1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.448 2006/12/20 19:57:47 tom Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
10 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
17 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
18 leftover -K file (the existence of which was triggered by #402).
19 While we were at it, introduced process PID as part of the -K
20 filename. This should rule out race conditions when creating
23 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
24 processing considerably. Previous code took too long for large mails,
25 triggering a timeout which in turn triggers #401.
27 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
28 in the DK code in transports.c. sendfile() is not really portable,
29 hence the _LINUX specificness.
31 TF/01 In the add_headers option to the mail command in an Exim filter,
32 there was a bug that Exim would claim a syntax error in any
33 header after the first one which had an odd number of characters
36 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
37 callout verification, Exim cached a "reject" for the entire domain. This
38 is correct for most verifications, but it is not correct for a recipient
39 verification with use_sender or use_postmaster set, because in that case
40 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
41 case of MAIL FROM:<> rejection from other early rejections (e.g.
42 rejection of HELO). When verifying a recipient using a non-null MAIL
43 address, the cache is ignored if it shows MAIL FROM:<> rejection.
44 Whatever the result of the callout, the value of the domain cache is
45 left unchanged (for any other kind of callout, getting as far as trying
46 RCPT means that the domain itself is ok).
48 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
51 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
52 manifest itself as EPIPE rather than ECONNECT. When tidying away a
53 session, the daemon ignores ECONNECT errors and logs others; it now
54 ignores EPIPE as well.
56 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
57 (quoted-printable decoding).
59 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
60 later the small subsequent patch to fix an introduced bug.
62 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
64 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
66 PH/08 An error is now given if message_size_limit is specified negative.
68 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
69 to be given (somewhat) arbitrary names.
71 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
74 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
75 miscellaneous code fixes
77 PH/10 Added the log_reject_target ACL modifier to specify where to log
80 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
81 hostname. This is wrong, because it relates to the incoming message (and
82 probably the interface on which it is arriving) and not to the outgoing
83 callout (which could be using a different interface). This has been
84 changed to use the value of the helo_data option from the smtp transport
85 instead - this is what is used when a message is actually being sent. If
86 there is no remote transport (possible with a router that sets up host
87 addresses), $smtp_active_hostname is used.
89 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
90 tweaks were necessary in order to get it to work (see also 21 below):
91 (a) The code assumed that strncpy() returns a negative number on buffer
92 overflow, which isn't the case. Replaced with Exim's string_format()
94 (b) There were several signed/unsigned issues. I just did the minimum
95 hacking in of casts. There is scope for a larger refactoring.
96 (c) The code used strcasecmp() which is not a standard C function.
97 Replaced with Exim's strcmpic() function.
98 (d) The code set only $1; it now sets $auth1 as well.
99 (e) A simple test gave the error "authentication client didn't specify
100 service in request". It would seem that Dovecot has changed its
101 interface. Fortunately there's a specification; I followed it and
102 changed what the client sends and it appears to be working now.
104 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
107 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
108 address A is aliased to B and C, where B exists and C does not. Without
109 -v the output is "A verified" because verification stops after a
110 successful redirection if more than one address is generated. However,
111 with -v the child addresses are also verified. Exim was outputting "A
112 failed to verify" and then showing the successful verification for C,
113 with its parentage. It now outputs "B failed to verify", showing B's
114 parentage before showing the successful verification of C.
116 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
117 look up a TXT record in a specific list after matching in a combined
120 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
121 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
122 they consult the DNS. I had assumed they would set it the way they
123 wanted; and indeed my experiments on Linux seem to show that in some
124 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
125 To be on the safe side, however, I have now made the interface to
126 host_find_byname() similar to host_find_bydns(), with an argument
127 containing the DNS resolver options. The host_find_byname() function now
128 sets these options at its start, just as host_find_bydns() does. The smtp
129 transport options dns_qualify_single and dns_search_parents are passed to
130 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
131 of host_find_byname() use the default settings of RES_DEFNAMES
132 (qualify_single) but not RES_DNSRCH (search_parents).
134 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
135 spool_read_header() do less string testing, by means of a preliminary
136 switch on the second character of optional "-foo" lines. (This is
137 overdue, caused by the large number of possibilities that now exist.
138 Originally there were few.) While I was there, I also converted the
139 str(n)cmp tests so they don't re-test the leading "-" and the first
140 character, in the hope this might squeeze out yet more improvement.
142 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
143 flag allowing group syntax was set by the header_syntax check but not
144 turned off, possible causing trouble later; (2) The flag was not being
145 set at all for the header_verify test, causing "group"-style headers to
146 be rejected. I have now set it in this case, and also caused header_
147 verify to ignore an empty address taken from a group. While doing this, I
148 came across some other cases where the code for allowing group syntax
149 while scanning a header line wasn't quite right (mostly, not resetting
150 the flag correctly in the right place). These bugs could have caused
151 trouble for malformed header lines. I hope it is now all correct.
153 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
154 with the "reply" argument non-NULL. The code, however (which originally
155 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
156 but it didn't always do it. This confused somebody who was copying the
157 code for some other use. I have removed all the tests.
159 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
160 feature that was used to support insecure browsers during the U.S. crypto
161 embargo. It requires special client support, and Exim is probably the
162 only MTA that supported it -- and would never use it because real RSA is
163 always available. This code has been removed, because it had the bad
164 effect of slowing Exim down by computing (never used) parameters for the
165 RSA_EXPORT functionality.
167 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
168 authenticator to fail if there's a tab character in the incoming data
169 (there should never be unless someone is messing about, as it's supposed
170 to be base64-encoded). Also added, on Timo's advice, the "secured" option
171 if the connection is using TLS or if the remote IP is the same as the
172 local IP, and the "valid-client-cert option" if a client certificate has
175 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
176 authenticators. This can be used for authorization after authentication
177 succeeds. (In the case of plaintext, it servers for both authentication
180 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
181 if any retry times were supplied.
183 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
184 connection, where there is no client IP address to check. In this
185 situation, the verify now always succeeds.
187 PH/25 Applied John Jetmore's -Mset patch.
189 PH/26 Added -bem to be like -Mset, but loading a message from a file.
191 PH/27 In a string expansion for a processed (not raw) header when multiple
192 headers of the same name were present, leading whitespace was being
193 removed from all of them, but trailing whitespace was being removed only
194 from the last one. Now trailing whitespace is removed from each header
195 before concatenation. Completely empty headers in a concatenation (as
198 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
199 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
201 PH/29 [Removed. This was a change that I later backed out, and forgot to
202 correct the ChangeLog entry (that I had efficiently created) before
203 committing the later change.]
205 PH/30 Exim was sometimes attempting to deliver messages that had suffered
206 address errors (4xx response to RCPT) over the same connection as other
207 messages routed to the same hosts. Such deliveries are always "forced",
208 so retry times are not inspected. This resulted in far too many retries
209 for the affected addresses. The effect occurred only when there were more
210 hosts than the hosts_max_try setting in the smtp transport when it had
211 the 4xx errors. Those hosts that it had tried were not added to the list
212 of hosts for which the message was waiting, so if all were tried, there
213 was no problem. Two fixes have been applied:
215 (i) If there are any address or message errors in an SMTP delivery, none
216 of the hosts (tried or untried) are now added to the list of hosts
217 for which the message is waiting, so the message should not be a
218 candidate for sending over the same connection that was used for a
219 successful delivery of some other message. This seems entirely
220 reasonable: after all the message is NOT "waiting for some host".
221 This is so "obvious" that I'm not sure why it wasn't done
222 previously. Hope I haven't missed anything, but it can't do any
223 harm, as the worst effect is to miss an optimization.
225 (ii) If, despite (i), such a delivery is accidentally attempted, the
226 routing retry time is respected, so at least it doesn't keep
227 hammering the server.
229 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
230 in ${readsocket because some servers need this prod.
232 PH/32 Added some extra debug output when updating a wait-xxx database.
234 PH/33 The hint "could be header name not terminated by colon", which has been
235 given for certain expansion errors for a long time, was not being given
236 for the ${if def:h_colon_omitted{... case.
238 PH/34 The spec says: "With one important exception, whenever a domain list is
239 being scanned, $domain contains the subject domain." There was at least
240 one case where this was not true.
242 PH/35 The error "getsockname() failed: connection reset by peer" was being
243 written to the panic log as well as the main log, but it isn't really
244 panic-worthy as it just means the connection died rather early on. I have
245 removed the panic log writing for the ECONNRESET error when getsockname()
248 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
249 runs only) independently of the message's sender address. This meant
250 that, if the 4xx error was in fact related to the sender, a different
251 message to the same recipient with a different sender could confuse
252 things. In particualar, this can happen when sending to a greylisting
253 server, but other circumstances could also provoke similar problems.
254 I have changed the default so that the retry time for these errors is now
255 based a combination of the sender and recipient addresses. This change
256 can be overridden by setting address_retry_include_sender=false in the
259 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
260 remote server are returned as part of bounce messages. This was not
261 happening for LMTP over a pipe (the lmtp transport), but now it is the
262 same for both kinds of LMTP.
264 PH/38 Despite being documented as not happening, Exim was rewriting addresses
265 in header lines that were in fact CNAMEs. This is no longer the case.
267 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
268 and queue runs started by the daemon processed all messages. This has
269 been fixed so that -R and -S can now usefully be given with -q<time>.
271 PH/40 Import PCRE release 6.7 (fixes some bugs).
273 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
275 PH/42 Give an error if -q is specified more than once.
277 PH/43 Renamed the variables $interface_address and $interface_port as
278 $received_ip_address and $received_port, to make it clear that these
279 values apply to message reception, and not to the outgoing interface when
280 a message is delivered. (The old names remain recognized, of course.)
282 PH/44 There was no timeout on the connect() call when using a Unix domain
283 socket in the ${readsocket expansion. There now is.
285 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
286 be meaningful with "accept".
288 SC/01 Eximstats V1.43
289 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
291 SC/02 Eximstats V1.44
292 Use a glob alias rather than an array ref in the generated
293 parser. This improves both readability and performance.
295 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
296 Collect SpamAssassin and rejection statistics.
297 Don't display local sender or destination tables unless
298 there is data to show.
299 Added average volumes into the top table text output.
301 SC/04 Eximstats V1.46
302 Collect data on the number of addresses (recipients)
303 as well as the number of messages.
305 SC/05 Eximstats V1.47
306 Added 'Message too big' to the list of mail rejection
307 reasons (thanks to Marco Gaiarin).
309 SC/06 Eximstats V1.48
310 Mainlog lines which have GMT offsets and are too short to
311 have a flag are now skipped.
313 SC/07 Eximstats V1.49 (Alain Williams)
314 Added the -emptyok flag.
316 SC/08 Eximstats V1.50
317 Fixes for obtaining the IP address from reject messages.
319 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
320 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
321 whitesspace changes from 4.64-PH/27
323 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
326 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
327 are found, allow negative numbers in numeric criteria)
329 JJ/06 exipick.20061117.2, added new $message_body_missing variable
331 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
332 to match changes made in 4.64-PH/43
334 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
336 PH/47 Put in an explicit test for a DNS lookup of an address record where the
337 "domain" is actually an IP address, and force a failure. This locks out
338 those revolvers/nameservers that support "A-for-A" lookups, in
339 contravention of the specifications.
341 PH/48 When a host name was looked up from an IP address, and the subsequent
342 forward lookup of the name timed out, the host name was left in
343 $sender_host_name, contrary to the specification.
345 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
346 restricted to single-key lookups, Exim was not diagnosing an error if
347 * or *@ was used with a query-style lookup.
349 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
351 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
352 long ago, but noone interested enough thought of it. Let's just say that
353 the "1.1" means that there are some new functions that weren't there at
354 some point in the past.
356 PH/51 Error processing for expansion failure of helo_data from an smtp
357 transport during callout processing was broken.
359 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
360 tested/used via the -bh/-bhc/-bs options.
362 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
363 bug, fixed in subsequent PCRE releases).
365 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
366 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
368 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
374 SC/01 Use a glob alias rather than an array ref in eximstats generated
375 parser. This improves both readability and performance.
377 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
378 Don't display local sender or destination tables in eximstats unless
379 there is data to show.
380 Added average volumes into the eximstats top table text output.
382 SC/03 Collect data on the number of addresses (recipients) as well
383 as the number of messages in eximstats.
385 TF/01 Correct an error in the documentation for the redirect router. Exim
386 does (usually) call initgroups() when daemonizing.
388 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
389 with consistent privilege compared to when running as a daemon.
391 TF/03 Note in the spec that $authenticated_id is not set for local
392 submissions from trusted users.
394 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
395 Thanks to Dean Brooks <dean@iglou.com> for the patch.
397 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
398 by adding some example configuration directives to the default
399 configuration file. A little bit of work is required to uncomment the
400 directives and define how usernames and passwords are checked, but
401 there is now a framework to start from.
403 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
404 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
405 without this. I don't know how relevant this is to other LDAP libraries.
407 PH/02 Add the verb name to the "unknown ACL verb" error.
409 PH/03 Magnus Holmgren's patch for filter_prepend_home.
411 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
413 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
414 directory not expanded when it should be if an expanded home directory
415 was set for the address (which is overridden by the transport).
417 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
420 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
421 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
422 because it is too late at that time, and has no effect.
424 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
425 security issue with \' (bugzilla #107). I could not use the
426 PQescapeStringConn() function, because it needs a PGconn value as one of
429 PH/08 When testing addresses using -bt, indicate those final addresses that
430 are duplicates that would not cause an additional delivery. At least one
431 person was confused, thinking that -bt output corresponded to deliveries.
432 (Suppressing duplicates isn't a good idea as you lose the information
433 about possibly different redirections that led to the duplicates.)
435 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
436 systems where poll() doesn't work, in particular OS X.
438 PH/10 Added more information to debugging output for retry time not reached.
440 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
441 operations in malware.c.
443 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
446 PH/13 If write_rejectlog was set false when logging was sent to syslog with
447 syslog_duplication set false, log lines that would normally be written
448 both the the main log and to the reject log were not written to syslog at
451 PH/14 In the default configuration, change the use of "message" in ACL warn
452 statements to "add_header".
454 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
455 not followed by a command (e.g. "seen endif").
457 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
458 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
461 PH/17 Added extra conditions to the default value of delay_warning_condition
465 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
466 { match{$h_precedence:}{(?i)bulk|list|junk} } \
467 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
470 The Auto-Submitted: and various List- headers are standardised, whereas I
471 don't think Precedence: ever was.
473 PH/18 Refactored debugging code in route_finduser() to show more information,
474 in particular, the error code if getpwnam() issues one.
476 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
477 This is apparently needed in addition to the PH/07 change above to avoid
478 any possible encoding problems.
480 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
481 but not after initializing Perl.
483 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
484 output them only if debugging. By default they are written stderr,
485 apparently, which is not desirable.
487 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
490 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
493 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
495 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
496 authenticated or an ident call has been made. Suppress the default
497 values for $authenticated_id and $authenticated_sender (but permit -oMai
498 and -oMas) when testing with -bh.
500 PH/24 Re-jigged the order of the tests in the default configuration so that the
501 tests for valid domains and recipients precede the DNS black list and CSA
502 tests, on the grounds that those ones are more expensive.
504 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
505 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
506 This bug exists in every version of Exim that I still have, right back to
509 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
510 However, an attempt to turn on case-sensitivity in a regex key by
511 including (?-i) didn't work because the subject string was already
512 lowercased, and the effects were non-intuitive. It turns out that a
513 one-line patch can be used to allow (?-i) to work as expected.
519 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
520 other effects) broke the use of negated acl sub-conditions.
522 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
525 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
526 "Deny" causes Exim to reject the incoming connection with a 554 error.
527 Unfortunately, if there is a major crisis, such as a disk failure,
528 tcp-wrappers gives "deny", whereas what one would like would be some
529 kind of temporary error. A kludge has been added to help with this.
530 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
531 554 error is used if errno is still zero or contains ENOENT (which occurs
532 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
535 PH/03 Add -lutil to the default FreeBSD LIBS setting.
537 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
538 errors. Otherwise a message that provokes a temporary error (when other
539 messages do not) can cause a whole host to time out.
541 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
542 addresses were routed directly to files or pipes from a redirect router.
543 File deliveries just didn't batch; pipe deliveries might have suffered
546 PH/06 A failure to get a lock for a hints database would erroneously always say
547 "Failed to get write lock", even when it was really a read lock.
549 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
550 of 0600. This has been changed to use the value of the lockfile_mode
551 option (which defaults to 0600).
553 PH/08 Applied small patch from the Sieve maintainer.
555 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
556 folder from quota calculations, a direct delivery into this folder messed
557 up the contents of the maildirsize file. This was because the regex was
558 used only to exclude .Trash (or whatever) when the size of the mailbox
559 was calculated. There was no check that a delivery was happening into an
560 excluded directory. This bug has been fixed by ignoring all quota
561 processing for deliveries into excluded directories.
563 PH/10 Added the maildirfolder_create_regex option to appendfile.
569 PH/01 The code for finding all the local interface addresses on a FreeBSD
570 system running IPv6 was broken. This may well have applied to all BSD
571 systems, as well as to others that have similar system calls. The broken
572 code found IPv4 interfaces correctly, but gave incorrect values for the
573 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
574 that it would not match correctly against @[] and not recognize the IPv6
577 PH/02 The ipliteral router was not recognizing addresses of the form user@
578 [ipv6:....] because it didn't know about the "ipv6:" prefix.
580 PH/03 Added disable_ipv6.
582 PH/04 Changed $reply_address to use the raw form of the headers instead of the
583 decoded form, because it is most often used to construct To: headers
584 lines in autoreplies, and the decoded form may well be syntactically
585 invalid. However, $reply_address has leading white space removed, and all
586 newlines turned into spaces so that the autoreply transport does not
589 PH/05 If group was specified without a user on a router, and no group or user
590 was specified on a transport, the group from the router was ignored.
592 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
593 for visible compile-time settings that can be used to change these
594 numbers, for those that want even more. Backwards compatibility with old
595 spool files has been maintained. However, going back to a previous Exim
596 release will lost any variables that are in spool files.
598 PH/07 Two small changes when running in the test harness: increase delay when
599 passing a TCP/IP connection to a new process, in case the original
600 process has to generate a bounce, and remove special handling of
601 127.0.0.2 (sic), which is no longer necessary.
603 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
604 be the same on different OS.
