 particular, if it contained a \" sequence, Exim got it wrong when building
2938 a From: or a Sender: header from that name. A second bug also caused
2939 incorrect handling when an unquoted " was present following a character
2940 that needed quoting.
2942 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2943 was not being matched caselessly.
2945 24. Arranged for all hyphens in the exim.8 source to be escaped with
2948 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2949 properly. Recipient callout cache records were still being keyed to include
2950 the sender, even when use_sender was set false. This led to far more
2951 callouts that were necessary. The sender is no longer included in the key
2952 when use_sender is false.
2954 26. Added "control = submission" modifier to ACLs.
2956 27. Added the ${base62d: operator to decode base 62 numbers.
2958 28. dnsdb lookups can now access SRV records.
2960 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2961 the configuration file.
2963 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2964 (-v) mode. This makes the output for a verbose queue run more intelligible.
2966 31. Added a use_postmaster feature to recipient callouts.
2968 32. Added the $body_zerocount variable, containing the number of binary zero
2969 bytes in the message body.
2971 33. The time of last modification of the "new" subdirectory is now used as the
2972 "mailbox time last read" when there is a quota error for a maildir
2975 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2977 35. Added +ignore_unknown as a special item in host lists.
2979 36. Code for decoding IPv6 addresses in host lists is now included, even if
2980 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2981 address was recognized as an IP address, but was then not correctly decoded
2982 into binary, causing unexpected and incorrect effects when compared with
2989 1. Very minor rewording of debugging text in manualroute to say "list of
2990 hosts" instead of "hostlist".
2992 2. If verify=header_syntax was set, and a header line with an unqualified
2993 address (no domain) and a large number of spaces between the end of the
2994 name and the colon was received, the reception process suffered a buffer
2995 overflow, and (when I tested it) crashed. This was caused by some obsolete
2996 code that should have been removed. The fix is to remove it!
2998 3. When running in the test harness, delay a bit after writing a bounce
2999 message to get a bit more predictability in the log output.
3001 4. Added a call to search_tidyup() just before forking a reception process. In
3002 theory, someone could use a lookup in the expansion of smtp_accept_max_
3003 per_host which, without the tidyup, could leave open a database connection.
3005 5. Added the variables $recipient_data and $sender_data which get set from a
3006 lookup success in an ACL "recipients" or "senders" condition, or a router
3007 "senders" option, similar to $domain_data and $local_part_data.
3009 6. Moved the writing of debug_print from before to after the "senders" test
3012 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3013 problems for message scanning, either using a data ACL, or using
3014 local_scan() because the Received: header was not generated till after they
3015 were called (in order to set the time as the time of reception completion).
3016 I have revised the way this works. The header is now generated after the
3017 body is received, but before the ACL or local_scan() are called. After they
3018 are run, the timestamp in the header is updated.
3024 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3025 before starting a queue runner without re-exec. This happened only when
3026 deliver_drop_privilege was set or when the Exim user was set to root. The
3027 effect of the bug was that timeouts during subsequent deliveries caused
3028 crashes instead of being properly handled. The handler is now left at its
3029 default (and expected) setting.
3031 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3032 message, again when deliver_drop_privilege is set or Exim is run as root.
3033 The bug described in (1) was not present in this case, but the tidying up
3034 of the other signals was missing. I have made the two cases consistent.
3036 3. The ignore_target_hosts setting on a manualroute router was being ignored
3037 for hosts that were looked up using the /MX notation.
3039 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3042 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3043 operated on the sender address. After changing the $sender_address to <>
3044 for the sender address verify, Exim was re-instated it as the original
3045 (before rewriting) address, but remembering that it had rewritten it, so it
3046 wasn't rewriting it again. This bug also had the effect of breaking the
3047 sender address verification caching when the sender address was rewritten.
3049 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3050 This has been changed so that if the ip literal address matches
3051 ignore_target_hosts, the router declines.
3053 7. Added expansion conditions match_domain, match_address, and match_local_
3054 part (NOT match_host).
3056 8. The placeholder for the Received: header didn't have a length field set.
3058 9. Added code to Exim itself and to exim_lock to test for a specific race
3059 condition that could lead to file corruption when using MBX delivery. The
3060 issue is with the lockfile that is created in /tmp. If this file is removed
3061 after a process has opened it but before that process has acquired a lock,
3062 there is the potential for a second process to recreate the file and also
3063 acquire a lock. This could lead to two Exim processes writing to the file
3064 at the same time. The added code performs the same test as UW imapd; it
3065 checks after acquiring the lock that its file descriptor still refers to
3066 the same named file.
3068 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3069 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3072 11. Added the smtp_active_hostname option. If used, this will typically be made
3073 to depend on the incoming interface address. Because $interface_address is
3074 not set up until the daemon has forked a reception process, error responses
3075 that can happen earlier (such as "too many connections") no longer contain
3078 12. If an expansion in a condition on a "warn" statement fails because a lookup
3079 defers, the "warn" statement is abandoned, and the next ACL statement is
3080 processed. Previously this caused the whole ACL to be aborted.
3082 13. Added the iplsearch lookup type.
3084 14. Added ident_timeout as a log selector.
3086 15. Added tls_certificate_verified as a log selector.
3088 16. Added a global option tls_require_ciphers (compare the smtp transport
3089 option of the same name). This controls incoming TLS connections.
3091 17. I finally figured out how to make tls_require_ciphers do a similar thing
3092 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3093 before starting the TLS session.
3095 18. Tabs are now shown as \t in -bP output.
3097 19. If the log selector return_path_on_delivery was set, Exim crashed when
3098 bouncing a message because it had too many Received: header lines.
3100 20. If two routers both had headers_remove settings, and the first one included
3101 a superfluous trailing colon, the final name in the first list and the
3102 first name in the second list were incorrectly joined into one item (with a
3103 colon in the middle).
3109 1. Added -C and -D options to the exinext utility, mainly to make it easier
3110 to include in the automated testing, but these could be helpful when
3111 multiple configurations are in use.
3113 2. The exinext utility was not formatting the output nicely when there was
3114 an alternate port involved in the retry record key, nor when there was a
3115 message id as well (for retries that were specific to a specific message
3116 and a specific host). It was also confused by IPv6 addresses, because of
3117 the additional colons they contain. I have fixed the IPv4 problem, and
3118 patched it up to do a reasonable job for IPv6.
3120 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3121 delivery, the log line now contains "pipelined" if PIPELINING was used.
3123 4. An SMTP transport process used to panic and die if the bind() call to set
3124 an explicit outgoing interface failed. This has been changed; it is now
3125 treated in the same way as a connect() failure.
3127 5. A reference to $sender_host_name in the part of a conditional expansion
3128 that was being skipped was still causing a DNS lookup. This no longer
3131 6. The def: expansion condition was not recognizing references to header lines
3132 that used bh_ and bheader_.
3134 7. Added the _cache feature to named lists.
3136 8. The code for checking quota_filecount in the appendfile transport was
3137 allowing one more file than it should have been.
3139 9. For compatibility with Sendmail, the command line option
3147 and sets the incoming protocol and host name (for trusted callers). The
3148 host name and its colon can be omitted when only the protocol is to be set.