606 PH/09 Moved a debug statement in filter processing to avoid a race problem when
609 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
610 whether --show-vars was specified or not
612 JJ/02 exipick: Added support for new ACL variable spool format introduced
615 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
616 syntactically invalid From: or Reply-to: line, and a filter used this to
617 generate an autoreply, and therefore failed to obtain an address for the
618 autoreply, Exim could try to deliver to a non-existent relative file
619 name, causing unrelated and misleading errors. What now happens is that
620 it logs this as a hard delivery error, but does not attempt to create a
623 PH/11 The exinext utility has a -C option for testing purposes, but although
624 the given file was scanned by exinext itself; it wasn't being passed on
625 when Exim was called.
627 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
628 an end-of-file indication when reading a command response.
630 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
631 compiled. In many other places in Exim, IPv6 addresses are always
632 recognized, so I have changed this. It also means that IPv4 domain
633 literals of the form [IPV4:n.n.n.n] are now always recognized.
635 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
636 used if the router is not running as root, for example, when verifying at
637 ACL time, or when using -bh. The debugging output from this situation was
638 non-existent - all you got was a failure to exec. I have made two
641 (a) Failures to set uid/gid, the current directory, or a process leader
642 in a subprocess such as that created by queryprogram now generate
643 suitable debugging ouput when -d is set.
645 (b) The queryprogram router detects when it is not running as root,
646 outputs suitable debugging information if -d is set, and then runs
647 the subprocess without attempting to change uid/gid.
649 PH/15 Minor change to Makefile for building test_host (undocumented testing
652 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
653 additional section of a DNS packet that returns MX or SRV records.
654 Instead, it always explicitly searches for A/AAAA records. This avoids
655 major problems that occur when a DNS server includes only records of one
656 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
657 fixed another bug: if SRV records were looked up and the corresponding
658 address records were *not* found in the additional section, the port
659 values from the SRV records were lost.
661 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
662 using the correct key (the original address) when searching the retry
663 rules in order to find which one to use for generating the retry hint.
665 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
666 adding the default one. Similarly, if it contains a Reply-To: header, the
667 errors_reply_to option, if set, is not used.
669 PH/19 When calculating a retry time, Exim used to measure the "time since
670 failure" by looking at the "first failed" field in the retry record. Now
671 it does not use this if it is later than than the arrival time of the
672 message. Instead it uses the arrival time. This makes for better
673 behaviour in cases where some deliveries succeed, thus re-setting the
674 "first failed" field. An example is a quota failure for a huge message
675 when small messages continue to be delivered. Without this change, the
676 "time since failure" will always be short, possible causing more frequent
677 delivery attempts for the huge message than are intended.
678 [Note: This change was subsequently modified - see PH/04 for 4.62.]
680 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
681 $1, $2, $3) because the numerical variables can be reset during some
682 expansion items (e.g. "match"), thereby losing the authentication data.
684 PH/21 Make -bV show the size of off_t variables so that the test suite can
685 decide whether to run tests for quotas > 2G.
687 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
688 mailbox_size, and mailbox_filecount in the appendfile transport. If a
689 filecount value is greater than 2G or if a quota value is greater than 2G
690 on a system where the size of off_t is not greater than 4, a panic error
693 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
694 never match. The debug and -bh output now contains an explicit error
695 message indicating a malformed IPv4 address or mask.
697 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
698 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
701 PH/25 Do not write to syslog when running in the test harness. The only
702 occasion when this arises is a failure to open the main or panic logs
703 (for which there is an explicit test).
705 PH/26 Added the /no_tell option to "control=freeze".
707 PH/27 If a host name lookup failed very early in a connection, for example, if
708 the IP address matched host_lookup and the reverse lookup yielded a name
709 that did not have a forward lookup, an error message of the form "no IP
710 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
711 could be logged. Now it outputs the IP address instead of "NULL".
713 PH/28 An enabling patch from MH: add new function child_open_exim2() which
714 allows the sender and the authenticated sender to be set when
715 submitting a message from within Exim. Since child_open_exim() is
716 documented for local_scan(), the new function should be too.
718 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
719 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
720 results in an empty string is now treated as unset.
722 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
724 PH/31 Added sender_verify_fail logging option.
726 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
727 needed by RFC 822 but not by RFC 2822 was commented out. I have now
728 tidied the source and removed it altogether.
730 PH/33 When a queue run was abandoned because the load average was too high, a
731 log line was always written; now it is written only if the queue_run log
732 selector is set. In addition, the log line for abandonment now contains
733 information about the queue run such as the pid. This is always present
734 in "start" and "stop" lines but was omitted from the "abandon" line.
736 PH/34 Omit spaces between a header name and the colon in the error message that
737 is given when verify = headers_syntax fails (if there are lots of them,
738 the message gets confusing).
740 PH/35 Change the default for dns_check_names_pattern to allow slashes within
741 names, as there are now some PTR records that contain slashes. This check
742 is only to protect against broken name servers that fall over on strange
743 characters, so the fact that it applies to all lookups doesn't matter.
745 PH/36 Now that the new test suite is complete, we can remove some of the
746 special code in Exim that was needed for the old test suite. For example,
747 sorting DNS records because real resolvers return them in an arbitrary
748 order. The new test suite's fake resolver always returns records in the
751 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
752 bounces) except when queue_only is set, to avoid logging races between
753 the different processes.
755 PH/38 Panic-die if .include specifies a non-absolute path.
757 PH/39 A tweak to the "H" retry rule from its user.
759 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
760 a label. They prevented compilation on older perls.
762 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
763 a warning to be raised on newish perls.
765 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
766 on queue. Changes to match documented behaviour of showing count of
767 messages matching specified criteria.
769 PH/40 Changed the default ident timeout from 30s to 5s.
771 PH/41 Added support for the use of login_cap features, on those BSD systems
772 that have them, for controlling the resources used by pipe deliveries.
774 PH/42 The content-scanning code uses fopen() to create files in which to put
775 message data. Previously it was not paying any attention to the mode of
776 the files. Exim runs with umask(0) because the rest of the code creates
777 files with open(), and sets the required mode explicitly. Thus, these
778 files were ending up world-writeable. This was not a big issue, because,
779 being within the spool directory, they were not world-accessible. I have
780 created a function called modefopen, which takes an additional mode
781 argument. It sets umask(777), creates the file, chmods it to the required
782 mode, then resets the umask. All the relevant calls to fopen() in the
783 content scanning code have been changed to use this function.
785 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
786 to 24 hours. This avoids potential overflow problems when processing G
787 and H retry rules. I suspect nobody ever tinkers with this value.
789 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
791 PH/45 When the plaintext authenticator is running as a client, the server's
792 challenges are checked to ensure they are valid base64 strings. By
793 default, the authentication attempt is cancelled if an invalid string is
794 received. Setting client_ignore_invalid_base64 true ignores these errors.
795 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
796 they are received. Thus, the responses can be made to depend on the
797 challenges. If an invalid string is ignored, an empty string is placed in
800 PH/46 Messages that are created by the autoreply transport now contains a
801 References: header, in accordance with RFCs 2822 and 3834.
803 PH/47 Added authenticated_sender_force to the smtp transport.
805 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
807 PH/49 Installed latest patch from the Sieve maintainer.
809 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
810 was also set, the appendfile transport was unnecessarily scanning a
811 directory of message files (e.g. for maildir delivery) to find the count
812 of files (along with the size), even though it did not need this
813 information. It now does the scan only if it needs to find either the
814 size of the count of files.
816 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
818 PH/52 Two bugs concerned with error handling when the smtp transport is
821 (i) Exim was not creating retry information for temporary errors given
822 for individual recipients after the DATA command when the smtp transport
823 was used in LMTP mode. This meant that they could be retried too
824 frequently, and not timed out correctly.
826 (ii) Exim was setting the flag that allows error details to be returned
827 for LMTP errors on RCPT commands, but not for LMTP errors for individual
828 recipients that were returned after the DATA command.
830 PH/53 This is related to PH/52, but is more general: for any failing address,
831 when detailed error information was permitted to be returned to the
832 sender, but the error was temporary, then after the final timeout, only
833 "retry timeout exceeded" was returned. Now it returns the full error as
834 well as "retry timeout exceeded".
836 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
837 do this, and (what is worse) MTAs that accept it.
839 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
840 will now be deprecated.
842 PH/56 New os.c-cygwin from the Cygwin maintainer.
844 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
845 formats (previously only available in exim formats via -bpr, -bpru,
846 and -bpra. Now also available in native and exiqgrep formats)
848 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
849 with very large, slow to parse queues
851 JJ/08 exipick: added ! as generic prefix to negate any criteria format
853 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
855 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
856 responses to authentication challenges, though it was showing the
857 challenges; (ii) I've removed the CR characters from the debug output for
860 PH/58 Allow for the insertion of a newline as well as a space when a string
861 is turned into more than one encoded-word during RFC 2047 encoding. The
862 Sieve code now uses this.
864 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
865 data_4xx, lost_connection, tls_required.
867 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
868 message was being sent as an SMTP response.
870 PH/61 Add -l and -k options to exicyclog.
872 PH/62 When verifying, if an address was redirected to one new address, so that
873 verification continued, and the new address failed or deferred after
874 having set something in $address_data, the value of $address_data was not
875 passed back to the ACL. This was different to the case when no
876 redirection occurred. The value is now passed back in both cases.
878 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
879 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
880 use login_cap.h, so on its own it isn't the distinguishing feature. The
881 new name refers directly to the setclassresources() function.
883 PH/65 Added configuration files for NetBSD3.
885 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
887 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
888 is preferred over IPv4.
890 PH/68 The bounce_return_message and bounce_return_body options were not being
891 honoured for bounces generated during the reception of non-SMTP messages.
892 In particular, this applied to messages rejected by the ACL. This bug has
893 been fixed. However, if bounce_return_message is true and bounce_return_
894 body is false, the headers that are returned for a non-SMTP message
895 include only those that have been read before the error was detected.
896 (In the case of an ACL rejection, they have all been read.)
898 PH/69 The HTML version of the specification is now built in a directory called
899 spec_html instead of spec.html, because the latter looks like a path with
900 a MIME-type, and this confuses some software.
902 PH/70 Catch two compiler warnings in sieve.c.
904 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
905 function verify_get_ident() calls ip_connect() to connect a socket, but
906 if the "connect()" function timed out, ip_connect() used to close the
907 socket. However, verify_get_ident() also closes the socket later, and in
908 between Exim writes to the log, which may get opened at this point. When
909 the socket was closed in ip_connect(), the log could get the same file
910 descriptor number as the socket. This naturally causes chaos. The fix is
911 not to close the socket in ip_connect(); the socket should be closed by
912 the function that creates it. There was only one place in the code where
913 this was missing, in the iplookup router, which I don't think anybody now
914 uses, but I've fixed it anyway.
916 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
917 well as to direct DNS lookups. Otherwise the handling of names in host
918 lists is inconsistent and therefore confusing.
924 PH/01 Two changes to the default runtime configuration:
926 (1) Move the checks for relay_from_hosts and authenticated clients from
927 after to before the (commented out) DNS black list checks.
929 (2) Add control=submission to the relay_from_hosts and authenticated
930 clients checks, on the grounds that messages accepted by these
931 statements are most likely to be submissions.
933 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
935 (1) Generate an error if the third argument for the ${prvs expansion is
938 (2) Treat a missing third argument of ${prvscheck as if it were an empty
941 (3) Reset the variables that are obtained from the first argument of
942 ${prvscheck and used in the second argument before leaving the code,
943 because their memory is reclaimed, so using them afterwards may do
946 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
947 one (it's much easier than Tom thought :-).
949 (5) Because of (4), we can now allow for the use of $prvscheck_result
950 inside the third argument.
952 PH/03 For some reason, the default setting of PATH when running a command from
953 a pipe transport was just "/usr/bin". I have changed it to
956 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
957 anything to be listed in the output from -bV.
959 PH/05 When a filter generated an autoreply, the entire To: header line was
960 quoted in the delivery log line, like this:
962 => >A.N.Other <ano@some.domain> <original@ddress> ...
964 This has been changed so that it extracts the operative address. There
965 may be more than one such address. If so, they are comma-separated, like
968 => >ano@some.domain,ona@other.domain <original@ddress> ...
970 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
971 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
972 not being looked up in the rDNS to get a host name, Exim was showing the
973 IP address twice in Received: lines, even though the IP addresses were
974 identical. For example:
976 Received: from [1.2.3.4] (helo=[1.2.3.4])
978 However, if the real host name was known, it was omitting the HELO data
979 if it matched the actual IP address. This has been tidied up so that it
980 doesn't show the same IP address twice.
982 PH/07 When both +timestamp and +memory debugging was on, the value given by
983 $tod_xxx expansions could be wrong, because the tod_stamp() function was
984 called by the debug printing, thereby overwriting the timestamp buffer.
985 Debugging no longer uses the tod_stamp() function when +timestamp is set.
987 PH/08 When the original message was included in an autoreply transport, it
988 always said "this is a copy of the message, including all the headers",
989 even if body_only or headers_only was set. It now gives an appropriate
992 PH/09 Applied a patch from the Sieve maintainer which:
994 o fixes some comments
995 o adds the (disabled) notify extension core
996 o adds some debug output for the result of if/elsif tests
997 o points to the current vacation draft in the documentation
998 and documents the missing references header update
1002 o fixes a bug in processing the envelope test (when testing
1003 multiple envelope elements, the last element determinted the
1006 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1007 Electronic Mail") by including:
1009 Auto-submitted: auto-generated
1011 in the messages that it generates (bounce messages and others, such as
1012 warnings). In the case of bounce messages for non-SMTP mesages, there was
1013 also a typo: it was using "Auto_submitted" (underscore instead of
1014 hyphen). Since every message generated by Exim is necessarily in response
1015 to another message, thes have all been changed to:
1017 Auto-Submitted: auto-replied
1019 in accordance with these statements in the RFC:
1021 The auto-replied keyword:
1023 - SHOULD be used on messages sent in direct response to another
1024 message by an automatic process,
1026 - MUST NOT be used on manually-generated messages,
1028 - MAY be used on Delivery Status Notifications (DSNs) and Message
1029 Disposition Notifications (MDNs),
1031 - MUST NOT be used on messages generated by automatic or periodic
1032 processes, except for messages which are automatic responses to
1035 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1036 to the default Received: header definition.
1038 PH/12 Added log selector acl_warn_skipped (default on).
1040 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1041 variables because (a) they are in the wrong storage pool and (b) even if
1042 they were copied, it wouldn't work properly because of the caching.
1044 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1045 checking when decoding. Apparently there are clients that generate
1046 overlong encoded strings. Why am I not surprised?
1048 PH/15 If the first argument of "${if match_address" was not empty, but did not
1049 contain an "@" character, Exim crashed. Now it writes a panic log message
1050 and treats the condition as false.
1052 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1054 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1055 extension "envelope-auth". The code is finished and in agreement with
1056 other implementations, but there is no documentation so far and in fact,
1057 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1058 not changing the active code.
1060 Print executed "if" and "elsif" statements when debugging is used. This
1061 helps a great deal to understand what a filter does.
1063 Document more things not specified clearly in RFC3028. I had all this
1064 sorted out, when out of a sudden new issues came to my mind. Oops."
1066 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1069 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1070 canonical form (as documented). However, after a host name lookup from
1071 the IP address, check_host() was doing a simple string comparison with
1072 addresses acquired from the DNS when checking that the found name did
1073 have the original IP as one of its addresses. Since any found IPv6
1074 addresses are likely to be in abbreviated form, the comparison could
1075 fail. Luckily, there already exists a function for doing the comparison
1076 by converting both addresses to binary, so now that is used instead of
1077 the text comparison.
1079 PH/20 There was another similar case to PH/19, when a complete host name was
1080 given in a host list; looking up its IP address could give an abbreviated
1081 form, whereas the current host's name might or might not be abbreviated.
1082 The same fix has been applied.
1088 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1089 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1092 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1093 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1095 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1097 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1098 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1099 Also, for "Auto-Submitted", treat anything other than "no" as
1100 non-personal, in accordance with RFC 3834. (Previously it treated
1101 anything starting "auto-" as non-personal.)
1103 TF/01 The control=submission/name=... option had a problem with syntax
1104 errors if the name included a slash character. The /name= option
1105 now slurps the rest of the string, so it can include any characters
1106 but it must come last in the list of options (after /sender_retain
1109 PH/05 Some modifications to the interface to the fake nameserver for the new
1117 TK/01 Added the "success_on_redirect" address verification option. See
1118 NewStuff for rationale and an example.
1120 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1122 PH/02 Patch to exigrep to allow it to work on syslog lines.
1124 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1125 fread() to skip over the body file's header line, because in Cygwin the
1126 header line is locked and is inaccessible.
1128 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1129 co-exist for some time) to make it clear that it is the Exim ID that is
1130 referenced, not the Message-ID: header line.
1132 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1133 string_format() function, because snprintf() does not exist on all
1136 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1137 ${stat: expansion item.
1139 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1140 protocol synchronization error", to keep the pedants happy.
1142 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1143 well as for IRIX systems, when gcc is being used. See the host.c source
1146 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1148 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1151 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1152 transport and to the smtp transport in LMTP mode.
1154 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1156 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1157 run a filter in a subprocess. This could lead to confusion in subsequent
1158 lookups in the parent process. There should also be a search_tidyup() at
1159 the end of the subprocess.
1161 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1162 only if the host matched helo_try_verify_hosts, which caused the
1163 verification to occur when the EHLO/HELO command was issued. The ACL just
1164 tested the remembered result. Now, if a previous verification attempt has
1165 not happened, "verify = helo" does it there and then.
1167 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1169 TK/03 Fix log output including CR from clamd.
1171 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1172 exist provoked a memory error which could cause a segfault.
1174 PH/15 Installed PCRE 6.2
1176 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1178 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1179 of the problem. Specifically, suggested +O2 rather than +O1 for the
1182 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1184 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1185 record, the port was not correctly logged when the outgoing_port log
1186 selector was set (it logged the transort's default port).
1188 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1189 fallback_hosts, and "hosts" in the smtp transport.
1191 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1192 host errors such as "Connection refused".
1194 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1195 authentication with radiusclient 0.4.9:
1197 - Error returned from rc_read_config was caught wrongly
1198 - Username/password not passed on to radius server due to wrong length.
1200 The presumption is that some radiusclient API changes for 4.51/PH/17
1201 were not taken care of correctly. The code is still untested by me (my
1202 Linux distribution still has 0.3.2 of radiusclient), but it was
1203 contributed by a Radius user.
1205 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1206 expanding the "port" option of the smtp transport.
1208 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1209 while reading a MIME header. Thanks to Tom Hughes for a patch.
1211 PH/24 Include config.h inside local_scan.h so that configuration settings are
1214 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1215 that in an ACL for RCPT (for example), you can examine exactly what was
1218 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1219 commands, but it was not correctly comparing the address with the actual
1220 client host address. Thus, it would show the EHLO address in Received:
1221 header lines when this was not necessary.
1223 PH/27 Added the % operator to ${eval:}.
1225 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1226 it should be ignoring failures (because with -C, for example, it has lost
1227 privilege). It wasn't ignoring creation failures other than "already
1230 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1233 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1234 because an input line was too long, either on its own, or by virtue of
1235 too many continuations, the temporary file was not being removed, and the
1236 return code was incorrect.
1238 PH/31 Missing "BOOL" in function definition in filtertest.c.
1240 PH/32 Applied Sieve patches from the maintainer.
1242 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1244 PH/33 Added "verify = not_blind".
1246 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1247 Local/Makefile (with some defaults set). These are used in built scripts
1248 such as exicyclog, but they have never been used in the exim_install
1249 script (though there are many overriding facilities there). I have
1250 arranged that the exim_install script now takes note of these two
1253 PH/35 Installed configuration files for Dragonfly.
1255 PH/36 When a locally submitted message by a trusted user did not contain a
1256 From: header, and the sender address was obtained from -f or from an SMTP
1257 MAIL command, and the trusted user did not use -F to supply a sender
1258 name, $originator_name was incorrectly used when constructing a From:
1259 header. Furthermore, $originator_name was used for submission mode
1260 messages from external hosts without From: headers in a similar way,
1261 which is clearly wrong.
1263 PH/37 Added control=suppress_local_fixups.
1265 PH/38 When log_selector = +received_sender was set, and the addition of the
1266 sender made the log line's construction buffer exactly full, or one byte
1267 less than full, an overflow happened when the terminating "\n" was
1270 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1271 when the result of a list match is failure because a DNS lookup failed.
1273 PH/40 RM_COMMAND is now used in the building process.
1275 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1276 the "build-* directories that it finds.
1278 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1279 address in a domain literal was a prefix of an interface address.
1281 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1282 when verifying a sender address, unless rewrite_headers is false.
1284 PH/44 Wrote a long comment about why errors_to addresses are verified as
1285 recipients, not senders.
1287 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1288 the ratelimit ACL was added.
1290 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1292 PH/46 Added extra information about PostgreSQL errors to the error string.
1294 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1295 avoiding the need to install special zones in a real server. This is
1296 backwards compatible; if it can't find the fake resolver, it drops back.
1297 Thus, both old and new test suites can be run.
1299 TF/02 Added util/ratelimit.pl
1301 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1302 clock is set back in time.
1304 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1305 Candler <B.Candler@pobox.com>.
1307 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1308 OK for addresses that are the result of redirections.
1310 PH/48 A number of further additions for the benefit of the new test suite,
1311 including a fake gethostbyname() that interfaces to the fake DNS resolver
1314 TF/06 The fix for widen_domains has also been applied to qualify_single and
1315 search_parents which are the other dnslookup options that can cause
1318 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1321 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1323 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1324 service" if it encountered an MX record with an empty target hostname.
1325 The message is now "an MX or SRV record indicated no SMTP service".
1327 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1328 if the DNS of the sending site is misconfigured. This is quite a
1329 common situation. This change restores the behaviour of treating a
1330 helo verification defer as a failure.
1332 PH/51 If self=fail was set on a router, the bounce message did not include the
1333 actual error message.
1339 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1341 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1342 command itself ended in error, the underlying message about the transport
1343 filter timeout was being overwritten with the pipe command error. Now the
1344 underlying error message should be appended to the second error message.
1346 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1348 PH/02 Reduce the amount of output that "make" produces by default. Full output
1349 can still be requested.
1351 PH/03 The warning log line about a condition test deferring for a "warn" verb
1352 was being output only once per connection, rather than after each
1353 occurrence (because it was using the same function as for successful
1354 "warn" verbs). This seems wrong, so I have changed it.
1356 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1357 it should not have, which might have caused a crash in the right
1358 circumstances, but probably never did.
1360 PH/04 Installed a modified version of Tony Finch's patch to make submission
1361 mode fix the return path as well as the Sender: header line, and to
1362 add a /name= option so that you can make the user's friendly name appear
1365 TF/03 Added the control = fakedefer ACL modifier.
1367 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1368 Mark Lowes for thorough testing.
1370 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1372 TK/03 Merged latest SRS patch from Miles Wilton.
1374 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1375 with the definition in sysexits.h (which is #included earlier).
1376 Fortunately, Exim does not actually use EX_OK. The code used to try to
1377 preserve the sysexits.h value, by assumimg that macro definitions were
1378 scanned for macro replacements. I have been disabused of this notion,
1379 so now the code just undefines EX_OK before #including unistd.h.
1381 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1382 in the smtp transport. When a block could not be written in a single
1383 write() function, the timeout was being re-applied to each part-write.
1384 This seems wrong - if the receiver was accepting one byte at a time it
1385 would take for ever. The timeout is now adjusted when this happens. It
1386 doesn't have to be particularly precise.
1388 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1389 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1391 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1392 without a domain if the check to <postmaster@domain> fails.
1394 SC/01 Eximstats: added -xls and the ability to specify output files
1395 (patch written by Frank Heydlauf).
1397 SC/02 Eximstats: use FileHandles for outputing results.
1399 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1401 SC/04 Eximstats: fixed display of large numbers with -nvr option
1403 SC/05 Eximstats: fixed merging of reports with empty tables.
1405 SC/06 Eximstats: added the -include_original_destination flag
1407 SC/07 Eximstats: removed tabs and trailing whitespace.
1409 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1411 TK/06 MBOX spool code: Add real "From " MBOX separator line
1412 so the .eml file is really in mbox format (even though
1413 most programs do not really care). Patch from Alex Miller.
1415 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1416 The latter is generated from $received_to and is only set if the
1417 message has one envelope recipient. SA can use these headers,
1418 obviously out-of-the-box. Patch from Alex Miller.
1420 PH/08 The ${def test on a variable was returning false if the variable's
1421 value was "0", contrary to what the specification has always said!
1422 The result should be true unless the variable is empty.
1424 PH/09 The syntax error of a character other than { following "${if
1425 def:variable_name" (after optional whitespace) was not being diagnosed.
1426 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1427 accidental colon was present, for example, could give incorrect results.
1429 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1430 result is used (not including appendfile, where other changes are about
1433 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1434 This involved changing a lot of size variables from int to off_t. It
1435 should work with maildirs and everything.
1437 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1438 spamd dying while we are connected to it.
1440 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1443 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1444 function for BDB 4.3.
1446 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1448 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1449 the sources for PCRE have been re-arranged and more files are now
1452 PH/15 The code I had for printing potentially long long variables in PH/11
1453 above was not the best (it lost precision). The length of off_t variables
1454 is now inspected at build time, and an appropriate printing format (%ld
1455 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1456 to be "long long int" or "long int". This is needed for the internal
1457 formatting function string_vformat().
1459 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1460 the configuration file to be ":syslog", then the script "guesses" where
1461 the logs files are, rather than using the compiled in default. In our
1462 case the guess is not the same as the compiled default, so the script
1463 suddenly stopped working when I started to use syslog. The patch checks
1464 to see if log_file_path is "". If so, it attempts to read it from exim
1465 with no configuration file to get the compiled in version, before it
1466 falls back to the previous guessing code."
1468 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1469 implementing BATV in an Exim configuration. See NewStuff for the gory
1472 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1473 Makefile that are specific to HP-UX.
1475 PH/18 If the "use_postmaster" option was set for a recipient callout together
1476 with the "random" option, the postmaster address was used as the MAIL
1477 FROM address for the random test, but not for the subsequent recipient
1478 test. It is now used for both.
1480 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1481 patch removes a few documentation additions to RFC 3028, because the
1482 latest draft now contains them. It adds the new en;ascii-case comparator
1483 and a new error check for 8bit text in MIME parts. Comparator and
1484 require names are now matched exactly. I enabled the subaddress
1485 extension, but it is not well tested yet (read: it works for me)."
1487 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1488 rework some of the code of TK/09 above to avoid the hardwired use of
1489 "%lld" and "long long". Replaced the call to snprintf() with a call to
1492 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1493 records point to non-existent hosts", "retry timeout exceeded", and
1494 "retry time not reached for any host after a long failure period".
1496 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1497 experimental DomainKeys support:
1499 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1500 (2) On an error such as an illegally used "control", the wrong name for
1501 the control was given.
1503 These problems did NOT occur unless DomainKeys support was compiled.
1505 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1507 PH/24 Added ${if match_ip condition.
1509 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1510 databases so that it will be absolutely obvious if a crash occurs in the
1511 DB library. This is a regular occurrence (often caused by mis-matched
1514 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1515 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1516 picked up on a user's system that detects such things. There doesn't seem
1517 to be a gcc warning option for this - only an attribute that has to be
1518 put on the function's prototype. It seems that in Fedora Core 4 they have
1519 set this on a number of new functions. No doubt there will be more in due
1522 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1523 specify a transport. However, if an address that was verified by such a
1524 router was the subject of a callout, Exim crashed because it tried to
1525 read the rcpt_include_affixes from the non-existent transport. Now it
1526 just assumes that the setting of that option is false. This bug was
1527 introduced by 4.51/PH/31.
1529 PH/28 Changed -d+all to exclude +memory, because that information is very
1530 rarely of interest, but it makes the output a lot bigger. People tend to
1531 do -d+all out of habit.
1533 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1534 code in os-type was giving problems when libc.so lives in lib64, like on
1537 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1538 aren't the modern standard, and it seems that some systems' include files
1539 don't always have them. Exim was already checking for some of the newer
1540 ones like T_AAAA, and defining it itself. I've added checks for all the
1541 record types that Exim uses.
1543 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1544 not automatically generating a new one, as it is supposed to. This
1545 prevented TLS from working. If the file did exist, but contained invalid
1546 data, a new version was generated, as expected. It was only the case of a
1547 non-existent file that was broken.
1549 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1550 with a change in libdomainkeys > 0.64.
1552 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1553 from DNS. If the selector record carries the flag, it now has
1554 precedence over the domain-wide flag.
1556 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1558 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1559 the use of a port name that isn't defined in /etc/services) occurred, the
1560 message was deferred as in a normal delivery, and thus remained on the
1561 spool, instead of being failed because of the mua_wrapper setting. This
1562 is now fixed, and I tidied up some of the mua_wrapper messages at the
1565 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1566 the messages in a hash of arrays rather than using individual hashes.
1567 This is a bit cleaner and results in dramatic memory savings, albeit
1568 at a slight CPU cost.
1570 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1571 as requested by Marc Sherman.
1573 SC/10 Eximstats: added histograms for user specified patterns as requested
1576 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1578 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1579 fopen() in the content-scanning modules that did not already have it.
1585 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1586 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1588 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1590 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1592 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1593 to test Sieve filters that use "vacation".
1595 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1596 that changes the way the GnuTLS parameters are stored in the cache file.
1597 The new format can be generated externally. For backward compatibility,
1598 if the data in the cache doesn't make sense, Exim assumes it has read an
1599 old-format file, and it generates new data and writes a new file. This
1600 means that you can't go back to an older release without removing the
1603 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1604 work if there are any delivery delays because "one_time" forces the
1605 parent to be marked "delivered", so its unseen clone is never tried
1606 again. For this reason, Exim now forbids the simultaneous setting of
1609 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1610 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1611 entry if you want to know the details. The fix, however, neglected to
1612 consider the case when local delivery batching is involved. The test for
1613 "previously delivered" was not happening when checking to see if an
1614 address could be batched with a previous (undelivered) one; under
1615 certain circumstances this could lead to multiple deliveries to the same
1618 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1619 in its include files, and this causes problems building Exim.
1621 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1622 header_syntax) but Exim was just ignoring anything given after a slash.
1623 In particular, this caused confusion with an attempt to use "verify =
1624 reverse_host_lookup/defer_ok". An error is now given when options are
1625 supplied for verify items that do not have them. (Maybe reverse_host_
1626 lookup should have a defer_ok option, but that's a different point.)
1628 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1629 defined by RFC 821) to 2048, because there were problems with some AUTH
1630 commands, and RFC 1869 says the size should be increased for extended
1631 SMTP commands that take arguments.
1633 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1636 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1637 "unknown" error; now it says that the functionality isn't in the binary.
1639 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1640 an address' error message when a string expansion fails (syntax or
1641 whatever). Otherwise the password may appear in the log. Following change
1642 PH/42 below, there is no longer a chance of it appearing in a bounce
1645 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1647 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1648 of its addresses was ever tried. (Bugzilla bug #2.)
1650 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1651 the result incorrectly in the debug output. (It correctly added a newline
1652 to what was transported.)
1654 TF/01 Added $received_time.
1656 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1657 commented out examples of how to interface to a virus scanner and to
1658 SpamAssassin. Also added commented examples of av_scanner and
1659 spamd_address settings.
1661 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1662 and controls are allowed in which ACLs. There were a couple of minor
1663 errors. Some of the entries in the conditions table (which is a table of
1664 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1665 negation of where the condition IS allowed.
1667 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1669 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1670 header file does not have a version number, so I've had to invent a new
1671 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1672 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1673 radiusclient), but it was contributed by a Radius user.
1675 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1676 files or directories, for OpenSSL.
1678 PH/19 When an Exim process that is running as root has to create an Exim log
1679 file, it does so in a subprocess that runs as exim:exim so as to get the
1680 ownership right at creation (otherwise, other Exim processes might see
1681 the file with the wrong ownership). There was no test for failure of this
1682 fork() call, which would lead to the process getting stuck as it waited
1683 for a non-existent subprocess. Forks do occasionally fail when resources
1684 run out. I reviewed all the other calls to fork(); they all seem to check
1687 PH/20 When checking for unexpected SMTP input at connect time (before writing
1688 the banner), Exim was not dealing correctly with a non-positive return
1689 from the read() function. If the client had disconnected by this time,
1690 the result was a log entry for a synchronization error with an empty
1691 string after "input=" when read() returned zero. If read() returned -1
1692 (an event I could not check), uninitialized data bytes were printed.
1693 There were reports of junk text (parts of files, etc) appearing after
1696 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1698 PH/22 Added support for macro redefinition, and (re)definition in between
1699 driver and ACL definitions.
1701 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1702 forgetting to use the resulting value; it was using the unexpanded value.
1704 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1705 hadn't been configured. The fix is from Juergen Kreileder, who
1706 understands it better than I do:
1708 "Here's what I see happening with three configured cyrus_sasl
1709 authenticators configured (plain, login, cram-md5):
1711 On startup auth_cyrus_sasl_init() gets called for each of these.
1712 This means three calls to sasl_listmech() without a specified mech_list.
1713 => SASL tests which mechs of all available mechs actually work
1714 => three warnings about OTP not working
1715 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1717 With the patch, sasl_listmech() also gets called three times. But now
1718 SASL's mech_list option is set to the server_mech specified in the the
1719 authenticator. Or in other words, the answer from sasl_listmech()
1720 gets limited to just the mech you're testing for (which is different
1722 => the return list contains just 'plain' or 'login', 'cram-md5' or
1723 nothing depending on the value of ob->server_mech.
1725 I've just tested the patch: Authentication still works fine,
1726 unavailable mechs specified in the exim configuration are still
1727 caught, and the auth.log warnings about OTP are gone."
1729 PH/25 When debugging is enabled, the contents of the command line are added
1730 to the debugging output, even when log_selector=+arguments is not
1733 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1734 answer is "GNU", and only if the return is "GNU/something" is the answer
1737 PH/27 $acl_verify_message is now set immediately after the failure of a
1738 verification in an ACL, and so is available in subsequent modifiers. In
1739 particular, the message can be preserved by coding like this:
1741 warn !verify = sender
1742 set acl_m0 = $acl_verify_message
1744 Previously, $acl_verify_message was set only while expanding "message"
1745 and "log_message" when a very denied access.
1747 PH/28 Modified OS/os.c-Linux with
1749 -#ifndef OS_LOAD_AVERAGE
1750 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1752 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1753 nomenclature these days.)
1755 PH/29 Installed patch from the Sieve maintainer that adds the options
1756 sieve_useraddress and sieve_subaddress to the redirect router.
1758 PH/30 In these circumstances:
1759 . Two addresses routed to the same list of hosts;
1760 . First host does not offer TLS;
1761 . First host accepts first address;
1762 . First host gives temporary error to second address;
1763 . Second host offers TLS and a TLS session is established;
1764 . Second host accepts second address.
1765 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1766 and peerdn, if requested) that were in fact used only for the second
1769 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1770 attention to any local part prefix or suffix that was matched by the
1771 router that accepted the address. It now behaves in the same way as it
1772 does for delivery: the affixes are removed from the local part unless
1773 rcpt_include_affixes is set on the transport.
1775 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1776 timeout during the DATA phase of an incoming message.
1778 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1779 applied a patch sanctioned by the Sieve maintainer.
1781 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1782 the uid or gid is negative. A case of a negative gid caused this to be
1783 noticed. The fix allows for either to be negative.
1785 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1786 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1789 PH/36 The change PH/12 above was broken. Fixed it.
1791 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1792 the grounds that routing the same address twice would always produce the
1793 same answer. This might have been true once, but it is certainly no
1794 longer true now. Routing a child address may depend on the previous
1795 routing that produced that child. Some complicated redirection strategies
1796 went wrong when messages had multiple recipients, and made Exim's
1797 behaviour dependent on the order in which the addresses were given.
1799 I have moved the duplicate checking until after the routing is complete.
1800 Exim scans the addresses that are assigned to local and remote
1801 transports, and removes any duplicates. This means that more work will be
1802 done, as duplicates will always all be routed, but duplicates are
1803 presumably rare, so I don't expect this is of any significance.
1805 For deliveries to pipes, files, and autoreplies, the duplicate checking
1806 still happens during the routing process, since they are not going to be
1809 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1810 It corrects a timeout issue with spamd. This is Ian's comment: "The
1811 background is that sometimes spamd either never reads data from a
1812 connection it has accepted, or it never writes response data. The exiscan
1813 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1814 blindly assumes that writes won't block so it may never time out."