3149 Note the Exim already has two private options, -pd and -ps, that refer to
3150 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3151 "s", but I don't think that's a major issue.
3153 10. A number of refactoring changes to the code, none of which should affect
3156 (a) The number of logging options was getting close to filling up the
3157 32-bit word that was used as a bit map. I have split them into two classes:
3158 those that are passed in the argument to log_write(), and those that are
3159 only ever tested independently outside of that function. These are now in
3160 separate 32-bit words, so there is plenty of room for expansion again.
3161 There is no change in the user interface or the logging behaviour.
3163 (b) When building, for example, log lines, the code previously used a
3164 macro that called string_cat() twice, in order to add two strings. This is
3165 not really sufficiently general. Furthermore, there was one instance where
3166 it was actually wrong because one of the argument was used twice, and in
3167 one call a function was used. (As it happened, calling the function twice
3168 did not affect the overall behaviour.) The macro has been replaced by a
3169 function that can join an arbitrary number of extra strings onto a growing
3172 (c) The code for expansion conditions now uses a table and a binary chop
3173 instead of a serial search (which was left over from when there were very
3174 few conditions). Also, it now recognizes conditions like "pam" even when
3175 the relevant support is not compiled in: a suitably worded error message is
3176 given if an attempt is made to use such a condition.
3178 11. Added ${time_interval:xxxxx}.
3180 12. A bug was causing one of the ddress fields not to be passed back correctly
3181 from remote delivery subprocesses. The field in question was not being
3182 subsequently used, so this caused to problems in practice.
3184 13. Added new log selectors queue_time and deliver_time.
3186 14. Might have fixed a bug in maildirsizefile handling that threw up
3187 "unexpected character" debug warnings, and recalculated the data
3188 unnecessarily. In any case, I expanded the warning message to give more
3191 15. Added the message "Restricted characters in address" to the statements in
3192 the default ACL that block characters like @ and % in local parts.
3194 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3195 Three changes have been made:
3197 (a) There was a serious bug; a negative response to MAIL caused the whole
3198 recipient domain to be cached as invalid, thereby blocking all messages
3199 to all local parts at the same domain, from all senders. This bug has
3200 been fixed. The domain is no longer cached after a negative response to
3201 MAIL if the sender used is not empty.
3203 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3206 (c) A new callout option, "use_sender" has been added for people who want
3207 the modified behaviour.
3213 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3216 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3217 indeed breaks things for older releases.
3219 3. Added additional logging to the case where there is a problem reading data
3220 from a filter that is running in a subprocess using a pipe, in order to
3221 try to track down a specific problem.
3223 4. Testing facility fudge: when running in the test harness and attempting
3224 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3225 getting "No route to host". Convert this to a timeout.
3227 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3230 6. Some OS don't have socklen_t but use size_t instead. This affects the
3231 fifth argument of getsockopt() amongst other things. This is now
3232 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3233 can be set for individual OS. I have set it for SunOS5, OSF1, and
3234 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3235 some earlier ones do not.
3237 7. Change 4.30/15 was not doing the test caselessly.
3239 8. The standard form for an IPv6 address literal was being rejected by address
3240 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3241 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3242 this, as well as the form without the "IPv6" on the front (but only when
3243 address literals are enabled, of course).
3245 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3247 10. Exim crashed if a message with an empty sender address specified by -f
3248 encountered a router with an errors_to setting. This could be provoked only
3249 by a command such as
3253 where an empty string was supplied; "<>" did not hit this bug.
3255 11. Installed PCRE release 4.5.
3257 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3258 remained set. It is now erased.
3260 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3261 times from message ids (which are base 36 rather than the normal 62).
3263 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3264 were being counted as actual protocol errors, and logged if the log
3265 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3266 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3267 and DATA following a set of rejected RCPTs do not count as protocol errors.
3268 In other words, Exim assumes they were pipelined, though this may not
3269 actually be the case. Of course, in all cases the client gets an
3270 appropriate error code.
3272 15. If a lookup fails in an ACL condition, a message about the failure may
3273 be available; it is used if testing the ACL cannot continue, because most
3274 such messages specify what the cause of the deferral is. However, some
3275 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3276 that caused an old message to be retained and used if a later statement
3277 caused a defer, replacing the real cause of the deferral.
3279 16. If an IP address had so many PTR records that the DNS lookup buffer
3280 was not large enough to hold them, Exim could crash while trying to process
3281 the truncated data. It now detects and logs this case.
3283 17. Further to 4.21/58, another change has been made: if (and only if) the
3284 first line of a message (the first header line) ends with CRLF, a bare LF
3285 in a subsequent header line has a space inserted after it, so as not to
3286 terminate the header.
3288 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3289 unnecessarily, used atoi() instead of strtol(), and didn't check the
3290 termination when getting file sizes from file names by regex.
3292 19. Completely re-implemented the support for maildirsize files, in the light
3293 of a number of problems with the previous contributed implementation
3294 (4.30/29). In particular:
3296 . If the quota is zero, the maildirsize file is maintained, but no quota is
3299 . If the maildir directory does not exist, it is created before any attempt
3300 to write a maildirsize file.
3302 . The quota value in the file is just a cache; if the quota is changed in
3303 the transport, the new value overrides.
3305 . A regular expression is available for excluding directories from the
3308 20. The autoreply transport checks the characters in options that define the
3309 message's headers; it allows continued headers, but it was checking with
3310 isspace() after an embedded newline instead of explicitly looking for a
3313 21. If all the "regular" hosts to which an address was routed had passed their
3314 expiry times, and had not reached their retry times, the address was
3315 bounced, even if fallback hosts were defined. Now Exim should go on to try
3318 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3319 equivalent code in the SMTP transport. Some hosts send humungous responses
3320 to HELO/EHLO, more than 1024 it seems.
3322 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3323 gives compiler warnings in some environments. I've now done it "properly",
3326 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3327 (because of problems with the built-in one) was declared to return uschar *
3328 instead of char *, causing compiler failure.
3330 25. Fixed a file descriptor leak when processing alias/forward files.
3332 26. Fixed a minor format string issue in dbfn.c.
3334 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3336 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3337 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3340 29. When a sender address is verified, it is cached, to save repeating the test
3341 when there is more than one recipient in a message. However, when the
3342 verification involves a callout, it is possible for different callout
3343 options to be set for different recipients. It is too complicated to keep
3344 track of this in the cache, so now Exim always runs a verification when a
3345 callout is required, relying on the callout cache for the optimization.
3346 The overhead is duplication of the address routing, but this should not be
3349 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3350 to be verified with callout=postmaster, and the main callout worked but the
3351 postmaster check failed, the verification correctly failed. However, if a
3352 subsequent RCPT command asked for sender verification *without* the
3353 postmaster check, incorrect caching caused this verification also to fail,
3356 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3357 it was not caching the DNS options (qualify_single, search_parents) that
3358 were used when the lookup failed. A subsequent lookup with different
3359 options therefore always gave the same answer, though there were cases
3360 where it should not have. (Example: a "domains = !$mx_any" option on a
3361 dnslookup router: the "domains" option is always processed without any
3362 widening, but the router might have qualify_single set.) Now Exim uses the
3363 cached value only when the same options are set.