1816 PH/39 Allow G after quota size as well as K and M.
1818 PH/40 The value set for $authenticated_id in an authenticator may not contain
1819 binary zeroes or newlines because the value is written to log lines and
1820 to spool files. There was no check on this. Now the value is run through
1821 the string_printing() function so that such characters are converted to
1822 printable escape sequences.
1824 PH/41 $message_linecount is a new variable that contains the total number of
1825 lines in the message. Compare $body_linecount, which is the count for the
1828 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1829 bounce and delay warning messages, except in certain special cases, which
1832 (a) An SMTP error message from a remote host;
1833 (b) A message specified in a :fail: redirection;
1834 (c) A message specified in a "fail" command in a system filter;
1835 (d) A message specified in a FAIL return from the queryprogram router;
1836 (e) A message specified by the cannot_route_message router option.
1838 In these cases only, Exim does include the error details in bounce and
1839 warning messages. There are also a few cases where bland messages such
1840 as "unrouteable address" or "local delivery error" are given.
1842 PH/43 $value is now also set for the "else" part of a ${run expansion.
1844 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1845 being worked on, but at least Exim now implements the latest version to
1848 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1849 process to complete was treated as a delivery failure, a timeout while
1850 writing the message to the pipe was logged, but erroneously treated as a
1851 successful delivery. Such timeouts include transport filter timeouts. For
1852 consistency with the overall process timeout, these timeouts are now
1853 treated as errors, giving rise to delivery failures by default. However,
1854 there is now a new Boolean option for the pipe transport called
1855 timeout_defer, which, if set TRUE, converts the failures into defers for
1856 both kinds of timeout. A transport filter timeout is now identified in
1859 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1860 systems where "make" and "gmake" are different, calling "gmake" at top
1861 level broke things. I've arranged for the value of $(MAKE) to be passed
1862 from the Makefile to this script so that it can call the same version of
1866 A note about Exim versions 4.44 and 4.50
1867 ----------------------------------------
1869 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1870 changes of various kinds. As a consequence, a big documentation update was
1871 needed. This delayed the release for rather longer than seemed good, especially
1872 in the light of a couple of (minor) security issues. Therefore, the changes
1873 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1874 release. So 4.44 and 4.50 are in effect two different branches that both start
1877 I have left the 4.50 change log unchanged; it contains all the changes since
1878 4.43. The change log for 4.44 is below; many of its items are identical to
1879 those for 4.50. This seems to be the most sensible way to preserve the
1880 historical information.
1886 1. Minor wording change to the doc/README.SIEVE file.
1888 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1889 computation of the current number of files was incorrect.
1891 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1892 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1895 4. Give more explanation in the error message when the command for a transport
1896 filter fails to execute.
1898 5. There are several places where Exim runs a non-Exim command in a
1899 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1900 was being done only for the command run by the queryprogram router. It is
1901 now done for all such subprocesses. The other cases are: ${run, transport
1902 filters, and the commands run by the lmtp and pipe transports.
1904 6. Added CONFIGURE_GROUP build-time option.
1906 7. Some older OS have a limit of 256 on the maximum number of file
1907 descriptors. Exim was using setrlimit() to set 1000 as a large value
1908 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1909 systems. I've change it so that if it can't get 1000, it tries for 256.
1911 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1912 was an oversight, and furthermore, ever since the addition of extra
1913 controls (e.g. 4.43/32), the checks on when to allow different forms of
1914 "control" were broken. There should now be diagnostics for all cases when a
1915 control that does not make sense is encountered.
1917 9. Added the /retain_sender option to "control=submission".
1919 10. $recipients is now available in the predata ACL (oversight).
1921 11. Tidy the search cache before the fork to do a delivery from a message
1922 received from the command line. Otherwise the child will trigger a lookup
1923 failure and thereby defer the delivery if it tries to use (for example) a
1924 cached ldap connection that the parent has called unbind on.
1926 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1927 of $address_data from the recipient verification was clobbered by the
1928 sender verification.
1930 13. The value of address_data from a sender verification is now available in
1931 $sender_address_data in subsequent conditions in the ACL statement.
1933 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1935 15. Added a new option "connect=<time>" to callout options, to set a different
1938 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1939 was its contents. (It was OK if the option was not defined at all.)
1941 17. A "Completed" log line is now written for messages that are removed from
1942 the spool by the -Mrm option.
1944 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1945 information about exactly what failed.
1947 19. Added -dd to debug only the daemon process.
1949 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1950 handles timeouts, both on the server side and network timeouts. Renamed the
1951 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1953 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1954 It is now set to "smtps".
1956 22. $host_address is now set to the target address during the checking of
1957 ignore_target_hosts.
1959 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1960 being passed; this would have caused $sender_host_name to have been used if
1961 matching the list had actually called for a host name (not very likely,
1962 since this list is usually IP addresses). A host name is now passed as
1965 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1966 code that specifies a non-restarting handler (typically sigaction() in
1967 modern systems) in an attempt to fix a rare and obscure crash bug.
1969 25. Narrowed the window for a race in the daemon that could cause it to ignore
1970 SIGCHLD signals. This is not a major problem, because they are used only to
1971 wake it up if nothing else does.
1973 26. A malformed maildirsize file could cause Exim to calculate negative values
1974 for the mailbox size or file count. Odd effects could occur as a result.
1975 The maildirsize information is now recalculated if the size or filecount
1978 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1979 support for a long time. Removed HAVE_SYS_VFS_H.
1981 28. Installed the latest version of exipick from John Jetmore.
1983 29. In an address list, if the pattern was not a regular expression, an empty
1984 subject address (from a bounce message) matched only if the pattern was an
1985 empty string. Non-empty patterns were not even tested. This was the wrong
1986 because it is perfectly reasonable to use an empty address as part of a
1987 database query. An empty address is now tested by patterns that are
1988 lookups. However, all the other forms of pattern expect the subject to
1989 contain a local part and a domain, and therefore, for them, an empty
1990 address still always fails if the pattern is not itself empty.
1992 30. Exim went into a mad DNS loop when attempting to do a callout where the
1993 host was specified on an smtp transport, and looking it up yielded more
1994 than one IP address.
1996 31. Re-factored the code for checking spool and log partition space into a
1997 function that finds that data and another that does the check. The former
1998 is then used to implement four new variables: $spool_space, $log_space,
1999 $spool_inodes, and $log_inodes.
2001 32. The RFC2047 encoding function was originally intended for short strings
2002 such as real names; it was not keeping to the 75-character limit for
2003 encoded words that the RFC imposes. It now respects the limit, and
2004 generates multiple encoded words if necessary. To be on the safe side, I
2005 have increased the buffer size for the ${rfc2047: expansion operator from
2008 33. It is now permitted to omit both strings after an "if" condition; if the
2009 condition is true, the result is "true". As before, when the second string
2010 is omitted, a false condition yields an empty string. This makes it less
2011 cumbersome to write custom ACL and router conditions.
2013 34. Failure to deliver a bounce message always caused it to be frozen, even if
2014 there was an errors_to setting on the router. The errors_to setting is now
2017 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2018 canonical form (fully expanded) before being placed in
2019 $sender_host_address.
2021 36. The table in the code that translates DNS record types into text (T_A to
2022 "A" for instance) was missing entries for NS and CNAME. It is just possible
2023 that this could have caused confusion if both these types were looked up
2024 for the same domain, because the text type is used as part of Exim's
2025 per-process caching. But the chance of anyone hitting this buglet seems
2028 37. The dnsdb lookup has been extended in a number of ways.
2030 (1) There is a new type, "zns", which walks up the domain tree until it
2031 finds some nameserver records. It should be used with care.
2033 (2) There is a new type, "mxh", which is like "mx" except that it returns
2034 just the host names, not the priorities.
2036 (3) It is now possible to give a list of domains (or IP addresses) to be
2037 looked up. The behaviour when one of the lookups defers can be
2038 controlled by a keyword.
2040 (4) It is now possible to specify the separator character for use when
2041 multiple records are returned.
2043 38. The dnslists ACL condition has been extended: it is now possible to supply
2044 a list of IP addresses and/or domains to be looked up in a particular DNS
2047 39. Added log_selector=+queue_time_overall.
2049 40. When running the queue in the test harness, wait just a tad after forking a
2050 delivery process, to get repeatability of debugging output.
2052 41. Include certificate and key file names in error message when GnuTLS fails
2053 to set them up, because the GnuTLS error message doesn't include the name
2054 of the failing file when there is a problem reading it.
2056 42. Allow both -bf and -bF in the same test run.
2058 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2060 44. The "Exiscan patch" is now merged into the mainline Exim source.
2062 45. Sometimes the final signoff response after QUIT could fail to get
2063 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2064 < 0 before doing a fflush(). This bug looks as though it goes back to the
2065 introduction of TLS in release 3.20, but "sometimes" must have been rare
2066 because the tests only now provoked it.
2068 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2069 (this can affect the format of dates).
2071 47. exim_tidydb, when checking for the continued existence of a message for
2072 which it has found a message-specific retry record, was not finding
2073 messages that were in split spool directories. Consequently, it was
2074 deleting retry records that should have stayed in existence.
2076 48. Steve fixed some bugs in eximstats.
2078 49. The SPA authentication driver was not abandoning authentication and moving
2079 on to the next authenticator when an expansion was forced to fail,
2080 contradicting the general specification for all authenticators. Instead it
2081 was generating a temporary error. It now behaves as specified.
2083 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2084 (the order specifies the preference for clients). The order is now AES256,
2085 AES128, 3DES, ARCFOUR128.
2087 51. Small patch to Sieve code - explicitly set From: when generating an
2090 52. Exim crashed if a remote delivery caused a very long error message to be
2091 recorded - for instance if somebody sent an entire SpamAssassin report back
2092 as a large number of 550 error lines. This bug was coincidentally fixed by
2093 increasing the size of one of Exim's internal buffers (big_buffer) that
2094 happened as part of the Exiscan merge. However, to be on the safe side, I
2095 have made the code more robust (and fixed the comments that describe what
2098 53. Now that there can be additional text after "Completed" in log lines (if
2099 the queue_time_overall log selector is set), a one-byte patch to exigrep
2100 was needed to allow it to recognize "Completed" as not the last thing in
2103 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2104 patch that reportedly fixes this has been added. I am not expert enough to
2105 create a test for it. This is what the patch creator wrote:
2107 "I found a little strange behaviour of ldap code when working with
2108 Windows 2003 AD Domain, where users was placed in more than one
2109 Organization Units. When I tried to give exim partial DN, the exit code
2110 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2111 But simultaneously result of request was absolutely normal ldap result,
2112 so I produce this patch..."
2114 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2115 REFERENCE, so I have modified the code to exclude the patch when that macro
2118 55. Some experimental protocols are using DNS PTR records for new purposes. The
2119 keys for these records are domain names, not reversed IP addresses. The
2120 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2121 leaves it alone. Component reversal etc. now happens only for IP addresses.
2124 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2126 57. Double the size of the debug message buffer (to 2048) so that more of very
2127 long debug lines gets shown.
2129 58. The exicyclog utility now does better if the number of log files to keep
2130 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2132 59. Two changes related to the smtp_active_hostname option:
2134 (1) $smtp_active_hostname is now available as a variable.
2135 (2) The default for smtp_banner uses $smtp_active_hostname instead
2136 of $primary_hostname.
2138 60. The host_aton() function is supposed to be passed a string that is known
2139 to be a valid IP address. However, in the case of IPv6 addresses, it was
2140 not checking this. This is a hostage to fortune. Exim now panics and dies
2141 if the condition is not met. A case was found where this could be provoked
2142 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2143 components; fortuitously, this particular loophole had already been fixed
2144 by change 4.50/55 above.
2146 If there are any other similar loopholes, the new check in host_aton()
2147 itself should stop them being exploited. The report I received stated that
2148 data on the command line could provoke the exploit when Exim was running as
2149 exim, but did not say which command line option was involved. All I could
2150 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2151 running as the user.
2154 61. There was a buffer overflow vulnerability in the SPA authentication code
2155 (which came originally from the Samba project). I have added a test to the
2156 spa_base64_to_bits() function which I hope fixes it.
2159 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2160 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2162 63. The daemon start-up calls getloadavg() while still root for those OS that
2163 need the first call to be done as root, but it missed one case: when
2164 deliver_queue_load_max is set with deliver_drop_privilege. This is
2165 necessary for the benefit of the queue runner, because there is no re-exec
2166 when deliver_drop_privilege is set.
2168 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2169 This has been fixed.
2171 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2172 was in use, was not putting the data itself into the right store pool;
2173 consequently, it could be overwritten for a subsequent message in the same
2174 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2177 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2179 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2180 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2181 the function were treating the return as a boolean value, which happened to
2182 work because 0=false and not-0=true, but is not correct code.
2184 68. The host_aton() function was not handling scoped IPv6 addresses (those
2185 with, for example, "%eth0" on the end) correctly.
2187 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2188 negated items (that is, ~something) in unsigned ints. Some compilers
2189 apparently mutter when there is no cast.
2191 70. If an address verification called from an ACL failed, and did not produce a
2192 user-specific message (i.e. there was only a "system" message), nothing was
2193 put in $acl_verify_message. In this situation, it now puts the system
2196 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2197 session; change 4.31/43 added the unwanted input to the log line - except
2198 that it did not do this in the start of session case. It now does.
2200 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2201 This is wrong and can cause the other end to generate a synchronization
2202 error if it is another Exim or anything else that does the synchronization
2203 check. A QUIT command is no longer sent after a timeout.
2205 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2206 during host lookups.
2208 74. The defer_ok option of callout verification was not working if it was used
2209 when verifying addresses in header lines, that is, for this case:
2211 verify = header_sender/callout=defer_ok
2213 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2214 those file descriptors could be used for SMTP connections. If anything
2215 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2216 could be sent to the SMTP client, causing chaos. The daemon now opens
2217 stdin, stdout, and stderr to /dev/null when it puts itself into the
2220 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2221 log by default. The user can override this with suitable Perl magic.
2223 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2224 the log message when discard triggers, was not working for the DATA ACL or
2225 for the non-SMTP ACL.
2227 78. Error message wording change in sieve.c.
2229 79. If smtp_accept_max_per_host was set, the number of connections could be
2230 restricted to fewer than expected, because the daemon was trying to set up
2231 a new connection before checking whether the processes handling previous
2232 connections had finished. The check for completed processes is now done
2233 earlier. On busy systems, this bug wouldn't be noticed because something
2234 else would have woken the daemon, and it would have reaped the completed
2237 80. If a message was submitted locally by a user whose login name contained one
2238 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2239 It caused a spool format error. I have fixed the spool reading code. A
2240 related problem was that the "from" clause in the Received: line became
2241 illegal because of the space(s). It is now covered by ${quote_local_part.
2243 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2246 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2248 83. Patch from Sieve maintainer to fix unterminated string problem in
2249 "vacation" handling.
2251 84. Some minor changes to the Linux configuration files to help with other
2252 OS variants using glibc.
2254 85. One more patch for Sieve to update vacation handling to latest spec.
2257 ----------------------------------------------------
2258 See the note above about the 4.44 and 4.50 releases.
2259 ----------------------------------------------------
2265 1. Change 4.43/35 introduced a bug that caused file counts to be
2266 incorrectly computed when quota_filecount was set in an appendfile
2269 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2270 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2273 3. Give more explanation in the error message when the command for a transport
2274 filter fails to execute.
2276 4. There are several places where Exim runs a non-Exim command in a
2277 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2278 was being done only for the command run by the queryprogram router. It is
2279 now done for all such subprocesses. The other cases are: ${run, transport
2280 filters, and the commands run by the lmtp and pipe transports.
2282 5. Some older OS have a limit of 256 on the maximum number of file
2283 descriptors. Exim was using setrlimit() to set 1000 as a large value
2284 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2285 systems. I've change it so that if it can't get 1000, it tries for 256.
2287 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2288 was an oversight, and furthermore, ever since the addition of extra
2289 controls (e.g. 4.43/32), the checks on when to allow different forms of
2290 "control" were broken. There should now be diagnostics for all cases when a
2291 control that does not make sense is encountered.
2293 7. $recipients is now available in the predata ACL (oversight).
2295 8. Tidy the search cache before the fork to do a delivery from a message
2296 received from the command line. Otherwise the child will trigger a lookup
2297 failure and thereby defer the delivery if it tries to use (for example) a
2298 cached ldap connection that the parent has called unbind on.
2300 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2301 of $address_data from the recipient verification was clobbered by the
2302 sender verification.
2304 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2305 was its contents. (It was OK if the option was not defined at all.)
2307 11. A "Completed" log line is now written for messages that are removed from
2308 the spool by the -Mrm option.
2310 12. $host_address is now set to the target address during the checking of
2311 ignore_target_hosts.
2313 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2314 being passed; this would have caused $sender_host_name to have been used if
2315 matching the list had actually called for a host name (not very likely,
2316 since this list is usually IP addresses). A host name is now passed as
2319 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2320 code that specifies a non-restarting handler (typically sigaction() in
2321 modern systems) in an attempt to fix a rare and obscure crash bug.
2323 15. Narrowed the window for a race in the daemon that could cause it to ignore
2324 SIGCHLD signals. This is not a major problem, because they are used only to
2325 wake it up if nothing else does.
2327 16. A malformed maildirsize file could cause Exim to calculate negative values
2328 for the mailbox size or file count. Odd effects could occur as a result.
2329 The maildirsize information is now recalculated if the size or filecount
2332 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2333 support for a long time. Removed HAVE_SYS_VFS_H.
2335 18. Updated exipick to current release from John Jetmore.
2337 19. Allow an empty sender to be matched against a lookup in an address list.
2338 Previously the only cases considered were a regular expression, or an
2341 20. Exim went into a mad DNS lookup loop when doing a callout where the
2342 host was specified on the transport, if the DNS lookup yielded more than
2345 21. The RFC2047 encoding function was originally intended for short strings
2346 such as real names; it was not keeping to the 75-character limit for
2347 encoded words that the RFC imposes. It now respects the limit, and
2348 generates multiple encoded words if necessary. To be on the safe side, I
2349 have increased the buffer size for the ${rfc2047: expansion operator from
2352 22. Failure to deliver a bounce message always caused it to be frozen, even if
2353 there was an errors_to setting on the router. The errors_to setting is now
2356 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2357 canonical form (fully expanded) before being placed in
2358 $sender_host_address.
2360 24. Updated eximstats to version 1.33
2362 25. Include certificate and key file names in error message when GnuTLS fails
2363 to set them up, because the GnuTLS error message doesn't include the name
2364 of the failing file when there is a problem reading it.