3365 32. Added John Jetmore's "exipick" utility to the distribution.
3367 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3368 than a timeout (e.g. a certificate is required, and is not provided), an
3369 Exim server now closes the connection immediately. Previously it waited for
3370 the client to close - but if the client is SSL, it seems that they each
3371 wait for each other, leading to a delay before one of them times out.
3373 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3374 maintained 0.8.x compatibility because I don't think many are using it, and
3375 it is clearly obsolete.
3377 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3380 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3381 client certificate was expired. A simple patch fixes this, though I don't
3382 understand the full logic of why the verify callback is called multiple
3385 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3386 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3387 which causes problems with some clients (such as the Certicom SSL Plus
3388 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3389 disables the coutermeasure allowing Eudora to connect."
3391 38. Exim was not checking that a write() to a log file succeeded. This could
3392 lead to Bad Things if a log got too big, in particular if it hit a file
3393 size limit. Exim now panics and dies if it cannot write to a log file, just
3394 as it does if it cannot open a log file.
3396 39. Modified OS/Makefile-Linux so that it now contains
3398 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3400 The two -D definitions ensure that Exim is compiled with large file
3401 support, which makes it possible to handle log files that are bigger than
3404 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3405 instance) a domain was checked against a named list that involved a lookup,
3406 causing $domain_data to be set, then another domain was checked against the
3407 same list, then the first domain was re-checked, the value of $domain_data
3408 after the final check could be wrong. In particular, if the second check
3409 failed, it could be set empty. This bug probably also applied to
3412 41. The strip_trailing_dot option was not being applied to the address given
3413 with the -f command-line option.
3415 42. The code for reading a message's header from the spool was incrementing
3416 $received_count, but never initializing it. This meant that the value was
3417 incorrect (doubled) while delivering a message in the same process in which
3418 it was received. In the most common configuration of Exim, this never
3419 happens - a fresh exec is done - but it can happen when
3420 deliver_drop_privilege is set.
3422 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3423 it now includes up to 150 characters of the unexpected data in the log
3426 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3427 and building data strings. The size of both of these buffers was 10 000
3428 bytes - far larger than anybody would *ever* want, thought I. Needless to
3429 say, somebody hit the limit. I have increased the maximum line length to
3430 20 000 and the maximum data length of concatenated lines to 100 000. I have
3431 also fixed two bugs, because there was no checking on these buffers. Tsk,
3432 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3433 buffer is too small.
3435 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3436 lsearch lookups. Now it does.
3438 46. When parsing a route_list item in a manualroute router, a fixed-length
3439 buffer was used for the list of hosts. I made this 1024 bytes long,
3440 thinking that nobody would ever have a list of hosts that long. Wrong.
3441 Somebody had a whole pile of complicated expansion conditions, and the
3442 string was silently truncated, leading to an expansion error. It turns out
3443 that it is easier to change to an unlimited length (owing to other changes
3444 that have happened since this code was originally written) than to build
3445 structure for giving a limitation error. The length of the item that
3446 expands into the list of hosts is now unlimited.
3448 47. The lsearch lookup could not handle data where the length of text line was
3449 more than 4095 characters. Such lines were truncated, leading to shortened
3450 data being returned. It should now handle lines of any length.
3452 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3453 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3456 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3457 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3458 no longer happens in comment lines. A list of replacements is now placed
3459 at the head of all of the source files, except those whose only change is
3460 to replace PERL_COMMAND in the very first #! line.
3462 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3463 messages on the queue, with a bottom-up merge sort, using code contributed
3464 by Michael Haardt. This should make operations like -bp somewhat faster on
3465 large queues. It won't affect queue runners, except when queue_run_in_order
3468 51. Installed eximstats 1.31 in the distribution.
3470 52. Added support for SRV lookups to the dnslookup router.
3472 53. If an ACL referred to $message_body or $message_body_end, the value was not
3473 reset for any messages that followed in the same SMTP session.
3475 54. The store-handling optimization for building very long strings was not
3476 differentiating between the different store pools. I don't think this
3477 actually made any difference in practice, but I've tidied it.
3479 55. While running the routers to verify a sender address, $sender_address
3480 was still set to the sender address. This is wrong, because when routing to
3481 send a bounce to the sender, it would be empty. Therefore, I have changed
3482 it so that, while verifying a sender address, $sender_address is set to <>.
3483 (There is no change to what happens when verifying a recipient address.)
3485 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3486 target A or AAAA records (if not already returned) without resetting the
3487 qualify_single or search_parents options of the DNS resolver. These are
3488 inappropriate in this case because the targets of MX and SRV records must
3489 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3490 target that, when qualified, matched something in the local domain. These
3491 two options are now turned off when doing these lookups.
3493 57. It seems that at least some releases of Reiserfs (which does not have the
3494 concept of a fixed number of inodes) returns zero and not -1 for the
3495 number of available inodes. This interacted badly with check_spool_inodes,
3496 which assumed that -1 was the "no such thing" setting. What I have done is
3497 to check that the total number of inodes is greater than zero before doing
3498 the test of how many are available.
3500 58. When a "warn" ACL statement has a log_message modifier, the message is
3501 remembered, and not repeated. This is to avoid a lot of repetition when a
3502 message has many recipients that cause the same warning to be written.
3503 Howewer, Exim was preserving the list of already written lines for an
3504 entire SMTP session, which doesn't seem right. The memory is now reset if a
3505 new message is started.
3507 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3508 debugging output unless log_rewrite was also set.
3510 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3511 of (int)(handle) when we know that handle contains (void *)(-1).
3513 61. The Exim daemon panic-logs an error return when it closes the incoming
3514 connection. However "connection reset by peer" seems to be common, and
3515 isn't really an error worthy of noting specially, so that particular error
3518 62. When Exim is trying to find all the local interfaces, it used to panic and
3519 die if the ioctl to get the interface flags failed. However, it seems that
3520 on at least one OS (Solaris 9) it is possible to have an interface that is
3521 included in the list of interfaces, but for which you get a failure error
3522 for this call. This happens when the interface is not "plumbed" into a
3523 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3524 failure of the "get flags" call assumes that the interface is down.
3526 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3527 makes life easier for people who are doing arithmetic on fields extracted
3528 from dates, where you often get leading zeros that should not be
3529 interpreted as octal.
3531 64. Added qualify_domain to the redirect router, to override the global
3534 65. If a pathologically long header line contained very many addresses (the
3535 report of this problem mentioned 10 000) and each of them was rewritten,
3536 Exim could use up a very large amount of memory. (It kept on making new
3537 copies of the header line as it rewrote, and never released the old ones.)
3538 At the expense of a bit more processing, the header rewriting function has
3539 been changed so that it no longer eats memory in this way.
3541 66. The generation of the Received: header has been moved from the time that a
3542 message starts to be received, to the time that it finishes. The timestamp
3543 in the Received: header should now be very close to that of the <= log
3544 line. There are two side-effects of this change:
3546 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3547 logged header lines no longer include the local Received: line, because
3548 it has not yet been created. The same applies to a copy of the message
3549 that is returned to a non-SMTP sender when a message is rejected.