2366 26. Expand error message when OpenSSL has problems setting up cert/key files.
2369 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2370 (this can affect the format of dates).
2372 28. exim_tidydb, when checking for the continued existence of a message for
2373 which it has found a message-specific retry record, was not finding
2374 messages that were in split spool directories. Consequently, it was
2375 deleting retry records that should have stayed in existence.
2377 29. eximstats updated to version 1.35
2378 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2379 1.35 - bugfix such that pie charts by volume are generated correctly
2381 30. The SPA authentication driver was not abandoning authentication and moving
2382 on to the next authenticator when an expansion was forced to fail,
2383 contradicting the general specification for all authenticators. Instead it
2384 was generating a temporary error. It now behaves as specified.
2386 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2387 (the order specifies the preference for clients). The order is now AES256,
2388 AES128, 3DES, ARCFOUR128.
2390 31. Small patch to Sieve code - explicitly set From: when generating an
2393 32. Exim crashed if a remote delivery caused a very long error message to be
2394 recorded - for instance if somebody sent an entire SpamAssassin report back
2395 as a large number of 550 error lines. This bug was coincidentally fixed by
2396 increasing the size of one of Exim's internal buffers (big_buffer) that
2397 happened as part of the Exiscan merge. However, to be on the safe side, I
2398 have made the code more robust (and fixed the comments that describe what
2401 33. Some experimental protocols are using DNS PTR records for new purposes. The
2402 keys for these records are domain names, not reversed IP addresses. The
2403 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2404 leaves it alone. Component reversal etc. now happens only for IP addresses.
2407 34. The host_aton() function is supposed to be passed a string that is known
2408 to be a valid IP address. However, in the case of IPv6 addresses, it was
2409 not checking this. This is a hostage to fortune. Exim now panics and dies
2410 if the condition is not met. A case was found where this could be provoked
2411 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2412 components; fortuitously, this particular loophole had already been fixed
2413 by change 4.50/55 or 4.44/33 above.
2415 If there are any other similar loopholes, the new check in host_aton()
2416 itself should stop them being exploited. The report I received stated that
2417 data on the command line could provoke the exploit when Exim was running as
2418 exim, but did not say which command line option was involved. All I could
2419 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2420 running as the user.
2423 35. There was a buffer overflow vulnerability in the SPA authentication code
2424 (which came originally from the Samba project). I have added a test to the
2425 spa_base64_to_bits() function which I hope fixes it.
2428 36. The daemon start-up calls getloadavg() while still root for those OS that
2429 need the first call to be done as root, but it missed one case: when
2430 deliver_queue_load_max is set with deliver_drop_privilege. This is
2431 necessary for the benefit of the queue runner, because there is no re-exec
2432 when deliver_drop_privilege is set.
2434 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2435 was in use, was not putting the data itself into the right store pool;
2436 consequently, it could be overwritten for a subsequent message in the same
2437 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2440 38. Sometimes the final signoff response after QUIT could fail to get
2441 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2442 < 0 before doing a fflush(). This bug looks as though it goes back to the
2443 introduction of TLS in release 3.20, but "sometimes" must have been rare
2444 because the tests only now provoked it.
2450 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2451 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2452 Now it returns an integer. A number of calls were still expecting a T/F
2453 return. Fortuitously, in all cases, the tests worked in OK situations,
2454 which is the norm. However, things would have gone wrong on any write
2455 failures on the smtp file descriptor. This function is used when sending
2456 messages over SMTP and also when doing verify callouts.
2458 2. When Exim is called to do synchronous delivery of a locally submitted
2459 message (the -odf or -odi options), it no longer closes stderr before doing
2462 3. Implemented the mua_wrapper option.
2464 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2466 5. Implemented the functions header_remove(), header_testname(),
2467 header_add_at_position(), and receive_remove_recipient(), and exported them
2470 6. If an ACL "warn" statement specified the addition of headers, Exim already
2471 inserted X-ACL-Warn: at the start if there was no header name. However, it
2472 was not making this test for the second and subsequent header lines if
2473 there were newlines in the string. This meant that an invalid header could
2474 be inserted if Exim was badly configured.
2476 7. Allow an ACL "warn" statement to add header lines at the start or after all
2477 the Received: headers, as well as at the end.
2479 8. Added the rcpt_4xx retry error code.
2481 9. Added postmaster_mailfrom=xxx to callout verification option.
2483 10. Added mailfrom=xxxx to the callout verification option, for verify=
2486 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2487 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2489 12. Inserted some casts to stop certain compilers warning when using pointer
2490 differences as field lengths or precisions in printf-type calls (mostly
2491 affecting debugging statements).
2493 13. Added optional readline() support for -be (dynamically loaded).
2495 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2496 same clock tick as a message's arrival, so that its received time was the
2497 same as the "first fail" time on the retry record, and that message
2498 remained on the queue past the ultimate address timeout, every queue runner
2499 would try a delivery (because it was past the ultimate address timeout) but
2500 after another failure, the ultimate address timeout, which should have then
2501 bounced the address, did not kick in. This was a "< instead of <=" error;
2502 in most cases the first failure would have been in the next clock tick
2503 after the received time, and all would be well.
2505 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2506 being recognized when the domain list was tested by the match_domain
2507 condition in an expansion string.
2509 16. Added the ${str2b64: operator.
2511 17. Exim was always calling setrlimit() to set a large limit for the number of
2512 processes, without checking whether the existing limit was already
2513 adequate. (It did check for the limit on file descriptors.) Furthermore,
2514 errors from getrlimit() and setrlimit() were being ignored. Now they are
2515 logged to the main and panic logs, but Exim does carry on, to try to do its
2516 job under whatever limits there are.
2518 18. Imported PCRE 5.0.
2520 19. Trivial typo in log message " temporarily refused connection" (the leading
2523 20. If the log selector return_path_on_delivery was set and an address was
2524 redirected to /dev/null, the delivery process crashed because it assumed
2525 that a return path would always be set for a "successful" delivery. In this
2526 case, the whole delivery is bypassed as an optimization, and therefore no
2529 21. Internal re-arrangement: the function for sending a challenge and reading
2530 a response while authentication was assuming a zero-terminated challenge
2531 string. It's now changed to take a pointer and a length, to allow for
2532 binary data in such strings.
2534 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2536 23. Exim was not respecting finduser_retries when seeking the login of the
2537 uid under which it was called; it was always trying 10 times. (The default
2538 setting of finduser_retries is zero.) Also, it was sleeping after the final
2539 failure, which is pointless.
2541 24. Implemented tls_on_connect_ports.
2543 25. Implemented acl_smtp_predata.
2545 26. If the domain in control=submission is set empty, Exim assumes that the
2546 authenticated id is a complete email address when it generates From: or
2547 Sender: header lines.
2549 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2550 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2551 chown and chgrp in /bin and hostname in /usr/bin.
2553 28. Exim was keeping the "process log" file open after each use, just as it
2554 does for the main log. This opens the possibility of it remaining open for
2555 long periods when the USR1 signal hits a daemon. Occasional processlog
2556 errors were reported, that could have been caused by this. Anyway, it seems
2557 much more sensible not to leave this file open at all, so that is what now
2560 29. The long-running daemon process does not normally write to the log once it
2561 has entered its main loop, and it closes the log before doing so. This is
2562 so that log files can straightforwardly be renamed and moved. However,
2563 there are a couple of unusual error situations where the daemon does write
2564 log entries, and I had neglected to close the log afterwards.
2566 30. The text of an SMTP error response that was received during a remote
2567 delivery was being truncated at 512 bytes. This is too short for some of
2568 the long messages that one sometimes sees. I've increased the limit to
2571 31. It is now possible to make retry rules that apply only when a message has a
2572 specific sender, in particular, an empty sender.
2574 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2575 it possible to be selective about when SMTP synchronization is enforced.
2577 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2579 32. Implemented hosts_connection_nolog.
2581 33. Added an ACL for QUIT.
2583 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2586 35. Added mailbox_size and mailbox_filecount to appendfile.
2588 36. Added control = no_multiline_responses to ACLs.
2590 37. There was a bug in the logic of the code that waits for the clock to tick
2591 in the case where the clock went backwards by a substantial amount such
2592 that the microsecond fraction of "now" was more than the microsecond
2593 fraction of "then" (but the whole seconds number was less).
2595 38. Added support for the libradius Radius client library this is found on
2596 FreeBSD (previously only the radiusclient library was supported).
2602 1. When certain lookups returned multiple values in the form name=value, the
2603 quoting of the values was not always being done properly. Specifically:
2604 (a) If the value started with a double quote, but contained no whitespace,
2606 (b) If the value contained whitespace other than a space character (i.e.
2607 tabs or newlines or carriage returns) it was not quoted.
2608 This fix has been applied to the mysql and pgsql lookups by writing a
2609 separate quoting function and calling it from the lookup code. The fix
2610 should probably also be applied to nisplus, ibase and oracle lookups, but
2611 since I cannot test any of those, I have not disturbed their existing code.
2613 2. A hit in the callout cache for a specific address caused a log line with no
2614 reason for rejecting RCPT. Now it says "Previous (cached) callout
2615 verification failure".
2617 3. There was an off-by-one bug in the queryprogram router. An over-long
2618 return line was truncated at 256 instead of 255 characters, thereby
2619 overflowing its buffer with the terminating zero. As well as fixing this, I
2620 have increased the buffer size to 1024 (and made a note to document this).
2622 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2623 when Exim is waiting for an SMTP response from a remote server, Exim
2624 restarts its select() call on the socket, thereby resetting its timeout.
2625 This is not a problem when such interrupts are rare. Somebody set up a cron
2626 job to run exiwhat every 2 minutes, which is less than the normal select()
2627 timeout (5 or 10 minutes). This meant that the select() timeout never
2628 kicked in because it was always reset. I have fixed this by comparing the
2629 time when an interrupt arrives with the time at the start of the first call
2630 to select(). If more time than the timeout has elapsed, the interrupt is
2631 treated as a timeout.
2633 5. Some internal re-factoring in preparation for the addition of Sieve
2634 extensions (by MH). In particular, the "personal" test is moved to a
2635 separate function, and given an option for scanning Cc: and Bcc: (which is
2636 not set for Exim filters).
2638 6. When Exim created an email address using the login of the caller as the
2639 local part (e.g. when creating a From: or Sender: header line), it was not
2640 quoting the local part when it contained special characters such as @.
2642 7. Installed new OpenBSD configuration files.
2644 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2645 try to make them clearer.
2647 9. Callout options, other than the timeout value, were being ignored when
2648 verifying sender addresses in header lines. For example, when using
2650 verify = header_sender/callout=no_cache
2652 the cache was (incorrectly) being used.
2654 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2655 only the Cygwin environment.
2657 11. When return_path_on_delivery was set as a log selector, if different remote
2658 addresses in the same message used different return paths and parallel
2659 remote delivery occurred, the wrong values would sometimes be logged.
2660 (Whenever a remote delivery process finished, the return path value from
2661 the most recently started remote delivery process was logged.)
2663 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2664 lines when AUTH and/or TLS are in use. This is the "received protocol"
2665 field. Exim used to use "asmtp" for authenticated SMTP, without any
2666 indication (in the protocol name) for TLS use. Now it follows the RFC and
2667 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2668 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2669 names appear in log lines as well as in Received: header lines.
2671 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2672 extensions, and comparison tests, and to fix some bugs.
2674 14. Changes to the "personal" filter test:
2676 (1) The test was buggy in that it was just doing the equivalent of
2677 "contains" tests on header lines. For example, if a user's address was
2678 anne@some.where, the "personal" test would incorrectly be true for
2680 To: susanne@some.where
2682 This test is now done by extracting each address from the header in turn,
2683 and checking the entire address. Other tests that are part of "personal"
2684 are now done using regular expressions (for example, to check local parts
2685 of addresses in From: header lines).
2687 (2) The list of non-personal local parts in From: addresses has been
2688 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2689 taken from the Sieve specification recommendations.
2691 (3) If the message contains any header line starting with "List-" it is
2692 treated as non-personal.
2694 (4) The test for "circular" in the Subject: header line has been removed
2695 because it now seems ill-conceived.
2697 15. Minor typos in src/EDITME comments corrected.
2699 16. Installed latest exipick from John Jetmore.
2701 17. If headers_add on a router specified a text string that was too long for
2702 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2703 of string_sprintf() is now avoided.
2705 18. $message_body_size was not set (it was always zero) when running the DATA
2706 ACL and the local_scan() function.
2708 19. For the "mail" command in an Exim filter, no default was being set for
2709 the once_repeat time, causing a random time value to be used if "once" was
2710 specified. (If the value happened to be <= 0, no repeat happened.) The
2711 default is now 0s, meaning "never repeat". The "vacation" command was OK
2712 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2713 (I found it when inspecting the code).
2715 20. There is now an overall timeout for performing a callout verification. It
2716 defaults to 4 times the callout timeout, which applies to individual SMTP
2717 commands during the callout. The overall timeout applies when there is more
2718 than one host that can be tried. The timeout is checked before trying the
2719 next host. This prevents very long delays if there are a large number of
2720 hosts and all are timing out (e.g. when the network connections are timing
2721 out). The value of the overall timeout can be changed by specifying an
2722 additional sub-option for "callout", called "maxwait". For example:
2724 verify = sender/callout=5s,maxwait=20s
2726 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2727 to the end before writing, but this should make it even safer).
2729 22. Exim was forgetting that it had advertised PIPELINING for the second and
2730 subsequent messages on an SMTP connection. It was also not resetting its
2731 memory on STARTTLS and an internal HELO.
2733 23. When Exim logs an SMTP synchronization error within a session, it now
2734 records whether PIPELINING has been advertised or not.
2736 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2737 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2738 rather than long int.
2740 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2742 26. Added the never_mail option to autoreply.
2748 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2749 crash if the getsockname() call failed; this can happen if a connection is
2750 closed very soon after it is established. The problem was simply in the
2751 order in which certain operations were done, causing Exim to try to write
2752 to the SMTP stream before it had set up the file descriptor. The bug has
2753 been fixed by making things happen in the correct order.
2759 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2760 before the connection was closed, thus losing the rejection response.
2762 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2763 some early Solaris releases, but causes trouble in current releases where
2764 socklen_t is defined.
2766 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2769 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2772 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2773 that releasing the top of it at the end releases what was used for sub-
2774 expansions (unless the block got too big). However, discard this block if
2775 the first thing is a variable or header, so that we can use its block when
2776 it is dynamic (useful for very large $message_headers, for example).
2778 6. Lookups now cache *every* query, not just the most recent. A new, separate
2779 store pool is used for this. It can be recovered when all lookup caches are
2780 flushed. Lookups now release memory at the end of their result strings.
2781 This has involved some general refactoring of the lookup sources.
2783 7. Some code has been added to the store_xxx() functions to reduce the amount
2784 of flapping under certain conditions.
2786 8. log_incoming_interface used to affect only the <= reception log lines. Now
2787 it causes the local interface and port to be added to several more SMTP log
2788 lines, for example "SMTP connection from", and rejection lines.
2790 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2792 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2794 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2795 could be overwritten at the end of the current message (or the start of a
2796 new message if it was set in a HELO ACL). The value is now preserved for
2797 the duration of the SMTP connection.
2799 12. If a transport had a headers_rewrite setting, and a matching header line
2800 contained an unqualified address, that address was qualified, even if it
2801 did not match any rewriting rules. The underlying bug was that the values
2802 of the flags that permit the existence of unqualified sender and recipient
2803 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2804 non-local messages, and by -bnq for local messages) were not being
2805 preserved with the message after it was received.
2807 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2808 "next input=" as part of the text comprising the host identity instead of
2809 the correct text. The code was using the same buffer for two different
2810 strings. However, depending on which order the printing function evaluated
2811 its arguments, the bug did not always show up. Under Linux, for example, my
2812 test suite worked just fine.
2814 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2815 doesn't work with some older versions of Perl. It has been changed to "my",
2816 which in any case is probably the better facility to use.
2818 15. A really picky compiler found some instances of statements for creating
2819 error messages that either had too many or two few arguments for the format
2822 16. The size of the buffer for calls to the DNS resolver has been increased
2823 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2824 for addresses that have a lot of PTR records. This alleviates a problem; it
2825 does not fully solve it.
2827 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2828 buffer now truncates the list and logs the incident, which is the same
2829 action as happens when Exim is looking up a host name and its aliases.
2830 Previously in this situation something unpredictable would happen;
2831 sometimes it was "internal error: store_reset failed".
2833 18. If a server dropped the connection unexpectedly when an Exim client was
2834 using GnuTLS and trying to read a response, the client delivery process
2835 crashed while trying to generate an error log message.
2837 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2838 string, for example:
2840 warn message = H1: something\nH2: something
2842 the text was added as a single header line from Exim's point of view
2843 though it ended up OK in the delivered message. However, searching for the
2844 second and subsequent header lines using $h_h2: did not work. This has been
2845 fixed. Similarly, if a system filter added multiple headers in this way,
2846 the routers could not see them.
2848 20. Expanded the error message when iplsearch is called with an invalid key to
2849 suggest using net-iplsearch in a host list.
2851 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2852 ACLs are no longer actually imposed (and a message to that effect is
2855 22. If a "gecos" field in a passwd entry contained escaped characters, in
2856 particular, if it contained a \" sequence, Exim got it wrong when building
2857 a From: or a Sender: header from that name. A second bug also caused
2858 incorrect handling when an unquoted " was present following a character
2859 that needed quoting.
2861 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2862 was not being matched caselessly.
2864 24. Arranged for all hyphens in the exim.8 source to be escaped with
2867 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2868 properly. Recipient callout cache records were still being keyed to include
2869 the sender, even when use_sender was set false. This led to far more
2870 callouts that were necessary. The sender is no longer included in the key
2871 when use_sender is false.
2873 26. Added "control = submission" modifier to ACLs.
2875 27. Added the ${base62d: operator to decode base 62 numbers.
2877 28. dnsdb lookups can now access SRV records.
2879 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2880 the configuration file.
2882 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2883 (-v) mode. This makes the output for a verbose queue run more intelligible.
2885 31. Added a use_postmaster feature to recipient callouts.
2887 32. Added the $body_zerocount variable, containing the number of binary zero
2888 bytes in the message body.
2890 33. The time of last modification of the "new" subdirectory is now used as the
2891 "mailbox time last read" when there is a quota error for a maildir
2894 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2896 35. Added +ignore_unknown as a special item in host lists.