3551 (b) When a filter file is tested using -bf, no additional Received: header
3552 is added to the test message. After some thought, I decided that this
3555 This change does not affect the value of $received_for. It is still set
3556 after address rewriting, but before local_scan() is called.
3558 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3560 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3561 gave an unhelpful panic error message, and a defer error. I have managed to
3562 change this behaviour so that it now rejects any supplied certificate,
3563 which seems right, as the list of acceptable certificates is empty.
3565 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3566 gave an unhelpful defer error. I have not managed to make this reject any
3567 supplied certificates, but the error message it gives is "no certificate
3568 supplied", which is not helpful.
3570 70. exigrep's output now also includes lines that are not associated with any
3571 message, but which match the given pattern. Implemented by a patch from
3572 Martin Sluka, which also tidied up the Perl a bit.
3574 71. Recipient callout verification, like sender verification, was using <> in
3575 the MAIL FROM command. This isn't really the right thing, since the actual
3576 sender may affect whether the remote host accepts the recipient or not. I
3577 have changed it to use the actual sender in the callout; this means that
3578 the cache record is now keyed on a recipient/sender pair, not just the
3579 recipient address. There doesn't seem to be a real danger of callout loops,
3580 since a callout by the remote host to check the sender would use <>.
3581 [SEE ABOVE: changed after hitting problems.]
3583 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3584 temporary errors. However, in the case of such a code being given after
3585 the end of a data transmission (i.e. after ".") Exim was failing to write
3586 a retry record for the message. (Yes, there was some broken host that was
3587 actually sending 8xx at this point.)
3589 73. An unknown lookup type in a host list could cause Exim to panic-die when
3590 the list was checked. (An example that provoked this was putting <; in the
3591 middle of a list instead of at the start.) If this happened during a DATA
3592 ACL check, a -D file could be left lying around. This kind of configuration
3593 error no longer causes Exim to die; instead it causes a defer errror. The
3594 incident is still logged to the main and panic logs.
3596 74. Buglet left over from Exim 3 conversion. The message "too many messages
3597 in one connection" was written to the rejectlog but not the mainlog, except
3598 when address rewriting (yes!) was being logged.
3600 75. Added write_rejectlog option.
3602 76. When a system filter was run not as root (that is, when system_filter_user
3603 was set), the values of the $n variables were not being returned to the
3604 main process; thus, they were not subsequently available in the $sn
3607 77. Added +return_path_on_delivery log selector.
3609 78. A connection timeout was being treated differently from recipients deferred
3610 when testing hosts_max_try with a message that was older than the host's
3611 retry timeout. (The host should not be counted, thus allowing all hosts to
3612 be tried at least once before bouncing.) This may have been the cause of an
3613 occasionally reported bug whereby a message would remain on the queue
3614 longer than the retry timeout, but would be bounced if a delivery was
3615 forced. I say "may" because I never totally pinned down the problem;
3616 setting up timeout/retry tests is difficult. See also the next item.
3618 79. The ultimate address timeout was not being applied to errors that involved
3619 a combination of host plus message (for example, a timeout on a MAIL
3620 command). When an address resolved to a number of possible hosts, and they
3621 were not all tried for each delivery (e.g. because of hosts_max_try), a
3622 message could remain on the queue longer than the retry timeout.
3624 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3627 81. Fixed an obscure SMTP outgoing bug which required at least the following
3628 conditions: (a) there was another message waiting for the same server;
3629 (b) the server returned 5xx to all RCPT commands in the first message so
3630 that the message was not completed; (c) the server dropped the connection
3631 or gave a negative response to the RSET that Exim sends to abort the
3632 transaction. The observed case was a dropped connection after DATA that had
3633 been sent in pipelining mode. That is, the server had advertised PIPELINING
3634 but was not implementing it correctly. The effect of the bug was incorrect
3635 behaviour, such as trying another host, and this could lead to a crash.
3641 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3642 and daemon.c were passed as pointers to ints; they should have been
3643 pointers to socklen_t variables (which are typically unsigned ints).
3645 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3648 3. Fixed a really odd bug that affected only the testing scheme; patching a
3649 certain fixed string in the binary changed the value of another string that
3650 happened to be identical to the end of the original first string.
3652 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3653 name", it returns that address as the IP address. On some operating
3654 systems (e.g. Solaris), it also passes back the IP address string as the
3655 "host name". However, on others (e.g. Linux), it passes back an empty
3656 string. Exim wasn't checking for this, and was changing the host name to an
3657 empty string, assuming it had been canonicized.
3659 5. Although rare, it is permitted to have more than one PTR record for a given
3660 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3661 all the names associated with an address, because they do in Solaris.
3662 However, it seems that they do not in Linux for data that comes from the
3663 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3664 I found this out when I moved to a new Linux workstation and tried to run
3665 the Exim test suite.
3667 To get round this problem I have changed the code so that it now does its
3668 own call to the DNS to look up PTR records when searching for a host name.
3669 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3670 addresses that are only in /etc/hosts are still found.
3672 This behaviour is, however, controlled by an option called host_lookup_
3673 order, which defaults to "bydns:byaddr". If people want to use the other
3674 order, or indeed, just use one or the other means of lookup, they can
3675 specify it in this variable.
3677 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3678 some operating systems, this comes back from gethostbyaddr() as an empty
3679 string, and this is what Exim used to test for. However, it seems that in
3680 other systems, "." is yielded. Exim now tests for this case too.
3682 7. The values of check_spool_space and check_log_space are now held internally
3683 as a number of kilobytes instead of an absolute number of bytes. If a
3684 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3685 kilobyte. This means that much larger values can be stored.
3687 8. Exim monitor: an attempt to get the action menu when not actually pointing
3688 at a message produces an empty menu entitled "No message selected". This
3689 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3690 no entries in it ("Shell widget menu has zero width and/or height"). So I
3691 have added a single, blank menu entry in this case.
3693 9. Added ${quote_local_part.
3695 10. MIME decoding is now applied to the contents of Subject: header lines when
3698 11. Now that a reference to $sender_host_address automatically causes a reverse
3699 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3700 host lookup before query-style lookups in lists that might use this
3701 variable. This has therefore been abolished, and the "net-" prefix is no
3702 longer necessary for query-style lookups.
3704 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3705 have been a typo for LFLAGS, so it has been changed.
3707 13. The install script calls Exim with "-C /dev/null" in order to find the
3708 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3709 to be output. Howeve, since Exim outputs its version number before the
3710 error, it didn't break the script. It just looked ugly. I fixed this by
3711 always allowing "-C /dev/null" if the caller is root.
3713 14. Ignore overlarge ACL variable number when reading spool file - insurance
3714 against a later release with more variables having written the file.
3716 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3717 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3718 this, as well as the form without the "IPv6" on the front.
3720 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3721 OS/Makefile-Darwin file.
3723 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3724 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3726 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3727 variable. However, some OS use ints for inodes. I've added cast to long int
3728 to get rid of the compiler warning.
3730 19. I had forgotten to lock out "/../" in configuration file names when
3731 ALT_CONFIG_PREFIX was set.
3733 20. Routers used for verification do not need to specify transports. However,
3734 if such a router generated a host list, and callout was configured, Exim
3735 crashed, because it could not find a port number from the (non-existent)
3736 transport. It now assumes port 25 in this circumstance.
3738 21. Added the -t option to exigrep.
3740 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3741 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3742 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3743 them is set, LOOKUP_LSEARCH is forced.
3745 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3747 24. Added sender and host information to the "rejected by local_scan()" log
3748 line; previously there was no indication of these.
3750 25. Added .include_if_exists.
3752 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3753 Linux. It turns out that not all file systems support this. Apparently some
3754 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3755 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3756 not supported on the file descriptor, is now ignored when Exim is trying to
3757 sync a directory. This applies only to Linux.
3759 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3761 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3762 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3765 29. Added support for maildirsize files from supplied patch (modified a bit).
3767 30. The use of :fail: followed by an empty string could lead Exim to respond to
3768 sender verification failures with (e.g.):
3770 550 Verification failed for <xxx>
3771 550 Sender verify failed
3773 where the first response line was missing the '-' that indicates it is not
3774 the final line of the response.
3776 31. The loop for finding the name of the user that called Exim had a hardwired
3777 limit of 10; it now uses the value of finduser_retries, which is used for
3778 all other user lookups.
3780 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3783 33. Exim was neglecting to zero errno before one call of strtol() when
3784 expanding a string and expecting an integer value. On some systems this
3785 resulted in spurious "integer overflow" errors. Also, it was casting the
3786 result into an int without checking.
3788 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3789 did not work. The code looks as if it has *never* worked, though it appears
3790 to have been documented since at least releast 1.62. I have made it work.
3792 35. The "timeout_DNS" error in retry rules, also documented since at least
3793 1.62, also never worked. As it isn't clear exactly what this means, and
3794 clearly it isn't a major issue, I have abolished the feature by treating it
3795 as "timeout", and writing a warning to the main and panic logs.
3797 36. The display of retry rules for -brt wasn't always showing the error code
3800 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3801 timeout_connect_A, timeout_connect_MX.
3803 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3804 to the empty sender.
3806 39. The daemon was not analysing the content of -oX till after it had closed
3807 stderr and disconnected from the controlling terminal. This meant that any
3808 syntax errors were only noted on the panic log, and the return code from
3809 the command was 0. By re-arranging the code a little, I've made the
3810 decoding happen first, so such errors now appear on stderr, and the return
3811 code is 1. However, the actual setting up of the sockets still happens in
3812 the disconnected process, so errors there are still only recorded on the
3815 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3816 connections (as happens on some IP stacks) was logged at start up time as
3817 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3818 it from "IPv6 and IPv4", which means that two separate sockets are being
3821 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3822 says whether AF_INET or AF_INET6 was passed as an argument.
3824 42. Exiwhat output was messed up when time zones were included in log
3827 43. Exiwhat now gives more information about the daemon's listening ports,
3828 and whether -tls-on-connect was used.
3830 44. The "port" option of the smtp transport is now expanded.
3832 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3833 silently ignored. Now an error message is written to the main and panic
3836 46. There's a new ACL modifier called "logwrite" which writes to a log file
3837 as soon as it is encountered.
3839 47. Added $local_user_uid and $local_user_gid at routing time.
3841 48. Exim crashed when trying to verify a sender address that was being
3844 49. Exim was recognizing only a space character after ".include". It now also
3845 recognizes a tab character.
3847 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3848 extracting the relevant information from the specification. The man page no
3849 longer contains scrambled data for the -d option, and I've added a section
3850 at the front about calling Exim under different names.
3852 51. Added "extra_headers" argument to the "mail" command in filter files.
3854 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3857 53. Installed eximstats 1.29.
3859 54. Added transport_filter_timeout as a generic transport option.
3861 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3862 Cc: header lines. This was required by RFC 822, but it not required by RFC
3865 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3866 incoming messages that did not have them. Now it does so only if the
3867 message originates locally, that is, if there is no associated remote host
3868 address. When Resent- header lines are present, this applies to the Resent-
3869 lines rather than the non-Resent- lines.
3871 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3872 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3874 58. Messages for configuration errors now include the name of the main
3875 configuration file - useful now that there may be more than one file in a
3876 list (.included file names were always shown).
3878 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3879 for those rare installations that do not start the daemon as root or run it
3880 setuid root. I've cut out the call to initgroups() if the daemon is not
3883 60. The Exim user and group can now be bound into the binary as text strings
3884 that are looked up at the start of Exim's processing.
3886 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3888 62. Added $mailstore_basename variable.
3890 63. Installed patch to sieve.c from Michael Haardt.
3892 64. When Exim failed to open the panic log after failing to open the main log,
3893 the original message it was trying to log was written to stderr and debug
3894 output, but if they were not available (the usual case in production), it
3895 was lost. Now it is written to syslog before the two lines that record the
3896 failures to open the logs.
3898 65. Users' Exim filters run in subprocesses under the user's uid. It is
3899 possible for a "deliver" command or an alias in a "personal" command to
3900 provoke an address rewrite. If logging of address rewriting is configured,
3901 this fails because the process is not running as root or exim. There may be
3902 a better way of dealing with this, but for the moment (because 4.30 needs
3903 to be released), I have disabled address rewrite logging when running a
3904 filter in a non-root, non-exim process.
3910 1. The buildconfig auxiliary program wasn't quoting the value set for
3911 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3912 not defined. This bug was masked in 4.22 by the effect that was fixed in
3915 2. Some messages that were rejected after a message id was allocated were
3916 shown as "incomplete" by exigrep. It no longer does this for messages that
3917 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3919 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3920 have allow_domain_literals set, the ID did not get logged in the <= line.
3921 Domain literals are now always recognized in Message-ID: header lines.
3923 4. The first argument for a ${extract expansion item is the key name or field
3924 number. Leading and trailing spaces in this item were not being ignored,
3925 causing some misleading effects.
3927 5. When deliver_drop_privilege was set, single queue runner processes started
3928 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3929 same command in the process it spins off) were not dropping privilege.
3931 6. When the daemon running as "exim" started a queue runner, it always
3932 re-executed Exim in the spun-off process. This is a waste of effort when
3933 deliver_drop_privilege is set. The new process now just calls the
3934 queue-runner function directly.
3940 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3943 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3944 ignored. Though the use of -oP was forcing the writing of a pid file, it
3945 was always written to the default place.
3947 3. If the message "no IP address found for host xxxx" is generated during
3948 incoming verification, it is now followed by identification of the incoming
3949 connection (so you can more easily find what provoked it).
3951 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3953 5. Added some features to "harden" Exim a bit more against certain attacks:
3955 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3956 be put in Local/Makefile. This is like the never_users runtime option,
3957 but it cannot be overridden. The default setting is "root".
3959 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3960 prefix string with which any file named in a -C command line option
3963 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3964 is retained for -C and -D only if the caller of Exim is root. Without
3965 it, the exim user may also use -C and -D and retain privilege.
3967 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3968 command line option is disabled.
3970 6. Macro names set by the -D option must start with an upper case letter, just
3971 like macro names defined in the configuration file.
3973 7. Added "dereference=" facility to LDAP.
3975 8. Two instances of the typo "uknown" in the source files are fixed.
3977 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3978 the Configure-Makefile script screwed up while processing it.