2898 36. Code for decoding IPv6 addresses in host lists is now included, even if
2899 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2900 address was recognized as an IP address, but was then not correctly decoded
2901 into binary, causing unexpected and incorrect effects when compared with
2908 1. Very minor rewording of debugging text in manualroute to say "list of
2909 hosts" instead of "hostlist".
2911 2. If verify=header_syntax was set, and a header line with an unqualified
2912 address (no domain) and a large number of spaces between the end of the
2913 name and the colon was received, the reception process suffered a buffer
2914 overflow, and (when I tested it) crashed. This was caused by some obsolete
2915 code that should have been removed. The fix is to remove it!
2917 3. When running in the test harness, delay a bit after writing a bounce
2918 message to get a bit more predictability in the log output.
2920 4. Added a call to search_tidyup() just before forking a reception process. In
2921 theory, someone could use a lookup in the expansion of smtp_accept_max_
2922 per_host which, without the tidyup, could leave open a database connection.
2924 5. Added the variables $recipient_data and $sender_data which get set from a
2925 lookup success in an ACL "recipients" or "senders" condition, or a router
2926 "senders" option, similar to $domain_data and $local_part_data.
2928 6. Moved the writing of debug_print from before to after the "senders" test
2931 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2932 problems for message scanning, either using a data ACL, or using
2933 local_scan() because the Received: header was not generated till after they
2934 were called (in order to set the time as the time of reception completion).
2935 I have revised the way this works. The header is now generated after the
2936 body is received, but before the ACL or local_scan() are called. After they
2937 are run, the timestamp in the header is updated.
2943 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2944 before starting a queue runner without re-exec. This happened only when
2945 deliver_drop_privilege was set or when the Exim user was set to root. The
2946 effect of the bug was that timeouts during subsequent deliveries caused
2947 crashes instead of being properly handled. The handler is now left at its
2948 default (and expected) setting.
2950 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2951 message, again when deliver_drop_privilege is set or Exim is run as root.
2952 The bug described in (1) was not present in this case, but the tidying up
2953 of the other signals was missing. I have made the two cases consistent.
2955 3. The ignore_target_hosts setting on a manualroute router was being ignored
2956 for hosts that were looked up using the /MX notation.
2958 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2961 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2962 operated on the sender address. After changing the $sender_address to <>
2963 for the sender address verify, Exim was re-instated it as the original
2964 (before rewriting) address, but remembering that it had rewritten it, so it
2965 wasn't rewriting it again. This bug also had the effect of breaking the
2966 sender address verification caching when the sender address was rewritten.
2968 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2969 This has been changed so that if the ip literal address matches
2970 ignore_target_hosts, the router declines.
2972 7. Added expansion conditions match_domain, match_address, and match_local_
2973 part (NOT match_host).
2975 8. The placeholder for the Received: header didn't have a length field set.
2977 9. Added code to Exim itself and to exim_lock to test for a specific race
2978 condition that could lead to file corruption when using MBX delivery. The
2979 issue is with the lockfile that is created in /tmp. If this file is removed
2980 after a process has opened it but before that process has acquired a lock,
2981 there is the potential for a second process to recreate the file and also
2982 acquire a lock. This could lead to two Exim processes writing to the file
2983 at the same time. The added code performs the same test as UW imapd; it
2984 checks after acquiring the lock that its file descriptor still refers to
2985 the same named file.
2987 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2988 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2991 11. Added the smtp_active_hostname option. If used, this will typically be made
2992 to depend on the incoming interface address. Because $interface_address is
2993 not set up until the daemon has forked a reception process, error responses
2994 that can happen earlier (such as "too many connections") no longer contain
2997 12. If an expansion in a condition on a "warn" statement fails because a lookup
2998 defers, the "warn" statement is abandoned, and the next ACL statement is
2999 processed. Previously this caused the whole ACL to be aborted.
3001 13. Added the iplsearch lookup type.
3003 14. Added ident_timeout as a log selector.
3005 15. Added tls_certificate_verified as a log selector.
3007 16. Added a global option tls_require_ciphers (compare the smtp transport
3008 option of the same name). This controls incoming TLS connections.
3010 17. I finally figured out how to make tls_require_ciphers do a similar thing
3011 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3012 before starting the TLS session.
3014 18. Tabs are now shown as \t in -bP output.
3016 19. If the log selector return_path_on_delivery was set, Exim crashed when
3017 bouncing a message because it had too many Received: header lines.
3019 20. If two routers both had headers_remove settings, and the first one included
3020 a superfluous trailing colon, the final name in the first list and the
3021 first name in the second list were incorrectly joined into one item (with a
3022 colon in the middle).
3028 1. Added -C and -D options to the exinext utility, mainly to make it easier
3029 to include in the automated testing, but these could be helpful when
3030 multiple configurations are in use.
3032 2. The exinext utility was not formatting the output nicely when there was
3033 an alternate port involved in the retry record key, nor when there was a
3034 message id as well (for retries that were specific to a specific message
3035 and a specific host). It was also confused by IPv6 addresses, because of
3036 the additional colons they contain. I have fixed the IPv4 problem, and
3037 patched it up to do a reasonable job for IPv6.
3039 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3040 delivery, the log line now contains "pipelined" if PIPELINING was used.
3042 4. An SMTP transport process used to panic and die if the bind() call to set
3043 an explicit outgoing interface failed. This has been changed; it is now
3044 treated in the same way as a connect() failure.
3046 5. A reference to $sender_host_name in the part of a conditional expansion
3047 that was being skipped was still causing a DNS lookup. This no longer
3050 6. The def: expansion condition was not recognizing references to header lines
3051 that used bh_ and bheader_.
3053 7. Added the _cache feature to named lists.
3055 8. The code for checking quota_filecount in the appendfile transport was
3056 allowing one more file than it should have been.
3058 9. For compatibility with Sendmail, the command line option
3066 and sets the incoming protocol and host name (for trusted callers). The
3067 host name and its colon can be omitted when only the protocol is to be set.
3068 Note the Exim already has two private options, -pd and -ps, that refer to
3069 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3070 "s", but I don't think that's a major issue.
3072 10. A number of refactoring changes to the code, none of which should affect
3075 (a) The number of logging options was getting close to filling up the
3076 32-bit word that was used as a bit map. I have split them into two classes:
3077 those that are passed in the argument to log_write(), and those that are
3078 only ever tested independently outside of that function. These are now in
3079 separate 32-bit words, so there is plenty of room for expansion again.
3080 There is no change in the user interface or the logging behaviour.
3082 (b) When building, for example, log lines, the code previously used a
3083 macro that called string_cat() twice, in order to add two strings. This is
3084 not really sufficiently general. Furthermore, there was one instance where
3085 it was actually wrong because one of the argument was used twice, and in
3086 one call a function was used. (As it happened, calling the function twice
3087 did not affect the overall behaviour.) The macro has been replaced by a
3088 function that can join an arbitrary number of extra strings onto a growing
3091 (c) The code for expansion conditions now uses a table and a binary chop
3092 instead of a serial search (which was left over from when there were very
3093 few conditions). Also, it now recognizes conditions like "pam" even when
3094 the relevant support is not compiled in: a suitably worded error message is
3095 given if an attempt is made to use such a condition.
3097 11. Added ${time_interval:xxxxx}.
3099 12. A bug was causing one of the ddress fields not to be passed back correctly
3100 from remote delivery subprocesses. The field in question was not being
3101 subsequently used, so this caused to problems in practice.
3103 13. Added new log selectors queue_time and deliver_time.
3105 14. Might have fixed a bug in maildirsizefile handling that threw up
3106 "unexpected character" debug warnings, and recalculated the data
3107 unnecessarily. In any case, I expanded the warning message to give more
3110 15. Added the message "Restricted characters in address" to the statements in
3111 the default ACL that block characters like @ and % in local parts.
3113 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3114 Three changes have been made:
3116 (a) There was a serious bug; a negative response to MAIL caused the whole
3117 recipient domain to be cached as invalid, thereby blocking all messages
3118 to all local parts at the same domain, from all senders. This bug has
3119 been fixed. The domain is no longer cached after a negative response to
3120 MAIL if the sender used is not empty.
3122 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3125 (c) A new callout option, "use_sender" has been added for people who want
3126 the modified behaviour.
3132 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3135 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3136 indeed breaks things for older releases.
3138 3. Added additional logging to the case where there is a problem reading data
3139 from a filter that is running in a subprocess using a pipe, in order to
3140 try to track down a specific problem.
3142 4. Testing facility fudge: when running in the test harness and attempting
3143 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3144 getting "No route to host". Convert this to a timeout.
3146 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3149 6. Some OS don't have socklen_t but use size_t instead. This affects the
3150 fifth argument of getsockopt() amongst other things. This is now
3151 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3152 can be set for individual OS. I have set it for SunOS5, OSF1, and
3153 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3154 some earlier ones do not.
3156 7. Change 4.30/15 was not doing the test caselessly.
3158 8. The standard form for an IPv6 address literal was being rejected by address
3159 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3160 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3161 this, as well as the form without the "IPv6" on the front (but only when
3162 address literals are enabled, of course).
3164 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3166 10. Exim crashed if a message with an empty sender address specified by -f
3167 encountered a router with an errors_to setting. This could be provoked only
3168 by a command such as
3172 where an empty string was supplied; "<>" did not hit this bug.
3174 11. Installed PCRE release 4.5.
3176 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3177 remained set. It is now erased.
3179 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3180 times from message ids (which are base 36 rather than the normal 62).
3182 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3183 were being counted as actual protocol errors, and logged if the log
3184 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3185 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3186 and DATA following a set of rejected RCPTs do not count as protocol errors.
3187 In other words, Exim assumes they were pipelined, though this may not
3188 actually be the case. Of course, in all cases the client gets an
3189 appropriate error code.
3191 15. If a lookup fails in an ACL condition, a message about the failure may
3192 be available; it is used if testing the ACL cannot continue, because most
3193 such messages specify what the cause of the deferral is. However, some
3194 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3195 that caused an old message to be retained and used if a later statement
3196 caused a defer, replacing the real cause of the deferral.
3198 16. If an IP address had so many PTR records that the DNS lookup buffer
3199 was not large enough to hold them, Exim could crash while trying to process
3200 the truncated data. It now detects and logs this case.
3202 17. Further to 4.21/58, another change has been made: if (and only if) the
3203 first line of a message (the first header line) ends with CRLF, a bare LF
3204 in a subsequent header line has a space inserted after it, so as not to
3205 terminate the header.
3207 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3208 unnecessarily, used atoi() instead of strtol(), and didn't check the
3209 termination when getting file sizes from file names by regex.
3211 19. Completely re-implemented the support for maildirsize files, in the light
3212 of a number of problems with the previous contributed implementation
3213 (4.30/29). In particular:
3215 . If the quota is zero, the maildirsize file is maintained, but no quota is
3218 . If the maildir directory does not exist, it is created before any attempt
3219 to write a maildirsize file.
3221 . The quota value in the file is just a cache; if the quota is changed in
3222 the transport, the new value overrides.
3224 . A regular expression is available for excluding directories from the
3227 20. The autoreply transport checks the characters in options that define the
3228 message's headers; it allows continued headers, but it was checking with
3229 isspace() after an embedded newline instead of explicitly looking for a
3232 21. If all the "regular" hosts to which an address was routed had passed their
3233 expiry times, and had not reached their retry times, the address was
3234 bounced, even if fallback hosts were defined. Now Exim should go on to try
3237 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3238 equivalent code in the SMTP transport. Some hosts send humungous responses
3239 to HELO/EHLO, more than 1024 it seems.
3241 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3242 gives compiler warnings in some environments. I've now done it "properly",
3245 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3246 (because of problems with the built-in one) was declared to return uschar *
3247 instead of char *, causing compiler failure.
3249 25. Fixed a file descriptor leak when processing alias/forward files.
3251 26. Fixed a minor format string issue in dbfn.c.
3253 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3255 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3256 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3259 29. When a sender address is verified, it is cached, to save repeating the test
3260 when there is more than one recipient in a message. However, when the
3261 verification involves a callout, it is possible for different callout
3262 options to be set for different recipients. It is too complicated to keep
3263 track of this in the cache, so now Exim always runs a verification when a
3264 callout is required, relying on the callout cache for the optimization.
3265 The overhead is duplication of the address routing, but this should not be
3268 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3269 to be verified with callout=postmaster, and the main callout worked but the
3270 postmaster check failed, the verification correctly failed. However, if a
3271 subsequent RCPT command asked for sender verification *without* the
3272 postmaster check, incorrect caching caused this verification also to fail,
3275 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3276 it was not caching the DNS options (qualify_single, search_parents) that
3277 were used when the lookup failed. A subsequent lookup with different
3278 options therefore always gave the same answer, though there were cases
3279 where it should not have. (Example: a "domains = !$mx_any" option on a
3280 dnslookup router: the "domains" option is always processed without any
3281 widening, but the router might have qualify_single set.) Now Exim uses the
3282 cached value only when the same options are set.
3284 32. Added John Jetmore's "exipick" utility to the distribution.
3286 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3287 than a timeout (e.g. a certificate is required, and is not provided), an
3288 Exim server now closes the connection immediately. Previously it waited for
3289 the client to close - but if the client is SSL, it seems that they each
3290 wait for each other, leading to a delay before one of them times out.
3292 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3293 maintained 0.8.x compatibility because I don't think many are using it, and
3294 it is clearly obsolete.
3296 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3299 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3300 client certificate was expired. A simple patch fixes this, though I don't
3301 understand the full logic of why the verify callback is called multiple
3304 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3305 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3306 which causes problems with some clients (such as the Certicom SSL Plus
3307 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3308 disables the coutermeasure allowing Eudora to connect."
3310 38. Exim was not checking that a write() to a log file succeeded. This could
3311 lead to Bad Things if a log got too big, in particular if it hit a file
3312 size limit. Exim now panics and dies if it cannot write to a log file, just
3313 as it does if it cannot open a log file.
3315 39. Modified OS/Makefile-Linux so that it now contains
3317 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3319 The two -D definitions ensure that Exim is compiled with large file
3320 support, which makes it possible to handle log files that are bigger than
3323 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3324 instance) a domain was checked against a named list that involved a lookup,
3325 causing $domain_data to be set, then another domain was checked against the
3326 same list, then the first domain was re-checked, the value of $domain_data
3327 after the final check could be wrong. In particular, if the second check
3328 failed, it could be set empty. This bug probably also applied to
3331 41. The strip_trailing_dot option was not being applied to the address given
3332 with the -f command-line option.
3334 42. The code for reading a message's header from the spool was incrementing
3335 $received_count, but never initializing it. This meant that the value was
3336 incorrect (doubled) while delivering a message in the same process in which
3337 it was received. In the most common configuration of Exim, this never
3338 happens - a fresh exec is done - but it can happen when
3339 deliver_drop_privilege is set.
3341 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3342 it now includes up to 150 characters of the unexpected data in the log
3345 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3346 and building data strings. The size of both of these buffers was 10 000
3347 bytes - far larger than anybody would *ever* want, thought I. Needless to
3348 say, somebody hit the limit. I have increased the maximum line length to
3349 20 000 and the maximum data length of concatenated lines to 100 000. I have
3350 also fixed two bugs, because there was no checking on these buffers. Tsk,
3351 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3352 buffer is too small.
3354 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3355 lsearch lookups. Now it does.
3357 46. When parsing a route_list item in a manualroute router, a fixed-length
3358 buffer was used for the list of hosts. I made this 1024 bytes long,
3359 thinking that nobody would ever have a list of hosts that long. Wrong.
3360 Somebody had a whole pile of complicated expansion conditions, and the
3361 string was silently truncated, leading to an expansion error. It turns out
3362 that it is easier to change to an unlimited length (owing to other changes
3363 that have happened since this code was originally written) than to build
3364 structure for giving a limitation error. The length of the item that
3365 expands into the list of hosts is now unlimited.
3367 47. The lsearch lookup could not handle data where the length of text line was
3368 more than 4095 characters. Such lines were truncated, leading to shortened
3369 data being returned. It should now handle lines of any length.
3371 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3372 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3375 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3376 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3377 no longer happens in comment lines. A list of replacements is now placed
3378 at the head of all of the source files, except those whose only change is
3379 to replace PERL_COMMAND in the very first #! line.
3381 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3382 messages on the queue, with a bottom-up merge sort, using code contributed
3383 by Michael Haardt. This should make operations like -bp somewhat faster on
3384 large queues. It won't affect queue runners, except when queue_run_in_order
3387 51. Installed eximstats 1.31 in the distribution.
3389 52. Added support for SRV lookups to the dnslookup router.
3391 53. If an ACL referred to $message_body or $message_body_end, the value was not
3392 reset for any messages that followed in the same SMTP session.
3394 54. The store-handling optimization for building very long strings was not
3395 differentiating between the different store pools. I don't think this
3396 actually made any difference in practice, but I've tidied it.
3398 55. While running the routers to verify a sender address, $sender_address
3399 was still set to the sender address. This is wrong, because when routing to
3400 send a bounce to the sender, it would be empty. Therefore, I have changed
3401 it so that, while verifying a sender address, $sender_address is set to <>.
3402 (There is no change to what happens when verifying a recipient address.)
3404 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3405 target A or AAAA records (if not already returned) without resetting the
3406 qualify_single or search_parents options of the DNS resolver. These are
3407 inappropriate in this case because the targets of MX and SRV records must
3408 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3409 target that, when qualified, matched something in the local domain. These
3410 two options are now turned off when doing these lookups.
3412 57. It seems that at least some releases of Reiserfs (which does not have the
3413 concept of a fixed number of inodes) returns zero and not -1 for the
3414 number of available inodes. This interacted badly with check_spool_inodes,
3415 which assumed that -1 was the "no such thing" setting. What I have done is
3416 to check that the total number of inodes is greater than zero before doing
3417 the test of how many are available.
3419 58. When a "warn" ACL statement has a log_message modifier, the message is
3420 remembered, and not repeated. This is to avoid a lot of repetition when a
3421 message has many recipients that cause the same warning to be written.
3422 Howewer, Exim was preserving the list of already written lines for an
3423 entire SMTP session, which doesn't seem right. The memory is now reset if a
3424 new message is started.
3426 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3427 debugging output unless log_rewrite was also set.
3429 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3430 of (int)(handle) when we know that handle contains (void *)(-1).