3980 10. Incorporated PCRE 4.4.
3982 11. The SMTP synchronization check was not operating right at the start of an
3983 SMTP session. For example, it could not catch a HELO sent before the client
3984 waited for the greeting. There is now a check for outstanding input at the
3985 point when the greeting is written. Because of the duplex, asynchronous
3986 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3987 way, but not yet received, when the check is performed.
3989 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3990 on TCP/IP sockets, because this apparently causes some broken clients to
3993 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3994 unchanged) from the Cygwin maintainer.
3996 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3997 was supported instead of testing for mbx. Effectively a typo.
3999 15. The spa authenticator server code was not checking that the input it
4000 received was valid base64.
4002 16. The debug output line for the "set" modifier in ACLs was not showing the
4003 name of the variable that was being set.
4005 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4007 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4008 lookup on its own. Something else was needed to trigger the lookup. For
4009 example, a match in host_lookup or the need for a host name in a host list.
4010 Now, if $sender_host_name is referenced and the host name has not yet been
4011 looked up, a lookup is performed. If the lookup fails, the variable remains
4012 empty, and $host_lookup_failed is set to "1".
4014 19. Added "eqi" as a case-independent comparison operator.
4016 20. The saslauthd authentication condition could segfault if neither service
4017 nor realm was specified.
4019 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4020 error message that was logged was misleading, and incoming SMTP
4021 connections were dropped. The message is now more accurate, and temporary
4022 errors are given to SMTP connections.
4024 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4025 (except RSET) with a 421 error, until QUIT is received. However, it was
4026 failing to send a response to QUIT.
4028 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4029 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4030 ACL is now run after the helo_try_verify_hosts code.
4032 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4033 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4034 case-independent, so other case variants are also recognized). Apparently
4035 some systems use these upper case variants.
4037 25. If more than two messages were waiting for the same host, and a transport
4038 filter was specified for the transport, Exim sent two messages over the
4039 same TCP/IP connection, and then failed with "socket operation on non-
4040 socket" when it tried to send the third.
4042 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4044 27. The extern definition of crypt16() in expand.c was not being excluded when
4045 the OS had its own crypt16() function.
4047 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4048 as a preferred synonym for return_size_limit, both as an option and as an
4051 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4053 30. Changed the default configuration ACL to relax the local part checking rule
4054 for addresses that are not in any local domains. For these addresses,
4055 slashes and pipe symbols are allowed within local parts, but the sequence
4056 /../ is explicitly forbidden.
4058 31. SPA server authentication was not clearing the challenge buffer before
4061 32. log_message in a "warn" ACL statement was writing to the reject log as
4062 well as to the main log, which contradicts the documentation and doesn't
4063 seem right (because no rejection is happening). So I have stopped it.
4065 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4066 However, I am unable to do any testing of this.
4068 34. Fixed an infelicity in the appendfile transport. When checking directories
4069 for a mailbox, to see if any needed to be created, it was accidentally
4070 using path names with one or more superfluous leading slashes; tracing
4071 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4073 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4074 contents are added to the log line that is written for every discarded
4075 recipient. (Previously a log_message setting was ignored.)
4077 36. The ${quote: operator now quotes the string if it is empty.
4079 37. The install script runs exim in order to find its version number. If for
4080 some reason other than non-existence or emptiness, which it checks, it
4081 could not run './exim', it was installing it with an empty version number,
4082 i.e. as "exim-". This error state is now caught, and the installation is
4085 38. An argument was missing from the function that creates an error message
4086 when Exim fails to connect to the socket for saslauthd authentication.
4087 This could cause Exim to crash, or give a corrupted message.
4089 39. Added isip, isip4, and isip6 to ${if conditions.
4091 40. The ACL variables $acl_xx are now saved with the message, and can be
4092 accessed later in routers, transports, and filters.
4094 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4095 strings in the file are not string-expanded.
4097 42. If a MAIL command specified a SIZE value that was too large to fit into an
4098 int variable, the check against message_size_limit failed. Such values are
4099 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4100 day this will have to be increased, but I don't think I want to be around
4101 when emails are that large.
4108 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4109 iconv() is not standard in FreeBSD.
4111 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4112 IPv6 enabled. The observed symptom was a segmentation fault on return from
4113 the function os_common_find_running_interfaces() in src/os.c.
4115 3. In the check_special_case() function in daemon.c I had used "errno" as an
4116 argument name, which causes warnings on some systems. This was basically a
4117 typo, since it was named "eno" in the comments!
4119 4. The code that waits for the clock to tick (at a resolution of some fraction
4120 of a second) so as to ensure message-id uniqueness was always waiting for
4121 at least one whole tick, when it could have waited for less. [This is
4122 almost certainly not relevant at current processor speeds, where it is
4123 unlikely to ever wait at all. But we try to future-proof.]
4125 5. The function that sleeps for a time interval that includes fractions of a
4126 second contained a race. It did not block SIGALRM between setting the
4127 timer, and suspending (a couple of lines later). If the interval was short
4128 and the sigsuspend() was delayed until after it had expired, the suspension
4129 never ended. On busy systems this could lead to processes getting stuck for
4132 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4133 process, before it forks any delivery processes. The open lookup caching
4134 mechanism meant that the open file or database connection was passed into
4135 the delivery process. The problem was that delivery processes always tidy
4136 up cached lookup data. This could cause a problem for the next delivery
4137 process started by the queue runner, because the external queue runner
4138 process does not know about the closure. So the next delivery process
4139 still has data in the lookup cache. In the case of a file lookup, there was
4140 no problem because closing a file descriptor in a subprocess doesn't affect
4141 the parent. However, if the lookup was caching a connection to a database,
4142 the connection was closed, and the second delivery process was likely to
4143 see errors such as "PGSQL: query failed: server closed the connection
4144 unexpectedly". The problem has been fixed by closing all cached lookups
4145 in a queue runner before running a delivery process.
4147 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4148 seem to have the "const" qualifier which it has on other OS. I've
4151 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4152 *of the same type* (client or server) with the same public name that an
4153 error should be diagnosed.
4155 9. When Exim looked up a host name for an IP address, but failed to find the
4156 original IP address when looking up the host name (a safety check), it
4157 output the message "<ip address> does not match any IP for NULL", which was
4158 confusing, to say the least. The bug was that the host name should have
4159 appeared instead of "NULL".
4161 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4162 user that is built into the binary, and the -C or -D options is used, root
4163 privilege is dropped before the configuration file is read. In addition,
4164 logging is switched to stderr instead of the normal log files. If the
4165 configuration then re-defines the Exim user, the unprivileged environment
4166 is probably not what is expected, so Exim logs a panic warning message (but
4169 However, if deliver_drop_privilege is set, the unprivileged state may well
4170 be exactly what is intended, so the warning has been cut out in that case,
4171 and Exim is allowed to try to write to its normal log files.
4177 1. smtp_return_error_details was not giving details for temporary sender
4178 or receiver verification errors.
4180 2. Diagnose a configuration error if two authenticators have the same public
4183 3. Exim used not to create the message log file for a message until the first
4184 delivery attempt. This could be confusing when incoming messages were held
4185 for policy or load reasons. The message log file is now created at the time
4186 the message is received, and an initial "Received" line is written to it.