3432 61. The Exim daemon panic-logs an error return when it closes the incoming
3433 connection. However "connection reset by peer" seems to be common, and
3434 isn't really an error worthy of noting specially, so that particular error
3437 62. When Exim is trying to find all the local interfaces, it used to panic and
3438 die if the ioctl to get the interface flags failed. However, it seems that
3439 on at least one OS (Solaris 9) it is possible to have an interface that is
3440 included in the list of interfaces, but for which you get a failure error
3441 for this call. This happens when the interface is not "plumbed" into a
3442 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3443 failure of the "get flags" call assumes that the interface is down.
3445 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3446 makes life easier for people who are doing arithmetic on fields extracted
3447 from dates, where you often get leading zeros that should not be
3448 interpreted as octal.
3450 64. Added qualify_domain to the redirect router, to override the global
3453 65. If a pathologically long header line contained very many addresses (the
3454 report of this problem mentioned 10 000) and each of them was rewritten,
3455 Exim could use up a very large amount of memory. (It kept on making new
3456 copies of the header line as it rewrote, and never released the old ones.)
3457 At the expense of a bit more processing, the header rewriting function has
3458 been changed so that it no longer eats memory in this way.
3460 66. The generation of the Received: header has been moved from the time that a
3461 message starts to be received, to the time that it finishes. The timestamp
3462 in the Received: header should now be very close to that of the <= log
3463 line. There are two side-effects of this change:
3465 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3466 logged header lines no longer include the local Received: line, because
3467 it has not yet been created. The same applies to a copy of the message
3468 that is returned to a non-SMTP sender when a message is rejected.
3470 (b) When a filter file is tested using -bf, no additional Received: header
3471 is added to the test message. After some thought, I decided that this
3474 This change does not affect the value of $received_for. It is still set
3475 after address rewriting, but before local_scan() is called.
3477 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3479 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3480 gave an unhelpful panic error message, and a defer error. I have managed to
3481 change this behaviour so that it now rejects any supplied certificate,
3482 which seems right, as the list of acceptable certificates is empty.
3484 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3485 gave an unhelpful defer error. I have not managed to make this reject any
3486 supplied certificates, but the error message it gives is "no certificate
3487 supplied", which is not helpful.
3489 70. exigrep's output now also includes lines that are not associated with any
3490 message, but which match the given pattern. Implemented by a patch from
3491 Martin Sluka, which also tidied up the Perl a bit.
3493 71. Recipient callout verification, like sender verification, was using <> in
3494 the MAIL FROM command. This isn't really the right thing, since the actual
3495 sender may affect whether the remote host accepts the recipient or not. I
3496 have changed it to use the actual sender in the callout; this means that
3497 the cache record is now keyed on a recipient/sender pair, not just the
3498 recipient address. There doesn't seem to be a real danger of callout loops,
3499 since a callout by the remote host to check the sender would use <>.
3500 [SEE ABOVE: changed after hitting problems.]
3502 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3503 temporary errors. However, in the case of such a code being given after
3504 the end of a data transmission (i.e. after ".") Exim was failing to write
3505 a retry record for the message. (Yes, there was some broken host that was
3506 actually sending 8xx at this point.)
3508 73. An unknown lookup type in a host list could cause Exim to panic-die when
3509 the list was checked. (An example that provoked this was putting <; in the
3510 middle of a list instead of at the start.) If this happened during a DATA
3511 ACL check, a -D file could be left lying around. This kind of configuration
3512 error no longer causes Exim to die; instead it causes a defer errror. The
3513 incident is still logged to the main and panic logs.
3515 74. Buglet left over from Exim 3 conversion. The message "too many messages
3516 in one connection" was written to the rejectlog but not the mainlog, except
3517 when address rewriting (yes!) was being logged.
3519 75. Added write_rejectlog option.
3521 76. When a system filter was run not as root (that is, when system_filter_user
3522 was set), the values of the $n variables were not being returned to the
3523 main process; thus, they were not subsequently available in the $sn
3526 77. Added +return_path_on_delivery log selector.
3528 78. A connection timeout was being treated differently from recipients deferred
3529 when testing hosts_max_try with a message that was older than the host's
3530 retry timeout. (The host should not be counted, thus allowing all hosts to
3531 be tried at least once before bouncing.) This may have been the cause of an
3532 occasionally reported bug whereby a message would remain on the queue
3533 longer than the retry timeout, but would be bounced if a delivery was
3534 forced. I say "may" because I never totally pinned down the problem;
3535 setting up timeout/retry tests is difficult. See also the next item.
3537 79. The ultimate address timeout was not being applied to errors that involved
3538 a combination of host plus message (for example, a timeout on a MAIL
3539 command). When an address resolved to a number of possible hosts, and they
3540 were not all tried for each delivery (e.g. because of hosts_max_try), a
3541 message could remain on the queue longer than the retry timeout.
3543 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3546 81. Fixed an obscure SMTP outgoing bug which required at least the following
3547 conditions: (a) there was another message waiting for the same server;
3548 (b) the server returned 5xx to all RCPT commands in the first message so
3549 that the message was not completed; (c) the server dropped the connection
3550 or gave a negative response to the RSET that Exim sends to abort the
3551 transaction. The observed case was a dropped connection after DATA that had
3552 been sent in pipelining mode. That is, the server had advertised PIPELINING
3553 but was not implementing it correctly. The effect of the bug was incorrect
3554 behaviour, such as trying another host, and this could lead to a crash.
3560 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3561 and daemon.c were passed as pointers to ints; they should have been
3562 pointers to socklen_t variables (which are typically unsigned ints).
3564 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3567 3. Fixed a really odd bug that affected only the testing scheme; patching a
3568 certain fixed string in the binary changed the value of another string that
3569 happened to be identical to the end of the original first string.
3571 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3572 name", it returns that address as the IP address. On some operating
3573 systems (e.g. Solaris), it also passes back the IP address string as the
3574 "host name". However, on others (e.g. Linux), it passes back an empty
3575 string. Exim wasn't checking for this, and was changing the host name to an
3576 empty string, assuming it had been canonicized.
3578 5. Although rare, it is permitted to have more than one PTR record for a given
3579 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3580 all the names associated with an address, because they do in Solaris.
3581 However, it seems that they do not in Linux for data that comes from the
3582 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3583 I found this out when I moved to a new Linux workstation and tried to run
3584 the Exim test suite.
3586 To get round this problem I have changed the code so that it now does its
3587 own call to the DNS to look up PTR records when searching for a host name.
3588 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3589 addresses that are only in /etc/hosts are still found.
3591 This behaviour is, however, controlled by an option called host_lookup_
3592 order, which defaults to "bydns:byaddr". If people want to use the other
3593 order, or indeed, just use one or the other means of lookup, they can
3594 specify it in this variable.
3596 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3597 some operating systems, this comes back from gethostbyaddr() as an empty
3598 string, and this is what Exim used to test for. However, it seems that in
3599 other systems, "." is yielded. Exim now tests for this case too.
3601 7. The values of check_spool_space and check_log_space are now held internally
3602 as a number of kilobytes instead of an absolute number of bytes. If a
3603 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3604 kilobyte. This means that much larger values can be stored.
3606 8. Exim monitor: an attempt to get the action menu when not actually pointing
3607 at a message produces an empty menu entitled "No message selected". This
3608 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3609 no entries in it ("Shell widget menu has zero width and/or height"). So I
3610 have added a single, blank menu entry in this case.
3612 9. Added ${quote_local_part.
3614 10. MIME decoding is now applied to the contents of Subject: header lines when
3617 11. Now that a reference to $sender_host_address automatically causes a reverse
3618 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3619 host lookup before query-style lookups in lists that might use this
3620 variable. This has therefore been abolished, and the "net-" prefix is no
3621 longer necessary for query-style lookups.
3623 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3624 have been a typo for LFLAGS, so it has been changed.
3626 13. The install script calls Exim with "-C /dev/null" in order to find the
3627 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3628 to be output. Howeve, since Exim outputs its version number before the
3629 error, it didn't break the script. It just looked ugly. I fixed this by
3630 always allowing "-C /dev/null" if the caller is root.
3632 14. Ignore overlarge ACL variable number when reading spool file - insurance
3633 against a later release with more variables having written the file.
3635 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3636 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3637 this, as well as the form without the "IPv6" on the front.
3639 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3640 OS/Makefile-Darwin file.
3642 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3643 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3645 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3646 variable. However, some OS use ints for inodes. I've added cast to long int
3647 to get rid of the compiler warning.
3649 19. I had forgotten to lock out "/../" in configuration file names when
3650 ALT_CONFIG_PREFIX was set.
3652 20. Routers used for verification do not need to specify transports. However,
3653 if such a router generated a host list, and callout was configured, Exim
3654 crashed, because it could not find a port number from the (non-existent)
3655 transport. It now assumes port 25 in this circumstance.
3657 21. Added the -t option to exigrep.
3659 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3660 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3661 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3662 them is set, LOOKUP_LSEARCH is forced.
3664 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3666 24. Added sender and host information to the "rejected by local_scan()" log
3667 line; previously there was no indication of these.
3669 25. Added .include_if_exists.
3671 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3672 Linux. It turns out that not all file systems support this. Apparently some
3673 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3674 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3675 not supported on the file descriptor, is now ignored when Exim is trying to
3676 sync a directory. This applies only to Linux.
3678 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3680 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3681 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3684 29. Added support for maildirsize files from supplied patch (modified a bit).
3686 30. The use of :fail: followed by an empty string could lead Exim to respond to
3687 sender verification failures with (e.g.):
3689 550 Verification failed for <xxx>
3690 550 Sender verify failed
3692 where the first response line was missing the '-' that indicates it is not
3693 the final line of the response.
3695 31. The loop for finding the name of the user that called Exim had a hardwired
3696 limit of 10; it now uses the value of finduser_retries, which is used for
3697 all other user lookups.
3699 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3702 33. Exim was neglecting to zero errno before one call of strtol() when
3703 expanding a string and expecting an integer value. On some systems this
3704 resulted in spurious "integer overflow" errors. Also, it was casting the
3705 result into an int without checking.
3707 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3708 did not work. The code looks as if it has *never* worked, though it appears
3709 to have been documented since at least releast 1.62. I have made it work.
3711 35. The "timeout_DNS" error in retry rules, also documented since at least
3712 1.62, also never worked. As it isn't clear exactly what this means, and
3713 clearly it isn't a major issue, I have abolished the feature by treating it
3714 as "timeout", and writing a warning to the main and panic logs.
3716 36. The display of retry rules for -brt wasn't always showing the error code
3719 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3720 timeout_connect_A, timeout_connect_MX.
3722 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3723 to the empty sender.
3725 39. The daemon was not analysing the content of -oX till after it had closed
3726 stderr and disconnected from the controlling terminal. This meant that any
3727 syntax errors were only noted on the panic log, and the return code from
3728 the command was 0. By re-arranging the code a little, I've made the
3729 decoding happen first, so such errors now appear on stderr, and the return
3730 code is 1. However, the actual setting up of the sockets still happens in
3731 the disconnected process, so errors there are still only recorded on the
3734 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3735 connections (as happens on some IP stacks) was logged at start up time as
3736 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3737 it from "IPv6 and IPv4", which means that two separate sockets are being
3740 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3741 says whether AF_INET or AF_INET6 was passed as an argument.
3743 42. Exiwhat output was messed up when time zones were included in log
3746 43. Exiwhat now gives more information about the daemon's listening ports,
3747 and whether -tls-on-connect was used.
3749 44. The "port" option of the smtp transport is now expanded.
3751 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3752 silently ignored. Now an error message is written to the main and panic
3755 46. There's a new ACL modifier called "logwrite" which writes to a log file
3756 as soon as it is encountered.
3758 47. Added $local_user_uid and $local_user_gid at routing time.
3760 48. Exim crashed when trying to verify a sender address that was being
3763 49. Exim was recognizing only a space character after ".include". It now also
3764 recognizes a tab character.
3766 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3767 extracting the relevant information from the specification. The man page no
3768 longer contains scrambled data for the -d option, and I've added a section
3769 at the front about calling Exim under different names.
3771 51. Added "extra_headers" argument to the "mail" command in filter files.
3773 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3776 53. Installed eximstats 1.29.
3778 54. Added transport_filter_timeout as a generic transport option.
3780 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3781 Cc: header lines. This was required by RFC 822, but it not required by RFC
3784 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3785 incoming messages that did not have them. Now it does so only if the
3786 message originates locally, that is, if there is no associated remote host
3787 address. When Resent- header lines are present, this applies to the Resent-
3788 lines rather than the non-Resent- lines.
3790 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3791 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3793 58. Messages for configuration errors now include the name of the main
3794 configuration file - useful now that there may be more than one file in a
3795 list (.included file names were always shown).
3797 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3798 for those rare installations that do not start the daemon as root or run it
3799 setuid root. I've cut out the call to initgroups() if the daemon is not
3802 60. The Exim user and group can now be bound into the binary as text strings
3803 that are looked up at the start of Exim's processing.
3805 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3807 62. Added $mailstore_basename variable.
3809 63. Installed patch to sieve.c from Michael Haardt.
3811 64. When Exim failed to open the panic log after failing to open the main log,
3812 the original message it was trying to log was written to stderr and debug
3813 output, but if they were not available (the usual case in production), it
3814 was lost. Now it is written to syslog before the two lines that record the
3815 failures to open the logs.
3817 65. Users' Exim filters run in subprocesses under the user's uid. It is
3818 possible for a "deliver" command or an alias in a "personal" command to
3819 provoke an address rewrite. If logging of address rewriting is configured,
3820 this fails because the process is not running as root or exim. There may be
3821 a better way of dealing with this, but for the moment (because 4.30 needs
3822 to be released), I have disabled address rewrite logging when running a
3823 filter in a non-root, non-exim process.
3829 1. The buildconfig auxiliary program wasn't quoting the value set for
3830 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3831 not defined. This bug was masked in 4.22 by the effect that was fixed in
3834 2. Some messages that were rejected after a message id was allocated were
3835 shown as "incomplete" by exigrep. It no longer does this for messages that
3836 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3838 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3839 have allow_domain_literals set, the ID did not get logged in the <= line.
3840 Domain literals are now always recognized in Message-ID: header lines.
3842 4. The first argument for a ${extract expansion item is the key name or field
3843 number. Leading and trailing spaces in this item were not being ignored,
3844 causing some misleading effects.
3846 5. When deliver_drop_privilege was set, single queue runner processes started
3847 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3848 same command in the process it spins off) were not dropping privilege.
3850 6. When the daemon running as "exim" started a queue runner, it always
3851 re-executed Exim in the spun-off process. This is a waste of effort when
3852 deliver_drop_privilege is set. The new process now just calls the
3853 queue-runner function directly.
3859 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3862 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3863 ignored. Though the use of -oP was forcing the writing of a pid file, it
3864 was always written to the default place.
3866 3. If the message "no IP address found for host xxxx" is generated during
3867 incoming verification, it is now followed by identification of the incoming
3868 connection (so you can more easily find what provoked it).
3870 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3872 5. Added some features to "harden" Exim a bit more against certain attacks:
3874 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3875 be put in Local/Makefile. This is like the never_users runtime option,
3876 but it cannot be overridden. The default setting is "root".
3878 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3879 prefix string with which any file named in a -C command line option
3882 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3883 is retained for -C and -D only if the caller of Exim is root. Without
3884 it, the exim user may also use -C and -D and retain privilege.
3886 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3887 command line option is disabled.
3889 6. Macro names set by the -D option must start with an upper case letter, just
3890 like macro names defined in the configuration file.
3892 7. Added "dereference=" facility to LDAP.
3894 8. Two instances of the typo "uknown" in the source files are fixed.
3896 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3897 the Configure-Makefile script screwed up while processing it.
3899 10. Incorporated PCRE 4.4.
3901 11. The SMTP synchronization check was not operating right at the start of an
3902 SMTP session. For example, it could not catch a HELO sent before the client
3903 waited for the greeting. There is now a check for outstanding input at the
3904 point when the greeting is written. Because of the duplex, asynchronous
3905 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3906 way, but not yet received, when the check is performed.
3908 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3909 on TCP/IP sockets, because this apparently causes some broken clients to
3912 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3913 unchanged) from the Cygwin maintainer.
3915 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3916 was supported instead of testing for mbx. Effectively a typo.
3918 15. The spa authenticator server code was not checking that the input it
3919 received was valid base64.
3921 16. The debug output line for the "set" modifier in ACLs was not showing the
3922 name of the variable that was being set.
3924 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3926 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3927 lookup on its own. Something else was needed to trigger the lookup. For
3928 example, a match in host_lookup or the need for a host name in a host list.
3929 Now, if $sender_host_name is referenced and the host name has not yet been
3930 looked up, a lookup is performed. If the lookup fails, the variable remains
3931 empty, and $host_lookup_failed is set to "1".
3933 19. Added "eqi" as a case-independent comparison operator.
3935 20. The saslauthd authentication condition could segfault if neither service
3936 nor realm was specified.
3938 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3939 error message that was logged was misleading, and incoming SMTP
3940 connections were dropped. The message is now more accurate, and temporary
3941 errors are given to SMTP connections.
3943 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3944 (except RSET) with a 421 error, until QUIT is received. However, it was
3945 failing to send a response to QUIT.
3947 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3948 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3949 ACL is now run after the helo_try_verify_hosts code.
3951 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3952 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3953 case-independent, so other case variants are also recognized). Apparently
3954 some systems use these upper case variants.
3956 25. If more than two messages were waiting for the same host, and a transport
3957 filter was specified for the transport, Exim sent two messages over the
3958 same TCP/IP connection, and then failed with "socket operation on non-
3959 socket" when it tried to send the third.
3961 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3963 27. The extern definition of crypt16() in expand.c was not being excluded when
3964 the OS had its own crypt16() function.
3966 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3967 as a preferred synonym for return_size_limit, both as an option and as an
3970 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3972 30. Changed the default configuration ACL to relax the local part checking rule
3973 for addresses that are not in any local domains. For these addresses,
3974 slashes and pipe symbols are allowed within local parts, but the sequence
3975 /../ is explicitly forbidden.
3977 31. SPA server authentication was not clearing the challenge buffer before
3980 32. log_message in a "warn" ACL statement was writing to the reject log as
3981 well as to the main log, which contradicts the documentation and doesn't
3982 seem right (because no rejection is happening). So I have stopped it.
3984 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3985 However, I am unable to do any testing of this.
3987 34. Fixed an infelicity in the appendfile transport. When checking directories
3988 for a mailbox, to see if any needed to be created, it was accidentally
3989 using path names with one or more superfluous leading slashes; tracing
3990 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3992 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3993 contents are added to the log line that is written for every discarded
3994 recipient. (Previously a log_message setting was ignored.)
3996 36. The ${quote: operator now quotes the string if it is empty.
3998 37. The install script runs exim in order to find its version number. If for
3999 some reason other than non-existence or emptiness, which it checks, it
4000 could not run './exim', it was installing it with an empty version number,
4001 i.e. as "exim-". This error state is now caught, and the installation is
4004 38. An argument was missing from the function that creates an error message
4005 when Exim fails to connect to the socket for saslauthd authentication.
4006 This could cause Exim to crash, or give a corrupted message.
4008 39. Added isip, isip4, and isip6 to ${if conditions.
4010 40. The ACL variables $acl_xx are now saved with the message, and can be
4011 accessed later in routers, transports, and filters.
4013 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4014 strings in the file are not string-expanded.
4016 42. If a MAIL command specified a SIZE value that was too large to fit into an
4017 int variable, the check against message_size_limit failed. Such values are
4018 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4019 day this will have to be increased, but I don't think I want to be around
4020 when emails are that large.
4027 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4028 iconv() is not standard in FreeBSD.
4030 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4031 IPv6 enabled. The observed symptom was a segmentation fault on return from
4032 the function os_common_find_running_interfaces() in src/os.c.
4034 3. In the check_special_case() function in daemon.c I had used "errno" as an
4035 argument name, which causes warnings on some systems. This was basically a
4036 typo, since it was named "eno" in the comments!
4038 4. The code that waits for the clock to tick (at a resolution of some fraction
4039 of a second) so as to ensure message-id uniqueness was always waiting for
4040 at least one whole tick, when it could have waited for less. [This is
4041 almost certainly not relevant at current processor speeds, where it is
4042 unlikely to ever wait at all. But we try to future-proof.]
4044 5. The function that sleeps for a time interval that includes fractions of a
4045 second contained a race. It did not block SIGALRM between setting the
4046 timer, and suspending (a couple of lines later). If the interval was short
4047 and the sigsuspend() was delayed until after it had expired, the suspension
4048 never ended. On busy systems this could lead to processes getting stuck for
4051 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4052 process, before it forks any delivery processes. The open lookup caching
4053 mechanism meant that the open file or database connection was passed into
4054 the delivery process. The problem was that delivery processes always tidy
4055 up cached lookup data. This could cause a problem for the next delivery
4056 process started by the queue runner, because the external queue runner
4057 process does not know about the closure. So the next delivery process
4058 still has data in the lookup cache. In the case of a file lookup, there was
4059 no problem because closing a file descriptor in a subprocess doesn't affect
4060 the parent. However, if the lookup was caching a connection to a database,
4061 the connection was closed, and the second delivery process was likely to
4062 see errors such as "PGSQL: query failed: server closed the connection
4063 unexpectedly". The problem has been fixed by closing all cached lookups
4064 in a queue runner before running a delivery process.
4066 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4067 seem to have the "const" qualifier which it has on other OS. I've
4070 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4071 *of the same type* (client or server) with the same public name that an
4072 error should be diagnosed.
4074 9. When Exim looked up a host name for an IP address, but failed to find the
4075 original IP address when looking up the host name (a safety check), it
4076 output the message "<ip address> does not match any IP for NULL", which was
4077 confusing, to say the least. The bug was that the host name should have
4078 appeared instead of "NULL".
4080 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4081 user that is built into the binary, and the -C or -D options is used, root
4082 privilege is dropped before the configuration file is read. In addition,
4083 logging is switched to stderr instead of the normal log files. If the
4084 configuration then re-defines the Exim user, the unprivileged environment
4085 is probably not what is expected, so Exim logs a panic warning message (but
4088 However, if deliver_drop_privilege is set, the unprivileged state may well
4089 be exactly what is intended, so the warning has been cut out in that case,
4090 and Exim is allowed to try to write to its normal log files.
4096 1. smtp_return_error_details was not giving details for temporary sender
4097 or receiver verification errors.
4099 2. Diagnose a configuration error if two authenticators have the same public
4102 3. Exim used not to create the message log file for a message until the first
4103 delivery attempt. This could be confusing when incoming messages were held
4104 for policy or load reasons. The message log file is now created at the time
4105 the message is received, and an initial "Received" line is written to it.
4107 4. The automatically generated man page for command line options had a minor
4108 bug that caused no ill effects; however, a more serious problem was that
4109 the procedure for building the man page automatically didn't always
4110 operate. Consequently, release 4.20 contains an out-of-date version. This
4111 shouldn't happen again.
4113 5. When building Exim with embedded Perl support, the script that builds the
4114 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4115 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4117 6. The freeze_tell option was not being used for messages that were frozen on
4118 arrival, either by an ACL or by local_scan().
4120 7. Added the smtp_incomplete_transaction log selector.
4122 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4123 was accepting AUTH without a new EHLO.
4125 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4126 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4129 10. Logging of TCP/IP connections (when configured) now happens in the main
4130 daemon process instead of the child process, so that the TCP/IP connection
4131 count is more accurate (but it can never be perfect).
4133 11. The use of "drop" in a nested ACL was not being handled correctly in the
4134 outer ACL. Now, if condition failure induced by the nested "drop" causes
4135 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4136 or "require"), the connection is dropped.
4138 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4139 that yield "discard" can now be used with an "accept" or a "discard" verb,
4140 but an error is generated for any others (because I can't see a useful way
4141 to define what should happen).
4143 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4144 are now processed in the same way as lines in the Exim configuration file.
4145 In particular, continuation lines are supported.
4147 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4149 15. Added -ti meaning -t -i.
4151 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4152 domains, and warn by logging if others are found.
4154 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4155 returned from GIFCONF when Exim is trying to find the list of interfaces on
4156 a host. The code in os.c has been modified to copy each ifreq to an aligned
4157 structure in all cases.
4159 Also, in some cases, the returned ifreq's were being copied to a 'struct
4160 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4161 means the last couple of bytes of an IPv6 address could be chopped if the
4162 ifreq contained only a normal sockaddr (14 bytes storage).
4164 18. Named domain lists were not supported in the hosts_treat_as_local option.
4165 An entry such as +xxxx was not recognized, and was treated as a literal
4168 19. Ensure that header lines added by a DATA ACL are included in the reject log
4169 if the ACL subsequently rejects the message.
4171 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4172 MD5 (which is deprecated).
4174 21. When testing a filter file using -bf, Exim was writing a message when it
4175 took the sender from a "From " line in the message, but it was not doing so
4176 when it took $return_path from a Return-Path: header line. It now does.
4178 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4179 with a valid header line field name (a series of printing characters
4180 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4182 23. Changed "disc" in the source to "disk" to conform to the documentation and
4183 the book and for uniformity.
4185 24. Ignore Sendmail's -Ooption=value command line item.
4187 25. When execve() failed while trying to run a command in a pipe transport,
4188 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4189 could be confused with a return value of 69 from the command itself. This
4190 has been changed to 127, the value the shell returns if it is asked to run
4191 a non-existent command. The wording for the related log line suggests a
4192 non-existent command as the problem.
4194 26. If received_header_text expands to an empty string, do not add a Received:
4195 header line to the message. (Well, it adds a token one on the spool, but
4196 marks it "old" so that it doesn't get used or transmitted.)
4198 27. Installed eximstats 1.28 (addition of -nt option).
4200 28. There was no check for failure on the call to getsockname() in the daemon
4201 code. This can fail if there is a shortage of resources on the system, with
4202 ENOMEM, for example. A temporary error is now given on failure.
4204 29. Contrary to the C standard, it seems that in some environments, the
4205 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4206 program. Exim now does this explicitly; it affects the formatting of
4207 timestamps using strftime().
4209 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4210 complaints. I've now initialized all the variables, to avoid this.
4212 32. Header lines added by a system filter were not being "seen" during
4213 transport-time rewrites.
4215 33. The info_callback() function passed to OpenSSL is set up with type void
4216 (*)(SSL *, int, int), as described somewhere. However, when calling the
4217 function (actually a macro) that sets it up, the type void(*)() is
4218 expected. I've put in a cast to prevent warnings from picky compilers.
4220 34. If a DNS black list lookup found a CNAME record, but there were no A
4221 records associated with the domain it pointed at, Exim crashed.
4223 35. If a DNS black list lookup returned more than one A record, Exim ignored
4224 all but the first. It now scans all returned addresses if a particular IP
4225 value is being sought. In this situation, the contents of the
4226 $dnslist_value variable are a list of all the addresses, separated by a
4229 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4230 to accept a host name and all its aliases if the forward lookup for any of
4231 them yielded the IP address of the incoming connection. Now it accepts only
4232 those names whose forward lookup yields the correct IP address. Any other
4233 names are discarded. This closes a loophole whereby a rogue DNS
4234 administrator could create reverse DNS records to break through a
4235 wildcarded host restriction in an ACL.
4237 37. If a user filter or a system filter that ran in a subprocess used any of
4238 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4239 the wrong values were passed to the pipe command ($thisaddress had the
4240 value of $0, $0 had the value of $1, etc). This bug was introduced by
4241 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4243 38. Improved the line breaking for long SMTP error messages from ACLs.
4244 Previously, if there was no break point between 40 and 75 characters, Exim
4245 left the rest of the message alone. Two changes have been made: (a) I've
4246 reduced the minimum length to 35 characters; (b) if it can't find a break
4247 point between 35 and 75 characters, it looks ahead and uses the first one
4248 that it finds. This may give the occasional overlong line, but at least the
4249 remaining text gets split now.
4251 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4252 file descriptors might be low, and that setting 1000 would always raise it.
4253 It turns out that in some environments, the limit is already over 1000 and
4254 that lowering it causes trouble. So now Exim takes care not to decrease it.
4256 40. When delivering a message, the value of $return_path is set to $sender_
4257 address at the start of routing (routers may change the value). By an
4258 oversight, this default was not being set up when an address was tested by
4259 -bt or -bv, which affected the outcome if any router or filter referred to
4262 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4263 log only when "message" or "log_message" are set. However, if one of the
4264 conditions was an address verification, or a call to a nested ACL, the
4265 messages generated by the underlying test were being passed through. This
4266 no longer happens. The underlying message is available in $acl_verify_
4267 message for both "message" and "log_message" expansions, so it can be
4268 passed through if needed.
4270 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4271 new expansion $bh_ to give the encoded byte string without charset
4272 translation. Translation happens only if iconv() is available; HAVE_ICONV
4273 indicates this at build time. HEADERS_CHARSET gives the charset to
4274 translate to; headers_charset can change it in the configuration, and
4275 "headers charset" can change it in an individual filter file.
4277 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4278 that creates RFC 2047 encoded "words" labels them as that charset instead
4279 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4280 expansion operator; (ii) when Exim creates a From: line for a local
4281 message; (iii) when a header line is rewritten to include a "phrase" part.
4283 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4284 buggy, causing it to skip the first lines of messages whose message ID
4285 ended in 'D'. This would not have bitten before Exim release 4.14, because
4286 message IDs were unlikely to end in 'D' before then. The effect was to have
4287 incorrect size information for certain domains.
4289 45. #include "config.h" was missing at the start of the crypt16.c module. This
4290 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4293 46. If there was a timeout during a "random" callout check, Exim treated it as
4294 a failure of the random address, and carried on sending RSET and the real
4295 address. If the delay was just some slowness somewhere, the response to the
4296 original RCPT would be taken as a response to RSET and so on, causing
4297 mayhem of various kinds.
4299 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4300 when I implemented it. It didn't allow for the fact that some option values
4301 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4302 the right test for positive values.
4304 48. Domain names in DNS records are case-independent. Exim always looks them up
4305 in lower case. Some resolvers return domain names in exactly the case they
4306 appear in the zone file, that is, they may contain uppercase letters. Not
4307 all resolvers do this - some return always lower case. Exim was treating a
4308 change of case by a resolver as a change of domain, similar to a widening
4309 of a domain abbreviation. This triggered its re-routing code and so it was
4310 trying to route what was effectively the same domain again. This normally
4311 caused routing to fail (because the router wouldn't handle the domain
4312 twice). Now Exim checks for this case specially, and just changes the
4313 casing of the domain that it ultimately uses when it transmits the message
4316 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4319 50. If a filter generated a file delivery with a non-absolute name (possible if
4320 no home directory exists for the router), the forbid_file option was not
4323 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4324 the existing equality matching.
4326 52. Exim was using ints instead of ino_t variables in some places where it was
4327 dealing with inode numbers.
4329 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4330 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4331 called TMPDIR, and if it finds it is different, it changes its value.
4333 54. The smtp_printf() function is now made available to local_scan() so
4334 additional output lines can be written before returning. There is also an
4335 smtp_fflush() function to enable the detection of a dropped connection.
4336 The variables smtp_input and smtp_batched_input are exported to
4339 55. Changed the default runtime configuration: the message "Unknown user"
4340 has been removed from the ACL, and instead placed on the localuser router,
4341 using the cannot_route_message feature. This means that any verification
4342 failures that generate their own messages won't get overridden. Similarly,
4343 the "Unrouteable address" message that was in the ACL for unverifiable
4344 relay addresses has also been removed.
4346 56. Added hosts_avoid_esmtp to the smtp transport.
4348 57. The exicyclog script was not checking for the esoteric option
4349 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4350 will work only if exicyclog is run under the appropriate euid.
4352 58. Following a discussion on the list, the rules by which Exim recognises line
4353 endings on incoming messages have been changed. The -dropcr and drop_cr
4354 options are now no-ops, retained only for backwards compatibility. The
4355 following line terminators are recognized: LF CRLF CR. However, special
4356 processing applies to CR:
4358 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4359 nor a local message in the state where . is a terminator.
4361 (ii) If a bare CR is encountered in a header line, an extra space is added
4362 after the line terminator so as not to end the header. The reasoning
4363 behind this is that bare CRs in header lines are most likely either
4364 to be mistakes, or people trying to play silly games.
4366 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4367 was being incorrectly given as 18 bytes larger than it should have been.
4368 This is a VOB (very old bug).
4370 60. This may never have affected anything current, but just in case it has:
4371 When the local host is found other than at the start of a list of hosts,
4372 the local host, those with the same MX, and any that follow, are discarded.
4373 When the list in question was part of a longer list of hosts, the following
4374 hosts (not currently being processed) were also being discarded. This no
4375 longer happens. I'm not sure if this situation could ever has previously
4378 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4381 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4382 header. This is something that is recommended in a new Internet Draft, and
4383 is something that is documented as being done by Sendmail. There are two
4384 possible values. For messages generated by the autoreply transport, Exim
4387 Auto-Submitted: auto-replied
4389 whereas for all other generated messages (e.g. bounces) it adds
4391 Auto-Submitted: auto-generated
4393 63. The "personal" condition in filters now includes a test for the
4394 Auto-Submitted: header. If it contains the string "auto-" the message it
4395 not considered personal.
4397 64. Added rcpt_include_affixes as a generic transport option.
4399 65. Added queue_only_override (default true).
4401 66. Added the syslog_duplication option.
4403 67. If what should have been the first header line of a message consisted of
4404 a space followed by a colon, Exim was mis-interpreting it as a header line.
4405 It isn't of course - it is syntactically invalid and should therefore be
4406 treated as the start of the message body. The misbehaviour could have
4407 caused a number of strange effects, including loss of data in subsequent
4408 header lines, and spool format errors.
4410 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4411 client host had authenticated. This control can now be exercised by an ACL
4412 for more flexibility.
4414 69. By default, callouts do not happen when testing with -bh. There is now a
4415 variant, -bhc, which does actually run the callout code, including
4416 consulting and updating the callout cache.
4418 70. Added support for saslauthd authentication, courtesy of Alexander
4421 71. If statvfs() failed on the spool or log directories while checking their
4422 size for availability, Exim confusingly gave the error "space shortage".
4423 Furthermore, in debugging mode it crashed with a floating point exception.
4424 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4425 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4426 serious problem, Exim now writes to the main and panic logs when this
4427 happens, with details of the failure. It then refuses to accept the
4428 incoming message, giving the message "spool directory problem" or "log
4429 directory problem" with a 421 code for SMTP messages.
4431 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4432 0, 1, and 2 exist, because some OS complain for execs without them (see
4433 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4434 descriptors. However, the code omitted to check that the open succeeded,
4435 causing mysterious errors if for some reason the permissions on /dev/null
4436 got screwed. Now Exim writes a message to the main and panic logs, and
4437 bombs out if it can't open /dev/null.
4439 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4440 interact so that it is all more flexible. It is supposed to remain
4441 backwards compatible. Also added extra_local_interfaces.
4443 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4444 to bomb out with an assertion failure - to the client this appears as a
4445 connection drop. This problem occurs in the part of the code that was taken
4446 from the Samba project. Fortunately, the assertion is in a very simple
4447 function, so I have fixed this by reproducing the function inline in the
4448 one place where it is called, and arranging for authentication to fail
4449 instead of killing the process with assert().
4451 75. The SPA client code was not working when the server requested OEM rather
4452 than Unicode encoding.
4454 76. Added code to make require_files with a specific uid setting more usable in
4455 the case where statting the file as root fails - usually a non-root-mounted
4456 NFS file system. When this happens and the failure is EACCES, Exim now
4457 forks a subprocess and does the per-uid checking as the relevant uid.
4459 77. Added process_log_path.
4461 78. If log_file_path was not explicitly set, a setting of check_log_space or
4462 check_log_inodes was ignored.
4464 79. If a space check for the spool or log partitions fails, the incident is now
4465 logged. Of course, in the latter case the data may get lost...
4467 80. Added the %p formatting code to string_format() so that it can be used to
4468 print addresses in debug_print(). Adjusted all the address printing in the
4469 debugging in store.c to use %p rather than %d.
4471 81. There was a concern that a line of code in smtp_in.c could overflow a
4472 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4473 initially expressed, the concern was not well-founded, because trailing
4474 spaces are removed early. However, if the trailing spaces were followed by
4475 a NULL, they did not get removed, so the overflow was possible. Two fixes
4478 (a) I re-wrote the offending code in a cleaner fashion.
4479 (b) If an incoming SMTP command contains a NULL character, it is rejected
4482 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4483 runs initgroups(), so that if the Exim user is in any additional groups,
4484 they will be used during message reception.
4490 The change log for 4.20 and earlier releases has been archived.