4188 4. The automatically generated man page for command line options had a minor
4189 bug that caused no ill effects; however, a more serious problem was that
4190 the procedure for building the man page automatically didn't always
4191 operate. Consequently, release 4.20 contains an out-of-date version. This
4192 shouldn't happen again.
4194 5. When building Exim with embedded Perl support, the script that builds the
4195 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4196 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4198 6. The freeze_tell option was not being used for messages that were frozen on
4199 arrival, either by an ACL or by local_scan().
4201 7. Added the smtp_incomplete_transaction log selector.
4203 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4204 was accepting AUTH without a new EHLO.
4206 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4207 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4210 10. Logging of TCP/IP connections (when configured) now happens in the main
4211 daemon process instead of the child process, so that the TCP/IP connection
4212 count is more accurate (but it can never be perfect).
4214 11. The use of "drop" in a nested ACL was not being handled correctly in the
4215 outer ACL. Now, if condition failure induced by the nested "drop" causes
4216 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4217 or "require"), the connection is dropped.
4219 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4220 that yield "discard" can now be used with an "accept" or a "discard" verb,
4221 but an error is generated for any others (because I can't see a useful way
4222 to define what should happen).
4224 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4225 are now processed in the same way as lines in the Exim configuration file.
4226 In particular, continuation lines are supported.
4228 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4230 15. Added -ti meaning -t -i.
4232 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4233 domains, and warn by logging if others are found.
4235 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4236 returned from GIFCONF when Exim is trying to find the list of interfaces on
4237 a host. The code in os.c has been modified to copy each ifreq to an aligned
4238 structure in all cases.
4240 Also, in some cases, the returned ifreq's were being copied to a 'struct
4241 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4242 means the last couple of bytes of an IPv6 address could be chopped if the
4243 ifreq contained only a normal sockaddr (14 bytes storage).
4245 18. Named domain lists were not supported in the hosts_treat_as_local option.
4246 An entry such as +xxxx was not recognized, and was treated as a literal
4249 19. Ensure that header lines added by a DATA ACL are included in the reject log
4250 if the ACL subsequently rejects the message.
4252 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4253 MD5 (which is deprecated).
4255 21. When testing a filter file using -bf, Exim was writing a message when it
4256 took the sender from a "From " line in the message, but it was not doing so
4257 when it took $return_path from a Return-Path: header line. It now does.
4259 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4260 with a valid header line field name (a series of printing characters
4261 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4263 23. Changed "disc" in the source to "disk" to conform to the documentation and
4264 the book and for uniformity.
4266 24. Ignore Sendmail's -Ooption=value command line item.
4268 25. When execve() failed while trying to run a command in a pipe transport,
4269 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4270 could be confused with a return value of 69 from the command itself. This
4271 has been changed to 127, the value the shell returns if it is asked to run
4272 a non-existent command. The wording for the related log line suggests a
4273 non-existent command as the problem.
4275 26. If received_header_text expands to an empty string, do not add a Received:
4276 header line to the message. (Well, it adds a token one on the spool, but
4277 marks it "old" so that it doesn't get used or transmitted.)
4279 27. Installed eximstats 1.28 (addition of -nt option).
4281 28. There was no check for failure on the call to getsockname() in the daemon
4282 code. This can fail if there is a shortage of resources on the system, with
4283 ENOMEM, for example. A temporary error is now given on failure.
4285 29. Contrary to the C standard, it seems that in some environments, the
4286 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4287 program. Exim now does this explicitly; it affects the formatting of
4288 timestamps using strftime().
4290 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4291 complaints. I've now initialized all the variables, to avoid this.
4293 32. Header lines added by a system filter were not being "seen" during
4294 transport-time rewrites.
4296 33. The info_callback() function passed to OpenSSL is set up with type void
4297 (*)(SSL *, int, int), as described somewhere. However, when calling the
4298 function (actually a macro) that sets it up, the type void(*)() is
4299 expected. I've put in a cast to prevent warnings from picky compilers.
4301 34. If a DNS black list lookup found a CNAME record, but there were no A
4302 records associated with the domain it pointed at, Exim crashed.
4304 35. If a DNS black list lookup returned more than one A record, Exim ignored
4305 all but the first. It now scans all returned addresses if a particular IP
4306 value is being sought. In this situation, the contents of the
4307 $dnslist_value variable are a list of all the addresses, separated by a
4310 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4311 to accept a host name and all its aliases if the forward lookup for any of
4312 them yielded the IP address of the incoming connection. Now it accepts only
4313 those names whose forward lookup yields the correct IP address. Any other
4314 names are discarded. This closes a loophole whereby a rogue DNS
4315 administrator could create reverse DNS records to break through a
4316 wildcarded host restriction in an ACL.
4318 37. If a user filter or a system filter that ran in a subprocess used any of
4319 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4320 the wrong values were passed to the pipe command ($thisaddress had the
4321 value of $0, $0 had the value of $1, etc). This bug was introduced by
4322 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4324 38. Improved the line breaking for long SMTP error messages from ACLs.
4325 Previously, if there was no break point between 40 and 75 characters, Exim
4326 left the rest of the message alone. Two changes have been made: (a) I've
4327 reduced the minimum length to 35 characters; (b) if it can't find a break
4328 point between 35 and 75 characters, it looks ahead and uses the first one
4329 that it finds. This may give the occasional overlong line, but at least the
4330 remaining text gets split now.
4332 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4333 file descriptors might be low, and that setting 1000 would always raise it.
4334 It turns out that in some environments, the limit is already over 1000 and
4335 that lowering it causes trouble. So now Exim takes care not to decrease it.
4337 40. When delivering a message, the value of $return_path is set to $sender_
4338 address at the start of routing (routers may change the value). By an
4339 oversight, this default was not being set up when an address was tested by
4340 -bt or -bv, which affected the outcome if any router or filter referred to
4343 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4344 log only when "message" or "log_message" are set. However, if one of the
4345 conditions was an address verification, or a call to a nested ACL, the
4346 messages generated by the underlying test were being passed through. This
4347 no longer happens. The underlying message is available in $acl_verify_
4348 message for both "message" and "log_message" expansions, so it can be
4349 passed through if needed.
4351 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4352 new expansion $bh_ to give the encoded byte string without charset
4353 translation. Translation happens only if iconv() is available; HAVE_ICONV
4354 indicates this at build time. HEADERS_CHARSET gives the charset to
4355 translate to; headers_charset can change it in the configuration, and
4356 "headers charset" can change it in an individual filter file.
4358 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4359 that creates RFC 2047 encoded "words" labels them as that charset instead
4360 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4361 expansion operator; (ii) when Exim creates a From: line for a local
4362 message; (iii) when a header line is rewritten to include a "phrase" part.
4364 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4365 buggy, causing it to skip the first lines of messages whose message ID
4366 ended in 'D'. This would not have bitten before Exim release 4.14, because
4367 message IDs were unlikely to end in 'D' before then. The effect was to have
4368 incorrect size information for certain domains.
4370 45. #include "config.h" was missing at the start of the crypt16.c module. This
4371 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4374 46. If there was a timeout during a "random" callout check, Exim treated it as
4375 a failure of the random address, and carried on sending RSET and the real
4376 address. If the delay was just some slowness somewhere, the response to the
4377 original RCPT would be taken as a response to RSET and so on, causing
4378 mayhem of various kinds.
4380 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4381 when I implemented it. It didn't allow for the fact that some option values
4382 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4383 the right test for positive values.
4385 48. Domain names in DNS records are case-independent. Exim always looks them up
4386 in lower case. Some resolvers return domain names in exactly the case they
4387 appear in the zone file, that is, they may contain uppercase letters. Not
4388 all resolvers do this - some return always lower case. Exim was treating a
4389 change of case by a resolver as a change of domain, similar to a widening
4390 of a domain abbreviation. This triggered its re-routing code and so it was
4391 trying to route what was effectively the same domain again. This normally
4392 caused routing to fail (because the router wouldn't handle the domain
4393 twice). Now Exim checks for this case specially, and just changes the
4394 casing of the domain that it ultimately uses when it transmits the message
4397 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4400 50. If a filter generated a file delivery with a non-absolute name (possible if
4401 no home directory exists for the router), the forbid_file option was not
4404 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4405 the existing equality matching.
4407 52. Exim was using ints instead of ino_t variables in some places where it was
4408 dealing with inode numbers.
4410 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4411 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4412 called TMPDIR, and if it finds it is different, it changes its value.
4414 54. The smtp_printf() function is now made available to local_scan() so
4415 additional output lines can be written before returning. There is also an
4416 smtp_fflush() function to enable the detection of a dropped connection.
4417 The variables smtp_input and smtp_batched_input are exported to
4420 55. Changed the default runtime configuration: the message "Unknown user"
4421 has been removed from the ACL, and instead placed on the localuser router,
4422 using the cannot_route_message feature. This means that any verification
4423 failures that generate their own messages won't get overridden. Similarly,
4424 the "Unrouteable address" message that was in the ACL for unverifiable
4425 relay addresses has also been removed.
4427 56. Added hosts_avoid_esmtp to the smtp transport.
4429 57. The exicyclog script was not checking for the esoteric option
4430 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4431 will work only if exicyclog is run under the appropriate euid.
4433 58. Following a discussion on the list, the rules by which Exim recognises line
4434 endings on incoming messages have been changed. The -dropcr and drop_cr
4435 options are now no-ops, retained only for backwards compatibility. The
4436 following line terminators are recognized: LF CRLF CR. However, special
4437 processing applies to CR:
4439 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4440 nor a local message in the state where . is a terminator.
4442 (ii) If a bare CR is encountered in a header line, an extra space is added
4443 after the line terminator so as not to end the header. The reasoning
4444 behind this is that bare CRs in header lines are most likely either
4445 to be mistakes, or people trying to play silly games.
4447 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4448 was being incorrectly given as 18 bytes larger than it should have been.
4449 This is a VOB (very old bug).
4451 60. This may never have affected anything current, but just in case it has:
4452 When the local host is found other than at the start of a list of hosts,
4453 the local host, those with the same MX, and any that follow, are discarded.
4454 When the list in question was part of a longer list of hosts, the following
4455 hosts (not currently being processed) were also being discarded. This no
4456 longer happens. I'm not sure if this situation could ever has previously
4459 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4462 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4463 header. This is something that is recommended in a new Internet Draft, and
4464 is something that is documented as being done by Sendmail. There are two
4465 possible values. For messages generated by the autoreply transport, Exim
4468 Auto-Submitted: auto-replied
4470 whereas for all other generated messages (e.g. bounces) it adds
4472 Auto-Submitted: auto-generated
4474 63. The "personal" condition in filters now includes a test for the
4475 Auto-Submitted: header. If it contains the string "auto-" the message it
4476 not considered personal.
4478 64. Added rcpt_include_affixes as a generic transport option.
4480 65. Added queue_only_override (default true).
4482 66. Added the syslog_duplication option.
4484 67. If what should have been the first header line of a message consisted of
4485 a space followed by a colon, Exim was mis-interpreting it as a header line.
4486 It isn't of course - it is syntactically invalid and should therefore be
4487 treated as the start of the message body. The misbehaviour could have
4488 caused a number of strange effects, including loss of data in subsequent
4489 header lines, and spool format errors.
4491 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4492 client host had authenticated. This control can now be exercised by an ACL
4493 for more flexibility.
4495 69. By default, callouts do not happen when testing with -bh. There is now a
4496 variant, -bhc, which does actually run the callout code, including
4497 consulting and updating the callout cache.
4499 70. Added support for saslauthd authentication, courtesy of Alexander
4502 71. If statvfs() failed on the spool or log directories while checking their
4503 size for availability, Exim confusingly gave the error "space shortage".
4504 Furthermore, in debugging mode it crashed with a floating point exception.
4505 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4506 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4507 serious problem, Exim now writes to the main and panic logs when this
4508 happens, with details of the failure. It then refuses to accept the
4509 incoming message, giving the message "spool directory problem" or "log
4510 directory problem" with a 421 code for SMTP messages.
4512 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4513 0, 1, and 2 exist, because some OS complain for execs without them (see
4514 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4515 descriptors. However, the code omitted to check that the open succeeded,
4516 causing mysterious errors if for some reason the permissions on /dev/null
4517 got screwed. Now Exim writes a message to the main and panic logs, and
4518 bombs out if it can't open /dev/null.
4520 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4521 interact so that it is all more flexible. It is supposed to remain
4522 backwards compatible. Also added extra_local_interfaces.
4524 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4525 to bomb out with an assertion failure - to the client this appears as a
4526 connection drop. This problem occurs in the part of the code that was taken
4527 from the Samba project. Fortunately, the assertion is in a very simple
4528 function, so I have fixed this by reproducing the function inline in the
4529 one place where it is called, and arranging for authentication to fail
4530 instead of killing the process with assert().
4532 75. The SPA client code was not working when the server requested OEM rather
4533 than Unicode encoding.
4535 76. Added code to make require_files with a specific uid setting more usable in
4536 the case where statting the file as root fails - usually a non-root-mounted
4537 NFS file system. When this happens and the failure is EACCES, Exim now
4538 forks a subprocess and does the per-uid checking as the relevant uid.
4540 77. Added process_log_path.
4542 78. If log_file_path was not explicitly set, a setting of check_log_space or
4543 check_log_inodes was ignored.
4545 79. If a space check for the spool or log partitions fails, the incident is now
4546 logged. Of course, in the latter case the data may get lost...
4548 80. Added the %p formatting code to string_format() so that it can be used to
4549 print addresses in debug_print(). Adjusted all the address printing in the
4550 debugging in store.c to use %p rather than %d.
4552 81. There was a concern that a line of code in smtp_in.c could overflow a
4553 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4554 initially expressed, the concern was not well-founded, because trailing
4555 spaces are removed early. However, if the trailing spaces were followed by
4556 a NULL, they did not get removed, so the overflow was possible. Two fixes
4559 (a) I re-wrote the offending code in a cleaner fashion.
4560 (b) If an incoming SMTP command contains a NULL character, it is rejected
4563 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4564 runs initgroups(), so that if the Exim user is in any additional groups,
4565 they will be used during message reception.
4571 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob