1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.449 2006/12/24 12:12:05 magnus 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
13 MH/01 Don't check that the operands of numeric comparison operators are
14 integers when their expansion is in "skipping" mode (fixes bug
15 introduced by 4.64-PH/07).
21 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
22 leftover -K file (the existence of which was triggered by #402).
23 While we were at it, introduced process PID as part of the -K
24 filename. This should rule out race conditions when creating
27 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
28 processing considerably. Previous code took too long for large mails,
29 triggering a timeout which in turn triggers #401.
31 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
32 in the DK code in transports.c. sendfile() is not really portable,
33 hence the _LINUX specificness.
35 TF/01 In the add_headers option to the mail command in an Exim filter,
36 there was a bug that Exim would claim a syntax error in any
37 header after the first one which had an odd number of characters
40 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
41 callout verification, Exim cached a "reject" for the entire domain. This
42 is correct for most verifications, but it is not correct for a recipient
43 verification with use_sender or use_postmaster set, because in that case
44 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
45 case of MAIL FROM:<> rejection from other early rejections (e.g.
46 rejection of HELO). When verifying a recipient using a non-null MAIL
47 address, the cache is ignored if it shows MAIL FROM:<> rejection.
48 Whatever the result of the callout, the value of the domain cache is
49 left unchanged (for any other kind of callout, getting as far as trying
50 RCPT means that the domain itself is ok).
52 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
55 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
56 manifest itself as EPIPE rather than ECONNECT. When tidying away a
57 session, the daemon ignores ECONNECT errors and logs others; it now
58 ignores EPIPE as well.
60 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
61 (quoted-printable decoding).
63 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
64 later the small subsequent patch to fix an introduced bug.
66 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
68 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
70 PH/08 An error is now given if message_size_limit is specified negative.
72 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
73 to be given (somewhat) arbitrary names.
75 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
78 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
79 miscellaneous code fixes
81 PH/10 Added the log_reject_target ACL modifier to specify where to log
84 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
85 hostname. This is wrong, because it relates to the incoming message (and
86 probably the interface on which it is arriving) and not to the outgoing
87 callout (which could be using a different interface). This has been
88 changed to use the value of the helo_data option from the smtp transport
89 instead - this is what is used when a message is actually being sent. If
90 there is no remote transport (possible with a router that sets up host
91 addresses), $smtp_active_hostname is used.
93 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
94 tweaks were necessary in order to get it to work (see also 21 below):
95 (a) The code assumed that strncpy() returns a negative number on buffer
96 overflow, which isn't the case. Replaced with Exim's string_format()
98 (b) There were several signed/unsigned issues. I just did the minimum
99 hacking in of casts. There is scope for a larger refactoring.
100 (c) The code used strcasecmp() which is not a standard C function.
101 Replaced with Exim's strcmpic() function.
102 (d) The code set only $1; it now sets $auth1 as well.
103 (e) A simple test gave the error "authentication client didn't specify
104 service in request". It would seem that Dovecot has changed its
105 interface. Fortunately there's a specification; I followed it and
106 changed what the client sends and it appears to be working now.
108 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
111 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
112 address A is aliased to B and C, where B exists and C does not. Without
113 -v the output is "A verified" because verification stops after a
114 successful redirection if more than one address is generated. However,
115 with -v the child addresses are also verified. Exim was outputting "A
116 failed to verify" and then showing the successful verification for C,
117 with its parentage. It now outputs "B failed to verify", showing B's
118 parentage before showing the successful verification of C.
120 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
121 look up a TXT record in a specific list after matching in a combined
124 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
125 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
126 they consult the DNS. I had assumed they would set it the way they
127 wanted; and indeed my experiments on Linux seem to show that in some
128 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
129 To be on the safe side, however, I have now made the interface to
130 host_find_byname() similar to host_find_bydns(), with an argument
131 containing the DNS resolver options. The host_find_byname() function now
132 sets these options at its start, just as host_find_bydns() does. The smtp
133 transport options dns_qualify_single and dns_search_parents are passed to
134 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
135 of host_find_byname() use the default settings of RES_DEFNAMES
136 (qualify_single) but not RES_DNSRCH (search_parents).
138 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
139 spool_read_header() do less string testing, by means of a preliminary
140 switch on the second character of optional "-foo" lines. (This is
141 overdue, caused by the large number of possibilities that now exist.
142 Originally there were few.) While I was there, I also converted the
143 str(n)cmp tests so they don't re-test the leading "-" and the first
144 character, in the hope this might squeeze out yet more improvement.
146 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
147 flag allowing group syntax was set by the header_syntax check but not
148 turned off, possible causing trouble later; (2) The flag was not being
149 set at all for the header_verify test, causing "group"-style headers to
150 be rejected. I have now set it in this case, and also caused header_
151 verify to ignore an empty address taken from a group. While doing this, I
152 came across some other cases where the code for allowing group syntax
153 while scanning a header line wasn't quite right (mostly, not resetting
154 the flag correctly in the right place). These bugs could have caused
155 trouble for malformed header lines. I hope it is now all correct.
157 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
158 with the "reply" argument non-NULL. The code, however (which originally
159 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
160 but it didn't always do it. This confused somebody who was copying the
161 code for some other use. I have removed all the tests.
163 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
164 feature that was used to support insecure browsers during the U.S. crypto
165 embargo. It requires special client support, and Exim is probably the
166 only MTA that supported it -- and would never use it because real RSA is
167 always available. This code has been removed, because it had the bad
168 effect of slowing Exim down by computing (never used) parameters for the
169 RSA_EXPORT functionality.
171 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
172 authenticator to fail if there's a tab character in the incoming data
173 (there should never be unless someone is messing about, as it's supposed
174 to be base64-encoded). Also added, on Timo's advice, the "secured" option
175 if the connection is using TLS or if the remote IP is the same as the
176 local IP, and the "valid-client-cert option" if a client certificate has
179 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
180 authenticators. This can be used for authorization after authentication
181 succeeds. (In the case of plaintext, it servers for both authentication
184 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
185 if any retry times were supplied.
187 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
188 connection, where there is no client IP address to check. In this
189 situation, the verify now always succeeds.
191 PH/25 Applied John Jetmore's -Mset patch.
193 PH/26 Added -bem to be like -Mset, but loading a message from a file.
195 PH/27 In a string expansion for a processed (not raw) header when multiple
196 headers of the same name were present, leading whitespace was being
197 removed from all of them, but trailing whitespace was being removed only
198 from the last one. Now trailing whitespace is removed from each header
199 before concatenation. Completely empty headers in a concatenation (as
202 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
203 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
205 PH/29 [Removed. This was a change that I later backed out, and forgot to
206 correct the ChangeLog entry (that I had efficiently created) before
207 committing the later change.]
209 PH/30 Exim was sometimes attempting to deliver messages that had suffered
210 address errors (4xx response to RCPT) over the same connection as other
211 messages routed to the same hosts. Such deliveries are always "forced",
212 so retry times are not inspected. This resulted in far too many retries
213 for the affected addresses. The effect occurred only when there were more
214 hosts than the hosts_max_try setting in the smtp transport when it had
215 the 4xx errors. Those hosts that it had tried were not added to the list
216 of hosts for which the message was waiting, so if all were tried, there
217 was no problem. Two fixes have been applied:
219 (i) If there are any address or message errors in an SMTP delivery, none
220 of the hosts (tried or untried) are now added to the list of hosts
221 for which the message is waiting, so the message should not be a
222 candidate for sending over the same connection that was used for a
223 successful delivery of some other message. This seems entirely
224 reasonable: after all the message is NOT "waiting for some host".
225 This is so "obvious" that I'm not sure why it wasn't done
226 previously. Hope I haven't missed anything, but it can't do any
227 harm, as the worst effect is to miss an optimization.
229 (ii) If, despite (i), such a delivery is accidentally attempted, the
230 routing retry time is respected, so at least it doesn't keep
231 hammering the server.
233 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
234 in ${readsocket because some servers need this prod.
236 PH/32 Added some extra debug output when updating a wait-xxx database.
238 PH/33 The hint "could be header name not terminated by colon", which has been
239 given for certain expansion errors for a long time, was not being given
240 for the ${if def:h_colon_omitted{... case.
242 PH/34 The spec says: "With one important exception, whenever a domain list is
243 being scanned, $domain contains the subject domain." There was at least
244 one case where this was not true.
246 PH/35 The error "getsockname() failed: connection reset by peer" was being
247 written to the panic log as well as the main log, but it isn't really
248 panic-worthy as it just means the connection died rather early on. I have
249 removed the panic log writing for the ECONNRESET error when getsockname()
252 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
253 runs only) independently of the message's sender address. This meant
254 that, if the 4xx error was in fact related to the sender, a different
255 message to the same recipient with a different sender could confuse
256 things. In particualar, this can happen when sending to a greylisting
257 server, but other circumstances could also provoke similar problems.
258 I have changed the default so that the retry time for these errors is now
259 based a combination of the sender and recipient addresses. This change
260 can be overridden by setting address_retry_include_sender=false in the
263 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
264 remote server are returned as part of bounce messages. This was not
265 happening for LMTP over a pipe (the lmtp transport), but now it is the
266 same for both kinds of LMTP.
268 PH/38 Despite being documented as not happening, Exim was rewriting addresses
269 in header lines that were in fact CNAMEs. This is no longer the case.
271 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
272 and queue runs started by the daemon processed all messages. This has
273 been fixed so that -R and -S can now usefully be given with -q<time>.
275 PH/40 Import PCRE release 6.7 (fixes some bugs).
277 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
279 PH/42 Give an error if -q is specified more than once.
281 PH/43 Renamed the variables $interface_address and $interface_port as
282 $received_ip_address and $received_port, to make it clear that these
283 values apply to message reception, and not to the outgoing interface when
284 a message is delivered. (The old names remain recognized, of course.)
286 PH/44 There was no timeout on the connect() call when using a Unix domain
287 socket in the ${readsocket expansion. There now is.
289 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
290 be meaningful with "accept".
292 SC/01 Eximstats V1.43
293 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
295 SC/02 Eximstats V1.44
296 Use a glob alias rather than an array ref in the generated
297 parser. This improves both readability and performance.
299 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
300 Collect SpamAssassin and rejection statistics.
301 Don't display local sender or destination tables unless
302 there is data to show.
303 Added average volumes into the top table text output.
305 SC/04 Eximstats V1.46
306 Collect data on the number of addresses (recipients)
307 as well as the number of messages.
309 SC/05 Eximstats V1.47
310 Added 'Message too big' to the list of mail rejection
311 reasons (thanks to Marco Gaiarin).
313 SC/06 Eximstats V1.48
314 Mainlog lines which have GMT offsets and are too short to
315 have a flag are now skipped.
317 SC/07 Eximstats V1.49 (Alain Williams)
318 Added the -emptyok flag.
320 SC/08 Eximstats V1.50
321 Fixes for obtaining the IP address from reject messages.
323 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
324 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
325 whitesspace changes from 4.64-PH/27
327 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
330 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
331 are found, allow negative numbers in numeric criteria)
333 JJ/06 exipick.20061117.2, added new $message_body_missing variable
335 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
336 to match changes made in 4.64-PH/43
338 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
340 PH/47 Put in an explicit test for a DNS lookup of an address record where the
341 "domain" is actually an IP address, and force a failure. This locks out
342 those revolvers/nameservers that support "A-for-A" lookups, in
343 contravention of the specifications.
345 PH/48 When a host name was looked up from an IP address, and the subsequent
346 forward lookup of the name timed out, the host name was left in
347 $sender_host_name, contrary to the specification.
349 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
350 restricted to single-key lookups, Exim was not diagnosing an error if
351 * or *@ was used with a query-style lookup.
353 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
355 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
356 long ago, but noone interested enough thought of it. Let's just say that
357 the "1.1" means that there are some new functions that weren't there at
358 some point in the past.
360 PH/51 Error processing for expansion failure of helo_data from an smtp
361 transport during callout processing was broken.
363 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
364 tested/used via the -bh/-bhc/-bs options.
366 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
367 bug, fixed in subsequent PCRE releases).
369 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
370 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
372 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
378 SC/01 Use a glob alias rather than an array ref in eximstats generated
379 parser. This improves both readability and performance.
381 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
382 Don't display local sender or destination tables in eximstats unless
383 there is data to show.
384 Added average volumes into the eximstats top table text output.
386 SC/03 Collect data on the number of addresses (recipients) as well
387 as the number of messages in eximstats.
389 TF/01 Correct an error in the documentation for the redirect router. Exim
390 does (usually) call initgroups() when daemonizing.
392 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
393 with consistent privilege compared to when running as a daemon.
395 TF/03 Note in the spec that $authenticated_id is not set for local
396 submissions from trusted users.
398 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
399 Thanks to Dean Brooks <dean@iglou.com> for the patch.
401 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
402 by adding some example configuration directives to the default
403 configuration file. A little bit of work is required to uncomment the
404 directives and define how usernames and passwords are checked, but
405 there is now a framework to start from.
407 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
408 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
409 without this. I don't know how relevant this is to other LDAP libraries.
411 PH/02 Add the verb name to the "unknown ACL verb" error.
413 PH/03 Magnus Holmgren's patch for filter_prepend_home.
415 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
417 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
418 directory not expanded when it should be if an expanded home directory
419 was set for the address (which is overridden by the transport).
421 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
424 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
425 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
426 because it is too late at that time, and has no effect.
428 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
429 security issue with \' (bugzilla #107). I could not use the
430 PQescapeStringConn() function, because it needs a PGconn value as one of
433 PH/08 When testing addresses using -bt, indicate those final addresses that
434 are duplicates that would not cause an additional delivery. At least one
435 person was confused, thinking that -bt output corresponded to deliveries.
436 (Suppressing duplicates isn't a good idea as you lose the information
437 about possibly different redirections that led to the duplicates.)
439 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
440 systems where poll() doesn't work, in particular OS X.
442 PH/10 Added more information to debugging output for retry time not reached.
444 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
445 operations in malware.c.
447 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
450 PH/13 If write_rejectlog was set false when logging was sent to syslog with
451 syslog_duplication set false, log lines that would normally be written
452 both the the main log and to the reject log were not written to syslog at
455 PH/14 In the default configuration, change the use of "message" in ACL warn
456 statements to "add_header".
458 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
459 not followed by a command (e.g. "seen endif").
461 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
462 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
465 PH/17 Added extra conditions to the default value of delay_warning_condition
469 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
470 { match{$h_precedence:}{(?i)bulk|list|junk} } \
471 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
474 The Auto-Submitted: and various List- headers are standardised, whereas I
475 don't think Precedence: ever was.
477 PH/18 Refactored debugging code in route_finduser() to show more information,
478 in particular, the error code if getpwnam() issues one.
480 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
481 This is apparently needed in addition to the PH/07 change above to avoid
482 any possible encoding problems.
484 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
485 but not after initializing Perl.
487 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
488 output them only if debugging. By default they are written stderr,
489 apparently, which is not desirable.
491 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
494 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
497 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
499 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
500 authenticated or an ident call has been made. Suppress the default
501 values for $authenticated_id and $authenticated_sender (but permit -oMai
502 and -oMas) when testing with -bh.
504 PH/24 Re-jigged the order of the tests in the default configuration so that the
505 tests for valid domains and recipients precede the DNS black list and CSA
506 tests, on the grounds that those ones are more expensive.
508 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
509 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
510 This bug exists in every version of Exim that I still have, right back to
513 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
514 However, an attempt to turn on case-sensitivity in a regex key by
515 including (?-i) didn't work because the subject string was already
516 lowercased, and the effects were non-intuitive. It turns out that a
517 one-line patch can be used to allow (?-i) to work as expected.
523 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
524 other effects) broke the use of negated acl sub-conditions.
526 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
529 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
530 "Deny" causes Exim to reject the incoming connection with a 554 error.
531 Unfortunately, if there is a major crisis, such as a disk failure,
532 tcp-wrappers gives "deny", whereas what one would like would be some
533 kind of temporary error. A kludge has been added to help with this.
534 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
535 554 error is used if errno is still zero or contains ENOENT (which occurs
536 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
539 PH/03 Add -lutil to the default FreeBSD LIBS setting.
541 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
542 errors. Otherwise a message that provokes a temporary error (when other
543 messages do not) can cause a whole host to time out.
545 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
546 addresses were routed directly to files or pipes from a redirect router.
547 File deliveries just didn't batch; pipe deliveries might have suffered
550 PH/06 A failure to get a lock for a hints database would erroneously always say
551 "Failed to get write lock", even when it was really a read lock.
553 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
554 of 0600. This has been changed to use the value of the lockfile_mode
555 option (which defaults to 0600).
557 PH/08 Applied small patch from the Sieve maintainer.
559 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
560 folder from quota calculations, a direct delivery into this folder messed
561 up the contents of the maildirsize file. This was because the regex was
562 used only to exclude .Trash (or whatever) when the size of the mailbox
563 was calculated. There was no check that a delivery was happening into an
564 excluded directory. This bug has been fixed by ignoring all quota
565 processing for deliveries into excluded directories.
567 PH/10 Added the maildirfolder_create_regex option to appendfile.
573 PH/01 The code for finding all the local interface addresses on a FreeBSD
574 system running IPv6 was broken. This may well have applied to all BSD
575 systems, as well as to others that have similar system calls. The broken
576 code found IPv4 interfaces correctly, but gave incorrect values for the
577 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
578 that it would not match correctly against @[] and not recognize the IPv6
581 PH/02 The ipliteral router was not recognizing addresses of the form user@
582 [ipv6:....] because it didn't know about the "ipv6:" prefix.
584 PH/03 Added disable_ipv6.
586 PH/04 Changed $reply_address to use the raw form of the headers instead of the
587 decoded form, because it is most often used to construct To: headers
588 lines in autoreplies, and the decoded form may well be syntactically
589 invalid. However, $reply_address has leading white space removed, and all
590 newlines turned into spaces so that the autoreply transport does not
593 PH/05 If group was specified without a user on a router, and no group or user
594 was specified on a transport, the group from the router was ignored.
596 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
597 for visible compile-time settings that can be used to change these
598 numbers, for those that want even more. Backwards compatibility with old
599 spool files has been maintained. However, going back to a previous Exim
600 release will lost any variables that are in spool files.
602 PH/07 Two small changes when running in the test harness: increase delay when
603 passing a TCP/IP connection to a new process, in case the original
604 process has to generate a bounce, and remove special handling of
605 127.0.0.2 (sic), which is no longer necessary.
607 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
608 be the same on different OS.
610 PH/09 Moved a debug statement in filter processing to avoid a race problem when
613 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
614 whether --show-vars was specified or not
616 JJ/02 exipick: Added support for new ACL variable spool format introduced
619 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
620 syntactically invalid From: or Reply-to: line, and a filter used this to
621 generate an autoreply, and therefore failed to obtain an address for the
622 autoreply, Exim could try to deliver to a non-existent relative file
623 name, causing unrelated and misleading errors. What now happens is that
624 it logs this as a hard delivery error, but does not attempt to create a
627 PH/11 The exinext utility has a -C option for testing purposes, but although
628 the given file was scanned by exinext itself; it wasn't being passed on
629 when Exim was called.
631 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
632 an end-of-file indication when reading a command response.
634 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
635 compiled. In many other places in Exim, IPv6 addresses are always
636 recognized, so I have changed this. It also means that IPv4 domain
637 literals of the form [IPV4:n.n.n.n] are now always recognized.
639 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
640 used if the router is not running as root, for example, when verifying at
641 ACL time, or when using -bh. The debugging output from this situation was
642 non-existent - all you got was a failure to exec. I have made two
645 (a) Failures to set uid/gid, the current directory, or a process leader
646 in a subprocess such as that created by queryprogram now generate
647 suitable debugging ouput when -d is set.
649 (b) The queryprogram router detects when it is not running as root,
650 outputs suitable debugging information if -d is set, and then runs
651 the subprocess without attempting to change uid/gid.
653 PH/15 Minor change to Makefile for building test_host (undocumented testing
656 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
657 additional section of a DNS packet that returns MX or SRV records.
658 Instead, it always explicitly searches for A/AAAA records. This avoids
659 major problems that occur when a DNS server includes only records of one
660 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
661 fixed another bug: if SRV records were looked up and the corresponding
662 address records were *not* found in the additional section, the port
663 values from the SRV records were lost.
665 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
666 using the correct key (the original address) when searching the retry
667 rules in order to find which one to use for generating the retry hint.
669 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
670 adding the default one. Similarly, if it contains a Reply-To: header, the
671 errors_reply_to option, if set, is not used.
673 PH/19 When calculating a retry time, Exim used to measure the "time since
674 failure" by looking at the "first failed" field in the retry record. Now
675 it does not use this if it is later than than the arrival time of the
676 message. Instead it uses the arrival time. This makes for better
677 behaviour in cases where some deliveries succeed, thus re-setting the
678 "first failed" field. An example is a quota failure for a huge message
679 when small messages continue to be delivered. Without this change, the
680 "time since failure" will always be short, possible causing more frequent
681 delivery attempts for the huge message than are intended.
682 [Note: This change was subsequently modified - see PH/04 for 4.62.]
684 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
685 $1, $2, $3) because the numerical variables can be reset during some
686 expansion items (e.g. "match"), thereby losing the authentication data.
688 PH/21 Make -bV show the size of off_t variables so that the test suite can
689 decide whether to run tests for quotas > 2G.
691 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
692 mailbox_size, and mailbox_filecount in the appendfile transport. If a
693 filecount value is greater than 2G or if a quota value is greater than 2G
694 on a system where the size of off_t is not greater than 4, a panic error
697 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
698 never match. The debug and -bh output now contains an explicit error
699 message indicating a malformed IPv4 address or mask.
701 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
702 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
705 PH/25 Do not write to syslog when running in the test harness. The only
706 occasion when this arises is a failure to open the main or panic logs
707 (for which there is an explicit test).
709 PH/26 Added the /no_tell option to "control=freeze".
711 PH/27 If a host name lookup failed very early in a connection, for example, if
712 the IP address matched host_lookup and the reverse lookup yielded a name
713 that did not have a forward lookup, an error message of the form "no IP
714 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
715 could be logged. Now it outputs the IP address instead of "NULL".
717 PH/28 An enabling patch from MH: add new function child_open_exim2() which
718 allows the sender and the authenticated sender to be set when
719 submitting a message from within Exim. Since child_open_exim() is
720 documented for local_scan(), the new function should be too.
722 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
723 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
724 results in an empty string is now treated as unset.
726 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
728 PH/31 Added sender_verify_fail logging option.
730 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
731 needed by RFC 822 but not by RFC 2822 was commented out. I have now
732 tidied the source and removed it altogether.
734 PH/33 When a queue run was abandoned because the load average was too high, a
735 log line was always written; now it is written only if the queue_run log
736 selector is set. In addition, the log line for abandonment now contains
737 information about the queue run such as the pid. This is always present
738 in "start" and "stop" lines but was omitted from the "abandon" line.
740 PH/34 Omit spaces between a header name and the colon in the error message that
741 is given when verify = headers_syntax fails (if there are lots of them,
742 the message gets confusing).
744 PH/35 Change the default for dns_check_names_pattern to allow slashes within
745 names, as there are now some PTR records that contain slashes. This check
746 is only to protect against broken name servers that fall over on strange
747 characters, so the fact that it applies to all lookups doesn't matter.
749 PH/36 Now that the new test suite is complete, we can remove some of the
750 special code in Exim that was needed for the old test suite. For example,
751 sorting DNS records because real resolvers return them in an arbitrary
752 order. The new test suite's fake resolver always returns records in the
755 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
756 bounces) except when queue_only is set, to avoid logging races between
757 the different processes.
759 PH/38 Panic-die if .include specifies a non-absolute path.
761 PH/39 A tweak to the "H" retry rule from its user.
763 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
764 a label. They prevented compilation on older perls.
766 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
767 a warning to be raised on newish perls.
769 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
770 on queue. Changes to match documented behaviour of showing count of
771 messages matching specified criteria.
773 PH/40 Changed the default ident timeout from 30s to 5s.
775 PH/41 Added support for the use of login_cap features, on those BSD systems
776 that have them, for controlling the resources used by pipe deliveries.
778 PH/42 The content-scanning code uses fopen() to create files in which to put
779 message data. Previously it was not paying any attention to the mode of
780 the files. Exim runs with umask(0) because the rest of the code creates
781 files with open(), and sets the required mode explicitly. Thus, these
782 files were ending up world-writeable. This was not a big issue, because,
783 being within the spool directory, they were not world-accessible. I have
784 created a function called modefopen, which takes an additional mode
785 argument. It sets umask(777), creates the file, chmods it to the required
786 mode, then resets the umask. All the relevant calls to fopen() in the
787 content scanning code have been changed to use this function.
789 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
790 to 24 hours. This avoids potential overflow problems when processing G
791 and H retry rules. I suspect nobody ever tinkers with this value.
793 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
795 PH/45 When the plaintext authenticator is running as a client, the server's
796 challenges are checked to ensure they are valid base64 strings. By
797 default, the authentication attempt is cancelled if an invalid string is
798 received. Setting client_ignore_invalid_base64 true ignores these errors.
799 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
800 they are received. Thus, the responses can be made to depend on the
801 challenges. If an invalid string is ignored, an empty string is placed in
804 PH/46 Messages that are created by the autoreply transport now contains a
805 References: header, in accordance with RFCs 2822 and 3834.
807 PH/47 Added authenticated_sender_force to the smtp transport.
809 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
811 PH/49 Installed latest patch from the Sieve maintainer.
813 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
814 was also set, the appendfile transport was unnecessarily scanning a
815 directory of message files (e.g. for maildir delivery) to find the count
816 of files (along with the size), even though it did not need this
817 information. It now does the scan only if it needs to find either the
818 size of the count of files.
820 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
822 PH/52 Two bugs concerned with error handling when the smtp transport is
825 (i) Exim was not creating retry information for temporary errors given
826 for individual recipients after the DATA command when the smtp transport
827 was used in LMTP mode. This meant that they could be retried too
828 frequently, and not timed out correctly.
830 (ii) Exim was setting the flag that allows error details to be returned
831 for LMTP errors on RCPT commands, but not for LMTP errors for individual
832 recipients that were returned after the DATA command.
834 PH/53 This is related to PH/52, but is more general: for any failing address,
835 when detailed error information was permitted to be returned to the
836 sender, but the error was temporary, then after the final timeout, only
837 "retry timeout exceeded" was returned. Now it returns the full error as
838 well as "retry timeout exceeded".
840 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
841 do this, and (what is worse) MTAs that accept it.
843 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
844 will now be deprecated.
846 PH/56 New os.c-cygwin from the Cygwin maintainer.
848 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
849 formats (previously only available in exim formats via -bpr, -bpru,
850 and -bpra. Now also available in native and exiqgrep formats)
852 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
853 with very large, slow to parse queues
855 JJ/08 exipick: added ! as generic prefix to negate any criteria format
857 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
859 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
860 responses to authentication challenges, though it was showing the
861 challenges; (ii) I've removed the CR characters from the debug output for
864 PH/58 Allow for the insertion of a newline as well as a space when a string
865 is turned into more than one encoded-word during RFC 2047 encoding. The
866 Sieve code now uses this.
868 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
869 data_4xx, lost_connection, tls_required.
871 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
872 message was being sent as an SMTP response.
874 PH/61 Add -l and -k options to exicyclog.
876 PH/62 When verifying, if an address was redirected to one new address, so that
877 verification continued, and the new address failed or deferred after
878 having set something in $address_data, the value of $address_data was not
879 passed back to the ACL. This was different to the case when no
880 redirection occurred. The value is now passed back in both cases.
882 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
883 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
884 use login_cap.h, so on its own it isn't the distinguishing feature. The
885 new name refers directly to the setclassresources() function.
887 PH/65 Added configuration files for NetBSD3.
889 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
891 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
892 is preferred over IPv4.
894 PH/68 The bounce_return_message and bounce_return_body options were not being
895 honoured for bounces generated during the reception of non-SMTP messages.
896 In particular, this applied to messages rejected by the ACL. This bug has
897 been fixed. However, if bounce_return_message is true and bounce_return_
898 body is false, the headers that are returned for a non-SMTP message
899 include only those that have been read before the error was detected.
900 (In the case of an ACL rejection, they have all been read.)
902 PH/69 The HTML version of the specification is now built in a directory called
903 spec_html instead of spec.html, because the latter looks like a path with
904 a MIME-type, and this confuses some software.
906 PH/70 Catch two compiler warnings in sieve.c.
908 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
909 function verify_get_ident() calls ip_connect() to connect a socket, but
910 if the "connect()" function timed out, ip_connect() used to close the
911 socket. However, verify_get_ident() also closes the socket later, and in
912 between Exim writes to the log, which may get opened at this point. When
913 the socket was closed in ip_connect(), the log could get the same file
914 descriptor number as the socket. This naturally causes chaos. The fix is
915 not to close the socket in ip_connect(); the socket should be closed by
916 the function that creates it. There was only one place in the code where
917 this was missing, in the iplookup router, which I don't think anybody now
918 uses, but I've fixed it anyway.
920 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
921 well as to direct DNS lookups. Otherwise the handling of names in host
922 lists is inconsistent and therefore confusing.
928 PH/01 Two changes to the default runtime configuration:
930 (1) Move the checks for relay_from_hosts and authenticated clients from
931 after to before the (commented out) DNS black list checks.
933 (2) Add control=submission to the relay_from_hosts and authenticated
934 clients checks, on the grounds that messages accepted by these
935 statements are most likely to be submissions.
937 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
939 (1) Generate an error if the third argument for the ${prvs expansion is
942 (2) Treat a missing third argument of ${prvscheck as if it were an empty
945 (3) Reset the variables that are obtained from the first argument of
946 ${prvscheck and used in the second argument before leaving the code,
947 because their memory is reclaimed, so using them afterwards may do
950 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
951 one (it's much easier than Tom thought :-).
953 (5) Because of (4), we can now allow for the use of $prvscheck_result
954 inside the third argument.
956 PH/03 For some reason, the default setting of PATH when running a command from
957 a pipe transport was just "/usr/bin". I have changed it to
960 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
961 anything to be listed in the output from -bV.
963 PH/05 When a filter generated an autoreply, the entire To: header line was
964 quoted in the delivery log line, like this:
966 => >A.N.Other <ano@some.domain> <original@ddress> ...
968 This has been changed so that it extracts the operative address. There
969 may be more than one such address. If so, they are comma-separated, like
972 => >ano@some.domain,ona@other.domain <original@ddress> ...
974 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
975 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
976 not being looked up in the rDNS to get a host name, Exim was showing the
977 IP address twice in Received: lines, even though the IP addresses were
978 identical. For example:
980 Received: from [1.2.3.4] (helo=[1.2.3.4])
982 However, if the real host name was known, it was omitting the HELO data
983 if it matched the actual IP address. This has been tidied up so that it
984 doesn't show the same IP address twice.
986 PH/07 When both +timestamp and +memory debugging was on, the value given by
987 $tod_xxx expansions could be wrong, because the tod_stamp() function was
988 called by the debug printing, thereby overwriting the timestamp buffer.
989 Debugging no longer uses the tod_stamp() function when +timestamp is set.
991 PH/08 When the original message was included in an autoreply transport, it
992 always said "this is a copy of the message, including all the headers",
993 even if body_only or headers_only was set. It now gives an appropriate
996 PH/09 Applied a patch from the Sieve maintainer which:
998 o fixes some comments
999 o adds the (disabled) notify extension core
1000 o adds some debug output for the result of if/elsif tests
1001 o points to the current vacation draft in the documentation
1002 and documents the missing references header update
1006 o fixes a bug in processing the envelope test (when testing
1007 multiple envelope elements, the last element determinted the
1010 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1011 Electronic Mail") by including:
1013 Auto-submitted: auto-generated
1015 in the messages that it generates (bounce messages and others, such as
1016 warnings). In the case of bounce messages for non-SMTP mesages, there was
1017 also a typo: it was using "Auto_submitted" (underscore instead of
1018 hyphen). Since every message generated by Exim is necessarily in response
1019 to another message, thes have all been changed to:
1021 Auto-Submitted: auto-replied
1023 in accordance with these statements in the RFC:
1025 The auto-replied keyword:
1027 - SHOULD be used on messages sent in direct response to another
1028 message by an automatic process,
1030 - MUST NOT be used on manually-generated messages,
1032 - MAY be used on Delivery Status Notifications (DSNs) and Message
1033 Disposition Notifications (MDNs),
1035 - MUST NOT be used on messages generated by automatic or periodic
1036 processes, except for messages which are automatic responses to
1039 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1040 to the default Received: header definition.
1042 PH/12 Added log selector acl_warn_skipped (default on).
1044 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1045 variables because (a) they are in the wrong storage pool and (b) even if
1046 they were copied, it wouldn't work properly because of the caching.
1048 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1049 checking when decoding. Apparently there are clients that generate
1050 overlong encoded strings. Why am I not surprised?
1052 PH/15 If the first argument of "${if match_address" was not empty, but did not
1053 contain an "@" character, Exim crashed. Now it writes a panic log message
1054 and treats the condition as false.
1056 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1058 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1059 extension "envelope-auth". The code is finished and in agreement with
1060 other implementations, but there is no documentation so far and in fact,
1061 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1062 not changing the active code.
1064 Print executed "if" and "elsif" statements when debugging is used. This
1065 helps a great deal to understand what a filter does.
1067 Document more things not specified clearly in RFC3028. I had all this
1068 sorted out, when out of a sudden new issues came to my mind. Oops."
1070 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1073 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1074 canonical form (as documented). However, after a host name lookup from
1075 the IP address, check_host() was doing a simple string comparison with
1076 addresses acquired from the DNS when checking that the found name did
1077 have the original IP as one of its addresses. Since any found IPv6
1078 addresses are likely to be in abbreviated form, the comparison could
1079 fail. Luckily, there already exists a function for doing the comparison
1080 by converting both addresses to binary, so now that is used instead of
1081 the text comparison.
1083 PH/20 There was another similar case to PH/19, when a complete host name was
1084 given in a host list; looking up its IP address could give an abbreviated
1085 form, whereas the current host's name might or might not be abbreviated.
1086 The same fix has been applied.
1092 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1093 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1096 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1097 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1099 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1101 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1102 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1103 Also, for "Auto-Submitted", treat anything other than "no" as
1104 non-personal, in accordance with RFC 3834. (Previously it treated
1105 anything starting "auto-" as non-personal.)
1107 TF/01 The control=submission/name=... option had a problem with syntax
1108 errors if the name included a slash character. The /name= option
1109 now slurps the rest of the string, so it can include any characters
1110 but it must come last in the list of options (after /sender_retain
1113 PH/05 Some modifications to the interface to the fake nameserver for the new
1121 TK/01 Added the "success_on_redirect" address verification option. See
1122 NewStuff for rationale and an example.
1124 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1126 PH/02 Patch to exigrep to allow it to work on syslog lines.
1128 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1129 fread() to skip over the body file's header line, because in Cygwin the
1130 header line is locked and is inaccessible.
1132 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1133 co-exist for some time) to make it clear that it is the Exim ID that is
1134 referenced, not the Message-ID: header line.
1136 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1137 string_format() function, because snprintf() does not exist on all
1140 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1141 ${stat: expansion item.
1143 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1144 protocol synchronization error", to keep the pedants happy.
1146 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1147 well as for IRIX systems, when gcc is being used. See the host.c source
1150 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1152 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1155 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1156 transport and to the smtp transport in LMTP mode.
1158 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1160 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1161 run a filter in a subprocess. This could lead to confusion in subsequent
1162 lookups in the parent process. There should also be a search_tidyup() at
1163 the end of the subprocess.
1165 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1166 only if the host matched helo_try_verify_hosts, which caused the
1167 verification to occur when the EHLO/HELO command was issued. The ACL just
1168 tested the remembered result. Now, if a previous verification attempt has
1169 not happened, "verify = helo" does it there and then.
1171 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1173 TK/03 Fix log output including CR from clamd.
1175 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1176 exist provoked a memory error which could cause a segfault.
1178 PH/15 Installed PCRE 6.2
1180 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1182 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1183 of the problem. Specifically, suggested +O2 rather than +O1 for the
1186 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1188 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1189 record, the port was not correctly logged when the outgoing_port log
1190 selector was set (it logged the transort's default port).
1192 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1193 fallback_hosts, and "hosts" in the smtp transport.
1195 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1196 host errors such as "Connection refused".
1198 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1199 authentication with radiusclient 0.4.9:
1201 - Error returned from rc_read_config was caught wrongly
1202 - Username/password not passed on to radius server due to wrong length.
1204 The presumption is that some radiusclient API changes for 4.51/PH/17
1205 were not taken care of correctly. The code is still untested by me (my
1206 Linux distribution still has 0.3.2 of radiusclient), but it was
1207 contributed by a Radius user.
1209 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1210 expanding the "port" option of the smtp transport.
1212 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1213 while reading a MIME header. Thanks to Tom Hughes for a patch.
1215 PH/24 Include config.h inside local_scan.h so that configuration settings are
1218 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1219 that in an ACL for RCPT (for example), you can examine exactly what was
1222 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1223 commands, but it was not correctly comparing the address with the actual
1224 client host address. Thus, it would show the EHLO address in Received:
1225 header lines when this was not necessary.
1227 PH/27 Added the % operator to ${eval:}.
1229 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1230 it should be ignoring failures (because with -C, for example, it has lost
1231 privilege). It wasn't ignoring creation failures other than "already
1234 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1237 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1238 because an input line was too long, either on its own, or by virtue of
1239 too many continuations, the temporary file was not being removed, and the
1240 return code was incorrect.
1242 PH/31 Missing "BOOL" in function definition in filtertest.c.
1244 PH/32 Applied Sieve patches from the maintainer.
1246 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1248 PH/33 Added "verify = not_blind".
1250 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1251 Local/Makefile (with some defaults set). These are used in built scripts
1252 such as exicyclog, but they have never been used in the exim_install
1253 script (though there are many overriding facilities there). I have
1254 arranged that the exim_install script now takes note of these two
1257 PH/35 Installed configuration files for Dragonfly.
1259 PH/36 When a locally submitted message by a trusted user did not contain a
1260 From: header, and the sender address was obtained from -f or from an SMTP
1261 MAIL command, and the trusted user did not use -F to supply a sender
1262 name, $originator_name was incorrectly used when constructing a From:
1263 header. Furthermore, $originator_name was used for submission mode
1264 messages from external hosts without From: headers in a similar way,
1265 which is clearly wrong.
1267 PH/37 Added control=suppress_local_fixups.
1269 PH/38 When log_selector = +received_sender was set, and the addition of the
1270 sender made the log line's construction buffer exactly full, or one byte
1271 less than full, an overflow happened when the terminating "\n" was
1274 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1275 when the result of a list match is failure because a DNS lookup failed.
1277 PH/40 RM_COMMAND is now used in the building process.
1279 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1280 the "build-* directories that it finds.
1282 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1283 address in a domain literal was a prefix of an interface address.
1285 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1286 when verifying a sender address, unless rewrite_headers is false.
1288 PH/44 Wrote a long comment about why errors_to addresses are verified as
1289 recipients, not senders.
1291 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1292 the ratelimit ACL was added.
1294 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1296 PH/46 Added extra information about PostgreSQL errors to the error string.
1298 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1299 avoiding the need to install special zones in a real server. This is
1300 backwards compatible; if it can't find the fake resolver, it drops back.
1301 Thus, both old and new test suites can be run.
1303 TF/02 Added util/ratelimit.pl
1305 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1306 clock is set back in time.
1308 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1309 Candler <B.Candler@pobox.com>.
1311 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1312 OK for addresses that are the result of redirections.
1314 PH/48 A number of further additions for the benefit of the new test suite,
1315 including a fake gethostbyname() that interfaces to the fake DNS resolver
1318 TF/06 The fix for widen_domains has also been applied to qualify_single and
1319 search_parents which are the other dnslookup options that can cause
1322 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1325 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1327 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1328 service" if it encountered an MX record with an empty target hostname.
1329 The message is now "an MX or SRV record indicated no SMTP service".
1331 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1332 if the DNS of the sending site is misconfigured. This is quite a
1333 common situation. This change restores the behaviour of treating a
1334 helo verification defer as a failure.
1336 PH/51 If self=fail was set on a router, the bounce message did not include the
1337 actual error message.
1343 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1345 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1346 command itself ended in error, the underlying message about the transport
1347 filter timeout was being overwritten with the pipe command error. Now the
1348 underlying error message should be appended to the second error message.
1350 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1352 PH/02 Reduce the amount of output that "make" produces by default. Full output
1353 can still be requested.
1355 PH/03 The warning log line about a condition test deferring for a "warn" verb
1356 was being output only once per connection, rather than after each
1357 occurrence (because it was using the same function as for successful
1358 "warn" verbs). This seems wrong, so I have changed it.
1360 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1361 it should not have, which might have caused a crash in the right
1362 circumstances, but probably never did.
1364 PH/04 Installed a modified version of Tony Finch's patch to make submission
1365 mode fix the return path as well as the Sender: header line, and to
1366 add a /name= option so that you can make the user's friendly name appear
1369 TF/03 Added the control = fakedefer ACL modifier.
1371 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1372 Mark Lowes for thorough testing.
1374 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1376 TK/03 Merged latest SRS patch from Miles Wilton.
1378 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1379 with the definition in sysexits.h (which is #included earlier).
1380 Fortunately, Exim does not actually use EX_OK. The code used to try to
1381 preserve the sysexits.h value, by assumimg that macro definitions were
1382 scanned for macro replacements. I have been disabused of this notion,
1383 so now the code just undefines EX_OK before #including unistd.h.
1385 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1386 in the smtp transport. When a block could not be written in a single
1387 write() function, the timeout was being re-applied to each part-write.
1388 This seems wrong - if the receiver was accepting one byte at a time it
1389 would take for ever. The timeout is now adjusted when this happens. It
1390 doesn't have to be particularly precise.
1392 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1393 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1395 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1396 without a domain if the check to <postmaster@domain> fails.
1398 SC/01 Eximstats: added -xls and the ability to specify output files
1399 (patch written by Frank Heydlauf).
1401 SC/02 Eximstats: use FileHandles for outputing results.
1403 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1405 SC/04 Eximstats: fixed display of large numbers with -nvr option
1407 SC/05 Eximstats: fixed merging of reports with empty tables.
1409 SC/06 Eximstats: added the -include_original_destination flag
1411 SC/07 Eximstats: removed tabs and trailing whitespace.
1413 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1415 TK/06 MBOX spool code: Add real "From " MBOX separator line
1416 so the .eml file is really in mbox format (even though
1417 most programs do not really care). Patch from Alex Miller.
1419 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1420 The latter is generated from $received_to and is only set if the
1421 message has one envelope recipient. SA can use these headers,
1422 obviously out-of-the-box. Patch from Alex Miller.
1424 PH/08 The ${def test on a variable was returning false if the variable's
1425 value was "0", contrary to what the specification has always said!
1426 The result should be true unless the variable is empty.
1428 PH/09 The syntax error of a character other than { following "${if
1429 def:variable_name" (after optional whitespace) was not being diagnosed.
1430 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1431 accidental colon was present, for example, could give incorrect results.
1433 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1434 result is used (not including appendfile, where other changes are about
1437 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1438 This involved changing a lot of size variables from int to off_t. It
1439 should work with maildirs and everything.
1441 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1442 spamd dying while we are connected to it.
1444 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1447 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1448 function for BDB 4.3.
1450 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1452 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1453 the sources for PCRE have been re-arranged and more files are now
1456 PH/15 The code I had for printing potentially long long variables in PH/11
1457 above was not the best (it lost precision). The length of off_t variables
1458 is now inspected at build time, and an appropriate printing format (%ld
1459 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1460 to be "long long int" or "long int". This is needed for the internal
1461 formatting function string_vformat().
1463 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1464 the configuration file to be ":syslog", then the script "guesses" where
1465 the logs files are, rather than using the compiled in default. In our
1466 case the guess is not the same as the compiled default, so the script
1467 suddenly stopped working when I started to use syslog. The patch checks
1468 to see if log_file_path is "". If so, it attempts to read it from exim
1469 with no configuration file to get the compiled in version, before it
1470 falls back to the previous guessing code."
1472 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1473 implementing BATV in an Exim configuration. See NewStuff for the gory
1476 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1477 Makefile that are specific to HP-UX.
1479 PH/18 If the "use_postmaster" option was set for a recipient callout together
1480 with the "random" option, the postmaster address was used as the MAIL
1481 FROM address for the random test, but not for the subsequent recipient
1482 test. It is now used for both.
1484 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1485 patch removes a few documentation additions to RFC 3028, because the
1486 latest draft now contains them. It adds the new en;ascii-case comparator
1487 and a new error check for 8bit text in MIME parts. Comparator and
1488 require names are now matched exactly. I enabled the subaddress
1489 extension, but it is not well tested yet (read: it works for me)."
1491 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1492 rework some of the code of TK/09 above to avoid the hardwired use of
1493 "%lld" and "long long". Replaced the call to snprintf() with a call to
1496 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1497 records point to non-existent hosts", "retry timeout exceeded", and
1498 "retry time not reached for any host after a long failure period".
1500 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1501 experimental DomainKeys support:
1503 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1504 (2) On an error such as an illegally used "control", the wrong name for
1505 the control was given.
1507 These problems did NOT occur unless DomainKeys support was compiled.
1509 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1511 PH/24 Added ${if match_ip condition.
1513 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1514 databases so that it will be absolutely obvious if a crash occurs in the
1515 DB library. This is a regular occurrence (often caused by mis-matched
1518 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1519 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1520 picked up on a user's system that detects such things. There doesn't seem
1521 to be a gcc warning option for this - only an attribute that has to be
1522 put on the function's prototype. It seems that in Fedora Core 4 they have
1523 set this on a number of new functions. No doubt there will be more in due
1526 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1527 specify a transport. However, if an address that was verified by such a
1528 router was the subject of a callout, Exim crashed because it tried to
1529 read the rcpt_include_affixes from the non-existent transport. Now it
1530 just assumes that the setting of that option is false. This bug was
1531 introduced by 4.51/PH/31.
1533 PH/28 Changed -d+all to exclude +memory, because that information is very
1534 rarely of interest, but it makes the output a lot bigger. People tend to
1535 do -d+all out of habit.
1537 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1538 code in os-type was giving problems when libc.so lives in lib64, like on
1541 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1542 aren't the modern standard, and it seems that some systems' include files
1543 don't always have them. Exim was already checking for some of the newer
1544 ones like T_AAAA, and defining it itself. I've added checks for all the
1545 record types that Exim uses.
1547 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1548 not automatically generating a new one, as it is supposed to. This
1549 prevented TLS from working. If the file did exist, but contained invalid
1550 data, a new version was generated, as expected. It was only the case of a
1551 non-existent file that was broken.
1553 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1554 with a change in libdomainkeys > 0.64.
1556 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1557 from DNS. If the selector record carries the flag, it now has
1558 precedence over the domain-wide flag.
1560 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1562 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1563 the use of a port name that isn't defined in /etc/services) occurred, the
1564 message was deferred as in a normal delivery, and thus remained on the
1565 spool, instead of being failed because of the mua_wrapper setting. This
1566 is now fixed, and I tidied up some of the mua_wrapper messages at the
1569 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1570 the messages in a hash of arrays rather than using individual hashes.
1571 This is a bit cleaner and results in dramatic memory savings, albeit
1572 at a slight CPU cost.
1574 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1575 as requested by Marc Sherman.
1577 SC/10 Eximstats: added histograms for user specified patterns as requested
1580 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1582 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1583 fopen() in the content-scanning modules that did not already have it.
1589 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1590 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1592 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1594 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1596 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1597 to test Sieve filters that use "vacation".
1599 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1600 that changes the way the GnuTLS parameters are stored in the cache file.
1601 The new format can be generated externally. For backward compatibility,
1602 if the data in the cache doesn't make sense, Exim assumes it has read an
1603 old-format file, and it generates new data and writes a new file. This
1604 means that you can't go back to an older release without removing the
1607 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1608 work if there are any delivery delays because "one_time" forces the
1609 parent to be marked "delivered", so its unseen clone is never tried
1610 again. For this reason, Exim now forbids the simultaneous setting of
1613 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1614 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1615 entry if you want to know the details. The fix, however, neglected to
1616 consider the case when local delivery batching is involved. The test for
1617 "previously delivered" was not happening when checking to see if an
1618 address could be batched with a previous (undelivered) one; under
1619 certain circumstances this could lead to multiple deliveries to the same
1622 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1623 in its include files, and this causes problems building Exim.
1625 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1626 header_syntax) but Exim was just ignoring anything given after a slash.
1627 In particular, this caused confusion with an attempt to use "verify =
1628 reverse_host_lookup/defer_ok". An error is now given when options are
1629 supplied for verify items that do not have them. (Maybe reverse_host_
1630 lookup should have a defer_ok option, but that's a different point.)
1632 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1633 defined by RFC 821) to 2048, because there were problems with some AUTH
1634 commands, and RFC 1869 says the size should be increased for extended
1635 SMTP commands that take arguments.
1637 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1640 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1641 "unknown" error; now it says that the functionality isn't in the binary.
1643 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1644 an address' error message when a string expansion fails (syntax or
1645 whatever). Otherwise the password may appear in the log. Following change
1646 PH/42 below, there is no longer a chance of it appearing in a bounce
1649 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1651 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1652 of its addresses was ever tried. (Bugzilla bug #2.)
1654 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1655 the result incorrectly in the debug output. (It correctly added a newline
1656 to what was transported.)
1658 TF/01 Added $received_time.
1660 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1661 commented out examples of how to interface to a virus scanner and to
1662 SpamAssassin. Also added commented examples of av_scanner and
1663 spamd_address settings.
1665 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1666 and controls are allowed in which ACLs. There were a couple of minor
1667 errors. Some of the entries in the conditions table (which is a table of
1668 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1669 negation of where the condition IS allowed.
1671 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1673 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1674 header file does not have a version number, so I've had to invent a new
1675 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1676 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1677 radiusclient), but it was contributed by a Radius user.
1679 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1680 files or directories, for OpenSSL.
1682 PH/19 When an Exim process that is running as root has to create an Exim log
1683 file, it does so in a subprocess that runs as exim:exim so as to get the
1684 ownership right at creation (otherwise, other Exim processes might see
1685 the file with the wrong ownership). There was no test for failure of this
1686 fork() call, which would lead to the process getting stuck as it waited
1687 for a non-existent subprocess. Forks do occasionally fail when resources
1688 run out. I reviewed all the other calls to fork(); they all seem to check
1691 PH/20 When checking for unexpected SMTP input at connect time (before writing
1692 the banner), Exim was not dealing correctly with a non-positive return
1693 from the read() function. If the client had disconnected by this time,
1694 the result was a log entry for a synchronization error with an empty
1695 string after "input=" when read() returned zero. If read() returned -1
1696 (an event I could not check), uninitialized data bytes were printed.
1697 There were reports of junk text (parts of files, etc) appearing after
1700 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1702 PH/22 Added support for macro redefinition, and (re)definition in between
1703 driver and ACL definitions.
1705 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1706 forgetting to use the resulting value; it was using the unexpanded value.
1708 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1709 hadn't been configured. The fix is from Juergen Kreileder, who
1710 understands it better than I do:
1712 "Here's what I see happening with three configured cyrus_sasl
1713 authenticators configured (plain, login, cram-md5):
1715 On startup auth_cyrus_sasl_init() gets called for each of these.
1716 This means three calls to sasl_listmech() without a specified mech_list.
1717 => SASL tests which mechs of all available mechs actually work
1718 => three warnings about OTP not working
1719 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1721 With the patch, sasl_listmech() also gets called three times. But now
1722 SASL's mech_list option is set to the server_mech specified in the the
1723 authenticator. Or in other words, the answer from sasl_listmech()
1724 gets limited to just the mech you're testing for (which is different
1726 => the return list contains just 'plain' or 'login', 'cram-md5' or
1727 nothing depending on the value of ob->server_mech.
1729 I've just tested the patch: Authentication still works fine,
1730 unavailable mechs specified in the exim configuration are still
1731 caught, and the auth.log warnings about OTP are gone."
1733 PH/25 When debugging is enabled, the contents of the command line are added
1734 to the debugging output, even when log_selector=+arguments is not
1737 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1738 answer is "GNU", and only if the return is "GNU/something" is the answer
1741 PH/27 $acl_verify_message is now set immediately after the failure of a
1742 verification in an ACL, and so is available in subsequent modifiers. In
1743 particular, the message can be preserved by coding like this:
1745 warn !verify = sender
1746 set acl_m0 = $acl_verify_message
1748 Previously, $acl_verify_message was set only while expanding "message"
1749 and "log_message" when a very denied access.
1751 PH/28 Modified OS/os.c-Linux with
1753 -#ifndef OS_LOAD_AVERAGE
1754 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1756 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1757 nomenclature these days.)
1759 PH/29 Installed patch from the Sieve maintainer that adds the options
1760 sieve_useraddress and sieve_subaddress to the redirect router.
1762 PH/30 In these circumstances:
1763 . Two addresses routed to the same list of hosts;
1764 . First host does not offer TLS;
1765 . First host accepts first address;
1766 . First host gives temporary error to second address;
1767 . Second host offers TLS and a TLS session is established;
1768 . Second host accepts second address.
1769 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1770 and peerdn, if requested) that were in fact used only for the second
1773 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1774 attention to any local part prefix or suffix that was matched by the
1775 router that accepted the address. It now behaves in the same way as it
1776 does for delivery: the affixes are removed from the local part unless
1777 rcpt_include_affixes is set on the transport.
1779 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1780 timeout during the DATA phase of an incoming message.
1782 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1783 applied a patch sanctioned by the Sieve maintainer.
1785 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1786 the uid or gid is negative. A case of a negative gid caused this to be
1787 noticed. The fix allows for either to be negative.
1789 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1790 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1793 PH/36 The change PH/12 above was broken. Fixed it.
1795 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1796 the grounds that routing the same address twice would always produce the
1797 same answer. This might have been true once, but it is certainly no
1798 longer true now. Routing a child address may depend on the previous
1799 routing that produced that child. Some complicated redirection strategies
1800 went wrong when messages had multiple recipients, and made Exim's
1801 behaviour dependent on the order in which the addresses were given.
1803 I have moved the duplicate checking until after the routing is complete.
1804 Exim scans the addresses that are assigned to local and remote
1805 transports, and removes any duplicates. This means that more work will be
1806 done, as duplicates will always all be routed, but duplicates are
1807 presumably rare, so I don't expect this is of any significance.
1809 For deliveries to pipes, files, and autoreplies, the duplicate checking
1810 still happens during the routing process, since they are not going to be
1813 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1814 It corrects a timeout issue with spamd. This is Ian's comment: "The
1815 background is that sometimes spamd either never reads data from a
1816 connection it has accepted, or it never writes response data. The exiscan
1817 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1818 blindly assumes that writes won't block so it may never time out."
1820 PH/39 Allow G after quota size as well as K and M.
1822 PH/40 The value set for $authenticated_id in an authenticator may not contain
1823 binary zeroes or newlines because the value is written to log lines and
1824 to spool files. There was no check on this. Now the value is run through
1825 the string_printing() function so that such characters are converted to
1826 printable escape sequences.
1828 PH/41 $message_linecount is a new variable that contains the total number of
1829 lines in the message. Compare $body_linecount, which is the count for the
1832 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1833 bounce and delay warning messages, except in certain special cases, which
1836 (a) An SMTP error message from a remote host;
1837 (b) A message specified in a :fail: redirection;
1838 (c) A message specified in a "fail" command in a system filter;
1839 (d) A message specified in a FAIL return from the queryprogram router;
1840 (e) A message specified by the cannot_route_message router option.
1842 In these cases only, Exim does include the error details in bounce and
1843 warning messages. There are also a few cases where bland messages such
1844 as "unrouteable address" or "local delivery error" are given.
1846 PH/43 $value is now also set for the "else" part of a ${run expansion.
1848 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1849 being worked on, but at least Exim now implements the latest version to
1852 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1853 process to complete was treated as a delivery failure, a timeout while
1854 writing the message to the pipe was logged, but erroneously treated as a
1855 successful delivery. Such timeouts include transport filter timeouts. For
1856 consistency with the overall process timeout, these timeouts are now
1857 treated as errors, giving rise to delivery failures by default. However,
1858 there is now a new Boolean option for the pipe transport called
1859 timeout_defer, which, if set TRUE, converts the failures into defers for
1860 both kinds of timeout. A transport filter timeout is now identified in
1863 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1864 systems where "make" and "gmake" are different, calling "gmake" at top
1865 level broke things. I've arranged for the value of $(MAKE) to be passed
1866 from the Makefile to this script so that it can call the same version of
1870 A note about Exim versions 4.44 and 4.50
1871 ----------------------------------------
1873 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1874 changes of various kinds. As a consequence, a big documentation update was
1875 needed. This delayed the release for rather longer than seemed good, especially
1876 in the light of a couple of (minor) security issues. Therefore, the changes
1877 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1878 release. So 4.44 and 4.50 are in effect two different branches that both start
1881 I have left the 4.50 change log unchanged; it contains all the changes since
1882 4.43. The change log for 4.44 is below; many of its items are identical to
1883 those for 4.50. This seems to be the most sensible way to preserve the
1884 historical information.
1890 1. Minor wording change to the doc/README.SIEVE file.
1892 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1893 computation of the current number of files was incorrect.
1895 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1896 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1899 4. Give more explanation in the error message when the command for a transport
1900 filter fails to execute.
1902 5. There are several places where Exim runs a non-Exim command in a
1903 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1904 was being done only for the command run by the queryprogram router. It is
1905 now done for all such subprocesses. The other cases are: ${run, transport
1906 filters, and the commands run by the lmtp and pipe transports.
1908 6. Added CONFIGURE_GROUP build-time option.
1910 7. Some older OS have a limit of 256 on the maximum number of file
1911 descriptors. Exim was using setrlimit() to set 1000 as a large value
1912 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1913 systems. I've change it so that if it can't get 1000, it tries for 256.
1915 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1916 was an oversight, and furthermore, ever since the addition of extra
1917 controls (e.g. 4.43/32), the checks on when to allow different forms of
1918 "control" were broken. There should now be diagnostics for all cases when a
1919 control that does not make sense is encountered.
1921 9. Added the /retain_sender option to "control=submission".
1923 10. $recipients is now available in the predata ACL (oversight).
1925 11. Tidy the search cache before the fork to do a delivery from a message
1926 received from the command line. Otherwise the child will trigger a lookup
1927 failure and thereby defer the delivery if it tries to use (for example) a
1928 cached ldap connection that the parent has called unbind on.
1930 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1931 of $address_data from the recipient verification was clobbered by the
1932 sender verification.
1934 13. The value of address_data from a sender verification is now available in
1935 $sender_address_data in subsequent conditions in the ACL statement.
1937 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1939 15. Added a new option "connect=<time>" to callout options, to set a different
1942 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1943 was its contents. (It was OK if the option was not defined at all.)
1945 17. A "Completed" log line is now written for messages that are removed from
1946 the spool by the -Mrm option.
1948 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1949 information about exactly what failed.
1951 19. Added -dd to debug only the daemon process.
1953 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1954 handles timeouts, both on the server side and network timeouts. Renamed the
1955 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1957 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1958 It is now set to "smtps".
1960 22. $host_address is now set to the target address during the checking of
1961 ignore_target_hosts.
1963 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1964 being passed; this would have caused $sender_host_name to have been used if
1965 matching the list had actually called for a host name (not very likely,
1966 since this list is usually IP addresses). A host name is now passed as
1969 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1970 code that specifies a non-restarting handler (typically sigaction() in
1971 modern systems) in an attempt to fix a rare and obscure crash bug.
1973 25. Narrowed the window for a race in the daemon that could cause it to ignore
1974 SIGCHLD signals. This is not a major problem, because they are used only to
1975 wake it up if nothing else does.
1977 26. A malformed maildirsize file could cause Exim to calculate negative values
1978 for the mailbox size or file count. Odd effects could occur as a result.
1979 The maildirsize information is now recalculated if the size or filecount
1982 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1983 support for a long time. Removed HAVE_SYS_VFS_H.
1985 28. Installed the latest version of exipick from John Jetmore.
1987 29. In an address list, if the pattern was not a regular expression, an empty
1988 subject address (from a bounce message) matched only if the pattern was an
1989 empty string. Non-empty patterns were not even tested. This was the wrong
1990 because it is perfectly reasonable to use an empty address as part of a
1991 database query. An empty address is now tested by patterns that are
1992 lookups. However, all the other forms of pattern expect the subject to
1993 contain a local part and a domain, and therefore, for them, an empty
1994 address still always fails if the pattern is not itself empty.
1996 30. Exim went into a mad DNS loop when attempting to do a callout where the
1997 host was specified on an smtp transport, and looking it up yielded more
1998 than one IP address.
2000 31. Re-factored the code for checking spool and log partition space into a
2001 function that finds that data and another that does the check. The former
2002 is then used to implement four new variables: $spool_space, $log_space,
2003 $spool_inodes, and $log_inodes.
2005 32. The RFC2047 encoding function was originally intended for short strings
2006 such as real names; it was not keeping to the 75-character limit for
2007 encoded words that the RFC imposes. It now respects the limit, and
2008 generates multiple encoded words if necessary. To be on the safe side, I
2009 have increased the buffer size for the ${rfc2047: expansion operator from
2012 33. It is now permitted to omit both strings after an "if" condition; if the
2013 condition is true, the result is "true". As before, when the second string
2014 is omitted, a false condition yields an empty string. This makes it less
2015 cumbersome to write custom ACL and router conditions.
2017 34. Failure to deliver a bounce message always caused it to be frozen, even if
2018 there was an errors_to setting on the router. The errors_to setting is now
2021 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2022 canonical form (fully expanded) before being placed in
2023 $sender_host_address.
2025 36. The table in the code that translates DNS record types into text (T_A to
2026 "A" for instance) was missing entries for NS and CNAME. It is just possible
2027 that this could have caused confusion if both these types were looked up
2028 for the same domain, because the text type is used as part of Exim's
2029 per-process caching. But the chance of anyone hitting this buglet seems
2032 37. The dnsdb lookup has been extended in a number of ways.
2034 (1) There is a new type, "zns", which walks up the domain tree until it
2035 finds some nameserver records. It should be used with care.
2037 (2) There is a new type, "mxh", which is like "mx" except that it returns
2038 just the host names, not the priorities.
2040 (3) It is now possible to give a list of domains (or IP addresses) to be
2041 looked up. The behaviour when one of the lookups defers can be
2042 controlled by a keyword.
2044 (4) It is now possible to specify the separator character for use when
2045 multiple records are returned.
2047 38. The dnslists ACL condition has been extended: it is now possible to supply
2048 a list of IP addresses and/or domains to be looked up in a particular DNS
2051 39. Added log_selector=+queue_time_overall.
2053 40. When running the queue in the test harness, wait just a tad after forking a
2054 delivery process, to get repeatability of debugging output.
2056 41. Include certificate and key file names in error message when GnuTLS fails
2057 to set them up, because the GnuTLS error message doesn't include the name
2058 of the failing file when there is a problem reading it.
2060 42. Allow both -bf and -bF in the same test run.
2062 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2064 44. The "Exiscan patch" is now merged into the mainline Exim source.
2066 45. Sometimes the final signoff response after QUIT could fail to get
2067 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2068 < 0 before doing a fflush(). This bug looks as though it goes back to the
2069 introduction of TLS in release 3.20, but "sometimes" must have been rare
2070 because the tests only now provoked it.
2072 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2073 (this can affect the format of dates).
2075 47. exim_tidydb, when checking for the continued existence of a message for
2076 which it has found a message-specific retry record, was not finding
2077 messages that were in split spool directories. Consequently, it was
2078 deleting retry records that should have stayed in existence.
2080 48. Steve fixed some bugs in eximstats.
2082 49. The SPA authentication driver was not abandoning authentication and moving
2083 on to the next authenticator when an expansion was forced to fail,
2084 contradicting the general specification for all authenticators. Instead it
2085 was generating a temporary error. It now behaves as specified.
2087 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2088 (the order specifies the preference for clients). The order is now AES256,
2089 AES128, 3DES, ARCFOUR128.
2091 51. Small patch to Sieve code - explicitly set From: when generating an
2094 52. Exim crashed if a remote delivery caused a very long error message to be
2095 recorded - for instance if somebody sent an entire SpamAssassin report back
2096 as a large number of 550 error lines. This bug was coincidentally fixed by
2097 increasing the size of one of Exim's internal buffers (big_buffer) that
2098 happened as part of the Exiscan merge. However, to be on the safe side, I
2099 have made the code more robust (and fixed the comments that describe what
2102 53. Now that there can be additional text after "Completed" in log lines (if
2103 the queue_time_overall log selector is set), a one-byte patch to exigrep
2104 was needed to allow it to recognize "Completed" as not the last thing in
2107 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2108 patch that reportedly fixes this has been added. I am not expert enough to
2109 create a test for it. This is what the patch creator wrote:
2111 "I found a little strange behaviour of ldap code when working with
2112 Windows 2003 AD Domain, where users was placed in more than one
2113 Organization Units. When I tried to give exim partial DN, the exit code
2114 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2115 But simultaneously result of request was absolutely normal ldap result,
2116 so I produce this patch..."
2118 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2119 REFERENCE, so I have modified the code to exclude the patch when that macro
2122 55. Some experimental protocols are using DNS PTR records for new purposes. The
2123 keys for these records are domain names, not reversed IP addresses. The
2124 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2125 leaves it alone. Component reversal etc. now happens only for IP addresses.
2128 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2130 57. Double the size of the debug message buffer (to 2048) so that more of very
2131 long debug lines gets shown.
2133 58. The exicyclog utility now does better if the number of log files to keep
2134 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2136 59. Two changes related to the smtp_active_hostname option:
2138 (1) $smtp_active_hostname is now available as a variable.
2139 (2) The default for smtp_banner uses $smtp_active_hostname instead
2140 of $primary_hostname.
2142 60. The host_aton() function is supposed to be passed a string that is known
2143 to be a valid IP address. However, in the case of IPv6 addresses, it was
2144 not checking this. This is a hostage to fortune. Exim now panics and dies
2145 if the condition is not met. A case was found where this could be provoked
2146 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2147 components; fortuitously, this particular loophole had already been fixed
2148 by change 4.50/55 above.
2150 If there are any other similar loopholes, the new check in host_aton()
2151 itself should stop them being exploited. The report I received stated that
2152 data on the command line could provoke the exploit when Exim was running as
2153 exim, but did not say which command line option was involved. All I could
2154 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2155 running as the user.
2158 61. There was a buffer overflow vulnerability in the SPA authentication code
2159 (which came originally from the Samba project). I have added a test to the
2160 spa_base64_to_bits() function which I hope fixes it.
2163 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2164 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2166 63. The daemon start-up calls getloadavg() while still root for those OS that
2167 need the first call to be done as root, but it missed one case: when
2168 deliver_queue_load_max is set with deliver_drop_privilege. This is
2169 necessary for the benefit of the queue runner, because there is no re-exec
2170 when deliver_drop_privilege is set.
2172 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2173 This has been fixed.
2175 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2176 was in use, was not putting the data itself into the right store pool;
2177 consequently, it could be overwritten for a subsequent message in the same
2178 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2181 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2183 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2184 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2185 the function were treating the return as a boolean value, which happened to
2186 work because 0=false and not-0=true, but is not correct code.
2188 68. The host_aton() function was not handling scoped IPv6 addresses (those
2189 with, for example, "%eth0" on the end) correctly.
2191 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2192 negated items (that is, ~something) in unsigned ints. Some compilers
2193 apparently mutter when there is no cast.
2195 70. If an address verification called from an ACL failed, and did not produce a
2196 user-specific message (i.e. there was only a "system" message), nothing was
2197 put in $acl_verify_message. In this situation, it now puts the system
2200 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2201 session; change 4.31/43 added the unwanted input to the log line - except
2202 that it did not do this in the start of session case. It now does.
2204 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2205 This is wrong and can cause the other end to generate a synchronization
2206 error if it is another Exim or anything else that does the synchronization
2207 check. A QUIT command is no longer sent after a timeout.
2209 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2210 during host lookups.
2212 74. The defer_ok option of callout verification was not working if it was used
2213 when verifying addresses in header lines, that is, for this case:
2215 verify = header_sender/callout=defer_ok
2217 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2218 those file descriptors could be used for SMTP connections. If anything
2219 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2220 could be sent to the SMTP client, causing chaos. The daemon now opens
2221 stdin, stdout, and stderr to /dev/null when it puts itself into the
2224 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2225 log by default. The user can override this with suitable Perl magic.
2227 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2228 the log message when discard triggers, was not working for the DATA ACL or
2229 for the non-SMTP ACL.
2231 78. Error message wording change in sieve.c.
2233 79. If smtp_accept_max_per_host was set, the number of connections could be
2234 restricted to fewer than expected, because the daemon was trying to set up
2235 a new connection before checking whether the processes handling previous
2236 connections had finished. The check for completed processes is now done
2237 earlier. On busy systems, this bug wouldn't be noticed because something
2238 else would have woken the daemon, and it would have reaped the completed
2241 80. If a message was submitted locally by a user whose login name contained one
2242 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2243 It caused a spool format error. I have fixed the spool reading code. A
2244 related problem was that the "from" clause in the Received: line became
2245 illegal because of the space(s). It is now covered by ${quote_local_part.
2247 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2250 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2252 83. Patch from Sieve maintainer to fix unterminated string problem in
2253 "vacation" handling.
2255 84. Some minor changes to the Linux configuration files to help with other
2256 OS variants using glibc.
2258 85. One more patch for Sieve to update vacation handling to latest spec.
2261 ----------------------------------------------------
2262 See the note above about the 4.44 and 4.50 releases.
2263 ----------------------------------------------------
2269 1. Change 4.43/35 introduced a bug that caused file counts to be
2270 incorrectly computed when quota_filecount was set in an appendfile
2273 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2274 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2277 3. Give more explanation in the error message when the command for a transport
2278 filter fails to execute.
2280 4. There are several places where Exim runs a non-Exim command in a
2281 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2282 was being done only for the command run by the queryprogram router. It is
2283 now done for all such subprocesses. The other cases are: ${run, transport
2284 filters, and the commands run by the lmtp and pipe transports.
2286 5. Some older OS have a limit of 256 on the maximum number of file
2287 descriptors. Exim was using setrlimit() to set 1000 as a large value
2288 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2289 systems. I've change it so that if it can't get 1000, it tries for 256.
2291 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2292 was an oversight, and furthermore, ever since the addition of extra
2293 controls (e.g. 4.43/32), the checks on when to allow different forms of
2294 "control" were broken. There should now be diagnostics for all cases when a
2295 control that does not make sense is encountered.
2297 7. $recipients is now available in the predata ACL (oversight).
2299 8. Tidy the search cache before the fork to do a delivery from a message
2300 received from the command line. Otherwise the child will trigger a lookup
2301 failure and thereby defer the delivery if it tries to use (for example) a
2302 cached ldap connection that the parent has called unbind on.
2304 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2305 of $address_data from the recipient verification was clobbered by the
2306 sender verification.
2308 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2309 was its contents. (It was OK if the option was not defined at all.)
2311 11. A "Completed" log line is now written for messages that are removed from
2312 the spool by the -Mrm option.
2314 12. $host_address is now set to the target address during the checking of
2315 ignore_target_hosts.
2317 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2318 being passed; this would have caused $sender_host_name to have been used if
2319 matching the list had actually called for a host name (not very likely,
2320 since this list is usually IP addresses). A host name is now passed as
2323 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2324 code that specifies a non-restarting handler (typically sigaction() in
2325 modern systems) in an attempt to fix a rare and obscure crash bug.
2327 15. Narrowed the window for a race in the daemon that could cause it to ignore
2328 SIGCHLD signals. This is not a major problem, because they are used only to
2329 wake it up if nothing else does.
2331 16. A malformed maildirsize file could cause Exim to calculate negative values
2332 for the mailbox size or file count. Odd effects could occur as a result.
2333 The maildirsize information is now recalculated if the size or filecount
2336 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2337 support for a long time. Removed HAVE_SYS_VFS_H.
2339 18. Updated exipick to current release from John Jetmore.
2341 19. Allow an empty sender to be matched against a lookup in an address list.
2342 Previously the only cases considered were a regular expression, or an
2345 20. Exim went into a mad DNS lookup loop when doing a callout where the
2346 host was specified on the transport, if the DNS lookup yielded more than
2349 21. The RFC2047 encoding function was originally intended for short strings
2350 such as real names; it was not keeping to the 75-character limit for
2351 encoded words that the RFC imposes. It now respects the limit, and
2352 generates multiple encoded words if necessary. To be on the safe side, I
2353 have increased the buffer size for the ${rfc2047: expansion operator from
2356 22. Failure to deliver a bounce message always caused it to be frozen, even if
2357 there was an errors_to setting on the router. The errors_to setting is now
2360 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2361 canonical form (fully expanded) before being placed in
2362 $sender_host_address.
2364 24. Updated eximstats to version 1.33
2366 25. Include certificate and key file names in error message when GnuTLS fails
2367 to set them up, because the GnuTLS error message doesn't include the name
2368 of the failing file when there is a problem reading it.
2370 26. Expand error message when OpenSSL has problems setting up cert/key files.
2373 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2374 (this can affect the format of dates).
2376 28. exim_tidydb, when checking for the continued existence of a message for
2377 which it has found a message-specific retry record, was not finding
2378 messages that were in split spool directories. Consequently, it was
2379 deleting retry records that should have stayed in existence.
2381 29. eximstats updated to version 1.35
2382 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2383 1.35 - bugfix such that pie charts by volume are generated correctly
2385 30. The SPA authentication driver was not abandoning authentication and moving
2386 on to the next authenticator when an expansion was forced to fail,
2387 contradicting the general specification for all authenticators. Instead it
2388 was generating a temporary error. It now behaves as specified.
2390 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2391 (the order specifies the preference for clients). The order is now AES256,
2392 AES128, 3DES, ARCFOUR128.
2394 31. Small patch to Sieve code - explicitly set From: when generating an
2397 32. Exim crashed if a remote delivery caused a very long error message to be
2398 recorded - for instance if somebody sent an entire SpamAssassin report back
2399 as a large number of 550 error lines. This bug was coincidentally fixed by
2400 increasing the size of one of Exim's internal buffers (big_buffer) that
2401 happened as part of the Exiscan merge. However, to be on the safe side, I
2402 have made the code more robust (and fixed the comments that describe what
2405 33. Some experimental protocols are using DNS PTR records for new purposes. The
2406 keys for these records are domain names, not reversed IP addresses. The
2407 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2408 leaves it alone. Component reversal etc. now happens only for IP addresses.
2411 34. The host_aton() function is supposed to be passed a string that is known
2412 to be a valid IP address. However, in the case of IPv6 addresses, it was
2413 not checking this. This is a hostage to fortune. Exim now panics and dies
2414 if the condition is not met. A case was found where this could be provoked
2415 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2416 components; fortuitously, this particular loophole had already been fixed
2417 by change 4.50/55 or 4.44/33 above.
2419 If there are any other similar loopholes, the new check in host_aton()
2420 itself should stop them being exploited. The report I received stated that
2421 data on the command line could provoke the exploit when Exim was running as
2422 exim, but did not say which command line option was involved. All I could
2423 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2424 running as the user.
2427 35. There was a buffer overflow vulnerability in the SPA authentication code
2428 (which came originally from the Samba project). I have added a test to the
2429 spa_base64_to_bits() function which I hope fixes it.
2432 36. The daemon start-up calls getloadavg() while still root for those OS that
2433 need the first call to be done as root, but it missed one case: when
2434 deliver_queue_load_max is set with deliver_drop_privilege. This is
2435 necessary for the benefit of the queue runner, because there is no re-exec
2436 when deliver_drop_privilege is set.
2438 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2439 was in use, was not putting the data itself into the right store pool;
2440 consequently, it could be overwritten for a subsequent message in the same
2441 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2444 38. Sometimes the final signoff response after QUIT could fail to get
2445 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2446 < 0 before doing a fflush(). This bug looks as though it goes back to the
2447 introduction of TLS in release 3.20, but "sometimes" must have been rare
2448 because the tests only now provoked it.
2454 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2455 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2456 Now it returns an integer. A number of calls were still expecting a T/F
2457 return. Fortuitously, in all cases, the tests worked in OK situations,
2458 which is the norm. However, things would have gone wrong on any write
2459 failures on the smtp file descriptor. This function is used when sending
2460 messages over SMTP and also when doing verify callouts.
2462 2. When Exim is called to do synchronous delivery of a locally submitted
2463 message (the -odf or -odi options), it no longer closes stderr before doing
2466 3. Implemented the mua_wrapper option.
2468 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2470 5. Implemented the functions header_remove(), header_testname(),
2471 header_add_at_position(), and receive_remove_recipient(), and exported them
2474 6. If an ACL "warn" statement specified the addition of headers, Exim already
2475 inserted X-ACL-Warn: at the start if there was no header name. However, it
2476 was not making this test for the second and subsequent header lines if
2477 there were newlines in the string. This meant that an invalid header could
2478 be inserted if Exim was badly configured.
2480 7. Allow an ACL "warn" statement to add header lines at the start or after all
2481 the Received: headers, as well as at the end.
2483 8. Added the rcpt_4xx retry error code.
2485 9. Added postmaster_mailfrom=xxx to callout verification option.
2487 10. Added mailfrom=xxxx to the callout verification option, for verify=
2490 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2491 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2493 12. Inserted some casts to stop certain compilers warning when using pointer
2494 differences as field lengths or precisions in printf-type calls (mostly
2495 affecting debugging statements).
2497 13. Added optional readline() support for -be (dynamically loaded).
2499 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2500 same clock tick as a message's arrival, so that its received time was the
2501 same as the "first fail" time on the retry record, and that message
2502 remained on the queue past the ultimate address timeout, every queue runner
2503 would try a delivery (because it was past the ultimate address timeout) but
2504 after another failure, the ultimate address timeout, which should have then
2505 bounced the address, did not kick in. This was a "< instead of <=" error;
2506 in most cases the first failure would have been in the next clock tick
2507 after the received time, and all would be well.
2509 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2510 being recognized when the domain list was tested by the match_domain
2511 condition in an expansion string.
2513 16. Added the ${str2b64: operator.
2515 17. Exim was always calling setrlimit() to set a large limit for the number of
2516 processes, without checking whether the existing limit was already
2517 adequate. (It did check for the limit on file descriptors.) Furthermore,
2518 errors from getrlimit() and setrlimit() were being ignored. Now they are
2519 logged to the main and panic logs, but Exim does carry on, to try to do its
2520 job under whatever limits there are.
2522 18. Imported PCRE 5.0.
2524 19. Trivial typo in log message " temporarily refused connection" (the leading
2527 20. If the log selector return_path_on_delivery was set and an address was
2528 redirected to /dev/null, the delivery process crashed because it assumed
2529 that a return path would always be set for a "successful" delivery. In this
2530 case, the whole delivery is bypassed as an optimization, and therefore no
2533 21. Internal re-arrangement: the function for sending a challenge and reading
2534 a response while authentication was assuming a zero-terminated challenge
2535 string. It's now changed to take a pointer and a length, to allow for
2536 binary data in such strings.
2538 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2540 23. Exim was not respecting finduser_retries when seeking the login of the
2541 uid under which it was called; it was always trying 10 times. (The default
2542 setting of finduser_retries is zero.) Also, it was sleeping after the final
2543 failure, which is pointless.
2545 24. Implemented tls_on_connect_ports.
2547 25. Implemented acl_smtp_predata.
2549 26. If the domain in control=submission is set empty, Exim assumes that the
2550 authenticated id is a complete email address when it generates From: or
2551 Sender: header lines.
2553 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2554 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2555 chown and chgrp in /bin and hostname in /usr/bin.
2557 28. Exim was keeping the "process log" file open after each use, just as it
2558 does for the main log. This opens the possibility of it remaining open for
2559 long periods when the USR1 signal hits a daemon. Occasional processlog
2560 errors were reported, that could have been caused by this. Anyway, it seems
2561 much more sensible not to leave this file open at all, so that is what now
2564 29. The long-running daemon process does not normally write to the log once it
2565 has entered its main loop, and it closes the log before doing so. This is
2566 so that log files can straightforwardly be renamed and moved. However,
2567 there are a couple of unusual error situations where the daemon does write
2568 log entries, and I had neglected to close the log afterwards.
2570 30. The text of an SMTP error response that was received during a remote
2571 delivery was being truncated at 512 bytes. This is too short for some of
2572 the long messages that one sometimes sees. I've increased the limit to
2575 31. It is now possible to make retry rules that apply only when a message has a
2576 specific sender, in particular, an empty sender.
2578 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2579 it possible to be selective about when SMTP synchronization is enforced.
2581 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2583 32. Implemented hosts_connection_nolog.
2585 33. Added an ACL for QUIT.
2587 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2590 35. Added mailbox_size and mailbox_filecount to appendfile.
2592 36. Added control = no_multiline_responses to ACLs.
2594 37. There was a bug in the logic of the code that waits for the clock to tick
2595 in the case where the clock went backwards by a substantial amount such
2596 that the microsecond fraction of "now" was more than the microsecond
2597 fraction of "then" (but the whole seconds number was less).
2599 38. Added support for the libradius Radius client library this is found on
2600 FreeBSD (previously only the radiusclient library was supported).
2606 1. When certain lookups returned multiple values in the form name=value, the
2607 quoting of the values was not always being done properly. Specifically:
2608 (a) If the value started with a double quote, but contained no whitespace,
2610 (b) If the value contained whitespace other than a space character (i.e.
2611 tabs or newlines or carriage returns) it was not quoted.
2612 This fix has been applied to the mysql and pgsql lookups by writing a
2613 separate quoting function and calling it from the lookup code. The fix
2614 should probably also be applied to nisplus, ibase and oracle lookups, but
2615 since I cannot test any of those, I have not disturbed their existing code.
2617 2. A hit in the callout cache for a specific address caused a log line with no
2618 reason for rejecting RCPT. Now it says "Previous (cached) callout
2619 verification failure".
2621 3. There was an off-by-one bug in the queryprogram router. An over-long
2622 return line was truncated at 256 instead of 255 characters, thereby
2623 overflowing its buffer with the terminating zero. As well as fixing this, I
2624 have increased the buffer size to 1024 (and made a note to document this).
2626 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2627 when Exim is waiting for an SMTP response from a remote server, Exim
2628 restarts its select() call on the socket, thereby resetting its timeout.
2629 This is not a problem when such interrupts are rare. Somebody set up a cron
2630 job to run exiwhat every 2 minutes, which is less than the normal select()
2631 timeout (5 or 10 minutes). This meant that the select() timeout never
2632 kicked in because it was always reset. I have fixed this by comparing the
2633 time when an interrupt arrives with the time at the start of the first call
2634 to select(). If more time than the timeout has elapsed, the interrupt is
2635 treated as a timeout.
2637 5. Some internal re-factoring in preparation for the addition of Sieve
2638 extensions (by MH). In particular, the "personal" test is moved to a
2639 separate function, and given an option for scanning Cc: and Bcc: (which is
2640 not set for Exim filters).
2642 6. When Exim created an email address using the login of the caller as the
2643 local part (e.g. when creating a From: or Sender: header line), it was not
2644 quoting the local part when it contained special characters such as @.
2646 7. Installed new OpenBSD configuration files.
2648 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2649 try to make them clearer.
2651 9. Callout options, other than the timeout value, were being ignored when
2652 verifying sender addresses in header lines. For example, when using
2654 verify = header_sender/callout=no_cache
2656 the cache was (incorrectly) being used.
2658 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2659 only the Cygwin environment.
2661 11. When return_path_on_delivery was set as a log selector, if different remote
2662 addresses in the same message used different return paths and parallel
2663 remote delivery occurred, the wrong values would sometimes be logged.
2664 (Whenever a remote delivery process finished, the return path value from
2665 the most recently started remote delivery process was logged.)
2667 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2668 lines when AUTH and/or TLS are in use. This is the "received protocol"
2669 field. Exim used to use "asmtp" for authenticated SMTP, without any
2670 indication (in the protocol name) for TLS use. Now it follows the RFC and
2671 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2672 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2673 names appear in log lines as well as in Received: header lines.
2675 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2676 extensions, and comparison tests, and to fix some bugs.
2678 14. Changes to the "personal" filter test:
2680 (1) The test was buggy in that it was just doing the equivalent of
2681 "contains" tests on header lines. For example, if a user's address was
2682 anne@some.where, the "personal" test would incorrectly be true for
2684 To: susanne@some.where
2686 This test is now done by extracting each address from the header in turn,
2687 and checking the entire address. Other tests that are part of "personal"
2688 are now done using regular expressions (for example, to check local parts
2689 of addresses in From: header lines).
2691 (2) The list of non-personal local parts in From: addresses has been
2692 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2693 taken from the Sieve specification recommendations.
2695 (3) If the message contains any header line starting with "List-" it is
2696 treated as non-personal.
2698 (4) The test for "circular" in the Subject: header line has been removed
2699 because it now seems ill-conceived.
2701 15. Minor typos in src/EDITME comments corrected.
2703 16. Installed latest exipick from John Jetmore.
2705 17. If headers_add on a router specified a text string that was too long for
2706 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2707 of string_sprintf() is now avoided.
2709 18. $message_body_size was not set (it was always zero) when running the DATA
2710 ACL and the local_scan() function.
2712 19. For the "mail" command in an Exim filter, no default was being set for
2713 the once_repeat time, causing a random time value to be used if "once" was
2714 specified. (If the value happened to be <= 0, no repeat happened.) The
2715 default is now 0s, meaning "never repeat". The "vacation" command was OK
2716 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2717 (I found it when inspecting the code).
2719 20. There is now an overall timeout for performing a callout verification. It
2720 defaults to 4 times the callout timeout, which applies to individual SMTP
2721 commands during the callout. The overall timeout applies when there is more
2722 than one host that can be tried. The timeout is checked before trying the
2723 next host. This prevents very long delays if there are a large number of
2724 hosts and all are timing out (e.g. when the network connections are timing
2725 out). The value of the overall timeout can be changed by specifying an
2726 additional sub-option for "callout", called "maxwait". For example:
2728 verify = sender/callout=5s,maxwait=20s
2730 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2731 to the end before writing, but this should make it even safer).
2733 22. Exim was forgetting that it had advertised PIPELINING for the second and
2734 subsequent messages on an SMTP connection. It was also not resetting its
2735 memory on STARTTLS and an internal HELO.
2737 23. When Exim logs an SMTP synchronization error within a session, it now
2738 records whether PIPELINING has been advertised or not.
2740 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2741 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2742 rather than long int.
2744 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2746 26. Added the never_mail option to autoreply.
2752 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2753 crash if the getsockname() call failed; this can happen if a connection is
2754 closed very soon after it is established. The problem was simply in the
2755 order in which certain operations were done, causing Exim to try to write
2756 to the SMTP stream before it had set up the file descriptor. The bug has
2757 been fixed by making things happen in the correct order.
2763 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2764 before the connection was closed, thus losing the rejection response.
2766 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2767 some early Solaris releases, but causes trouble in current releases where
2768 socklen_t is defined.
2770 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2773 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2776 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2777 that releasing the top of it at the end releases what was used for sub-
2778 expansions (unless the block got too big). However, discard this block if
2779 the first thing is a variable or header, so that we can use its block when
2780 it is dynamic (useful for very large $message_headers, for example).
2782 6. Lookups now cache *every* query, not just the most recent. A new, separate
2783 store pool is used for this. It can be recovered when all lookup caches are
2784 flushed. Lookups now release memory at the end of their result strings.
2785 This has involved some general refactoring of the lookup sources.
2787 7. Some code has been added to the store_xxx() functions to reduce the amount
2788 of flapping under certain conditions.
2790 8. log_incoming_interface used to affect only the <= reception log lines. Now
2791 it causes the local interface and port to be added to several more SMTP log
2792 lines, for example "SMTP connection from", and rejection lines.
2794 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2796 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2798 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2799 could be overwritten at the end of the current message (or the start of a
2800 new message if it was set in a HELO ACL). The value is now preserved for
2801 the duration of the SMTP connection.
2803 12. If a transport had a headers_rewrite setting, and a matching header line
2804 contained an unqualified address, that address was qualified, even if it
2805 did not match any rewriting rules. The underlying bug was that the values
2806 of the flags that permit the existence of unqualified sender and recipient
2807 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2808 non-local messages, and by -bnq for local messages) were not being
2809 preserved with the message after it was received.
2811 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2812 "next input=" as part of the text comprising the host identity instead of
2813 the correct text. The code was using the same buffer for two different
2814 strings. However, depending on which order the printing function evaluated
2815 its arguments, the bug did not always show up. Under Linux, for example, my
2816 test suite worked just fine.
2818 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2819 doesn't work with some older versions of Perl. It has been changed to "my",
2820 which in any case is probably the better facility to use.
2822 15. A really picky compiler found some instances of statements for creating
2823 error messages that either had too many or two few arguments for the format
2826 16. The size of the buffer for calls to the DNS resolver has been increased
2827 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2828 for addresses that have a lot of PTR records. This alleviates a problem; it
2829 does not fully solve it.
2831 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2832 buffer now truncates the list and logs the incident, which is the same
2833 action as happens when Exim is looking up a host name and its aliases.
2834 Previously in this situation something unpredictable would happen;
2835 sometimes it was "internal error: store_reset failed".
2837 18. If a server dropped the connection unexpectedly when an Exim client was
2838 using GnuTLS and trying to read a response, the client delivery process
2839 crashed while trying to generate an error log message.
2841 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2842 string, for example:
2844 warn message = H1: something\nH2: something
2846 the text was added as a single header line from Exim's point of view
2847 though it ended up OK in the delivered message. However, searching for the
2848 second and subsequent header lines using $h_h2: did not work. This has been
2849 fixed. Similarly, if a system filter added multiple headers in this way,
2850 the routers could not see them.
2852 20. Expanded the error message when iplsearch is called with an invalid key to
2853 suggest using net-iplsearch in a host list.
2855 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2856 ACLs are no longer actually imposed (and a message to that effect is
2859 22. If a "gecos" field in a passwd entry contained escaped characters, in
2860 particular, if it contained a \" sequence, Exim got it wrong when building
2861 a From: or a Sender: header from that name. A second bug also caused
2862 incorrect handling when an unquoted " was present following a character
2863 that needed quoting.
2865 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2866 was not being matched caselessly.
2868 24. Arranged for all hyphens in the exim.8 source to be escaped with
2871 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2872 properly. Recipient callout cache records were still being keyed to include
2873 the sender, even when use_sender was set false. This led to far more
2874 callouts that were necessary. The sender is no longer included in the key
2875 when use_sender is false.
2877 26. Added "control = submission" modifier to ACLs.
2879 27. Added the ${base62d: operator to decode base 62 numbers.
2881 28. dnsdb lookups can now access SRV records.
2883 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2884 the configuration file.
2886 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2887 (-v) mode. This makes the output for a verbose queue run more intelligible.
2889 31. Added a use_postmaster feature to recipient callouts.
2891 32. Added the $body_zerocount variable, containing the number of binary zero
2892 bytes in the message body.
2894 33. The time of last modification of the "new" subdirectory is now used as the
2895 "mailbox time last read" when there is a quota error for a maildir
2898 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2900 35. Added +ignore_unknown as a special item in host lists.
2902 36. Code for decoding IPv6 addresses in host lists is now included, even if
2903 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2904 address was recognized as an IP address, but was then not correctly decoded
2905 into binary, causing unexpected and incorrect effects when compared with
2912 1. Very minor rewording of debugging text in manualroute to say "list of
2913 hosts" instead of "hostlist".
2915 2. If verify=header_syntax was set, and a header line with an unqualified
2916 address (no domain) and a large number of spaces between the end of the
2917 name and the colon was received, the reception process suffered a buffer
2918 overflow, and (when I tested it) crashed. This was caused by some obsolete
2919 code that should have been removed. The fix is to remove it!
2921 3. When running in the test harness, delay a bit after writing a bounce
2922 message to get a bit more predictability in the log output.
2924 4. Added a call to search_tidyup() just before forking a reception process. In
2925 theory, someone could use a lookup in the expansion of smtp_accept_max_
2926 per_host which, without the tidyup, could leave open a database connection.
2928 5. Added the variables $recipient_data and $sender_data which get set from a
2929 lookup success in an ACL "recipients" or "senders" condition, or a router
2930 "senders" option, similar to $domain_data and $local_part_data.
2932 6. Moved the writing of debug_print from before to after the "senders" test
2935 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2936 problems for message scanning, either using a data ACL, or using
2937 local_scan() because the Received: header was not generated till after they
2938 were called (in order to set the time as the time of reception completion).
2939 I have revised the way this works. The header is now generated after the
2940 body is received, but before the ACL or local_scan() are called. After they
2941 are run, the timestamp in the header is updated.
2947 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2948 before starting a queue runner without re-exec. This happened only when
2949 deliver_drop_privilege was set or when the Exim user was set to root. The
2950 effect of the bug was that timeouts during subsequent deliveries caused
2951 crashes instead of being properly handled. The handler is now left at its
2952 default (and expected) setting.
2954 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2955 message, again when deliver_drop_privilege is set or Exim is run as root.
2956 The bug described in (1) was not present in this case, but the tidying up
2957 of the other signals was missing. I have made the two cases consistent.
2959 3. The ignore_target_hosts setting on a manualroute router was being ignored
2960 for hosts that were looked up using the /MX notation.
2962 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2965 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2966 operated on the sender address. After changing the $sender_address to <>
2967 for the sender address verify, Exim was re-instated it as the original
2968 (before rewriting) address, but remembering that it had rewritten it, so it
2969 wasn't rewriting it again. This bug also had the effect of breaking the
2970 sender address verification caching when the sender address was rewritten.
2972 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2973 This has been changed so that if the ip literal address matches
2974 ignore_target_hosts, the router declines.
2976 7. Added expansion conditions match_domain, match_address, and match_local_
2977 part (NOT match_host).
2979 8. The placeholder for the Received: header didn't have a length field set.
2981 9. Added code to Exim itself and to exim_lock to test for a specific race
2982 condition that could lead to file corruption when using MBX delivery. The
2983 issue is with the lockfile that is created in /tmp. If this file is removed
2984 after a process has opened it but before that process has acquired a lock,
2985 there is the potential for a second process to recreate the file and also
2986 acquire a lock. This could lead to two Exim processes writing to the file
2987 at the same time. The added code performs the same test as UW imapd; it
2988 checks after acquiring the lock that its file descriptor still refers to
2989 the same named file.
2991 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2992 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2995 11. Added the smtp_active_hostname option. If used, this will typically be made
2996 to depend on the incoming interface address. Because $interface_address is
2997 not set up until the daemon has forked a reception process, error responses
2998 that can happen earlier (such as "too many connections") no longer contain
3001 12. If an expansion in a condition on a "warn" statement fails because a lookup
3002 defers, the "warn" statement is abandoned, and the next ACL statement is
3003 processed. Previously this caused the whole ACL to be aborted.
3005 13. Added the iplsearch lookup type.
3007 14. Added ident_timeout as a log selector.
3009 15. Added tls_certificate_verified as a log selector.
3011 16. Added a global option tls_require_ciphers (compare the smtp transport
3012 option of the same name). This controls incoming TLS connections.
3014 17. I finally figured out how to make tls_require_ciphers do a similar thing
3015 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3016 before starting the TLS session.
3018 18. Tabs are now shown as \t in -bP output.
3020 19. If the log selector return_path_on_delivery was set, Exim crashed when
3021 bouncing a message because it had too many Received: header lines.
3023 20. If two routers both had headers_remove settings, and the first one included
3024 a superfluous trailing colon, the final name in the first list and the
3025 first name in the second list were incorrectly joined into one item (with a
3026 colon in the middle).
3032 1. Added -C and -D options to the exinext utility, mainly to make it easier
3033 to include in the automated testing, but these could be helpful when
3034 multiple configurations are in use.
3036 2. The exinext utility was not formatting the output nicely when there was
3037 an alternate port involved in the retry record key, nor when there was a
3038 message id as well (for retries that were specific to a specific message
3039 and a specific host). It was also confused by IPv6 addresses, because of
3040 the additional colons they contain. I have fixed the IPv4 problem, and
3041 patched it up to do a reasonable job for IPv6.
3043 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3044 delivery, the log line now contains "pipelined" if PIPELINING was used.
3046 4. An SMTP transport process used to panic and die if the bind() call to set
3047 an explicit outgoing interface failed. This has been changed; it is now
3048 treated in the same way as a connect() failure.
3050 5. A reference to $sender_host_name in the part of a conditional expansion
3051 that was being skipped was still causing a DNS lookup. This no longer
3054 6. The def: expansion condition was not recognizing references to header lines
3055 that used bh_ and bheader_.
3057 7. Added the _cache feature to named lists.
3059 8. The code for checking quota_filecount in the appendfile transport was
3060 allowing one more file than it should have been.
3062 9. For compatibility with Sendmail, the command line option
3070 and sets the incoming protocol and host name (for trusted callers). The
3071 host name and its colon can be omitted when only the protocol is to be set.
3072 Note the Exim already has two private options, -pd and -ps, that refer to
3073 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3074 "s", but I don't think that's a major issue.
3076 10. A number of refactoring changes to the code, none of which should affect
3079 (a) The number of logging options was getting close to filling up the
3080 32-bit word that was used as a bit map. I have split them into two classes:
3081 those that are passed in the argument to log_write(), and those that are
3082 only ever tested independently outside of that function. These are now in
3083 separate 32-bit words, so there is plenty of room for expansion again.
3084 There is no change in the user interface or the logging behaviour.
3086 (b) When building, for example, log lines, the code previously used a
3087 macro that called string_cat() twice, in order to add two strings. This is
3088 not really sufficiently general. Furthermore, there was one instance where
3089 it was actually wrong because one of the argument was used twice, and in
3090 one call a function was used. (As it happened, calling the function twice
3091 did not affect the overall behaviour.) The macro has been replaced by a
3092 function that can join an arbitrary number of extra strings onto a growing
3095 (c) The code for expansion conditions now uses a table and a binary chop
3096 instead of a serial search (which was left over from when there were very
3097 few conditions). Also, it now recognizes conditions like "pam" even when
3098 the relevant support is not compiled in: a suitably worded error message is
3099 given if an attempt is made to use such a condition.
3101 11. Added ${time_interval:xxxxx}.
3103 12. A bug was causing one of the ddress fields not to be passed back correctly
3104 from remote delivery subprocesses. The field in question was not being
3105 subsequently used, so this caused to problems in practice.
3107 13. Added new log selectors queue_time and deliver_time.
3109 14. Might have fixed a bug in maildirsizefile handling that threw up
3110 "unexpected character" debug warnings, and recalculated the data
3111 unnecessarily. In any case, I expanded the warning message to give more
3114 15. Added the message "Restricted characters in address" to the statements in
3115 the default ACL that block characters like @ and % in local parts.
3117 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3118 Three changes have been made:
3120 (a) There was a serious bug; a negative response to MAIL caused the whole
3121 recipient domain to be cached as invalid, thereby blocking all messages
3122 to all local parts at the same domain, from all senders. This bug has
3123 been fixed. The domain is no longer cached after a negative response to
3124 MAIL if the sender used is not empty.
3126 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3129 (c) A new callout option, "use_sender" has been added for people who want
3130 the modified behaviour.
3136 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3139 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3140 indeed breaks things for older releases.
3142 3. Added additional logging to the case where there is a problem reading data
3143 from a filter that is running in a subprocess using a pipe, in order to
3144 try to track down a specific problem.
3146 4. Testing facility fudge: when running in the test harness and attempting
3147 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3148 getting "No route to host". Convert this to a timeout.
3150 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3153 6. Some OS don't have socklen_t but use size_t instead. This affects the
3154 fifth argument of getsockopt() amongst other things. This is now
3155 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3156 can be set for individual OS. I have set it for SunOS5, OSF1, and
3157 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3158 some earlier ones do not.
3160 7. Change 4.30/15 was not doing the test caselessly.
3162 8. The standard form for an IPv6 address literal was being rejected by address
3163 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3164 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3165 this, as well as the form without the "IPv6" on the front (but only when
3166 address literals are enabled, of course).
3168 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3170 10. Exim crashed if a message with an empty sender address specified by -f
3171 encountered a router with an errors_to setting. This could be provoked only
3172 by a command such as
3176 where an empty string was supplied; "<>" did not hit this bug.
3178 11. Installed PCRE release 4.5.
3180 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3181 remained set. It is now erased.
3183 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3184 times from message ids (which are base 36 rather than the normal 62).
3186 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3187 were being counted as actual protocol errors, and logged if the log
3188 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3189 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3190 and DATA following a set of rejected RCPTs do not count as protocol errors.
3191 In other words, Exim assumes they were pipelined, though this may not
3192 actually be the case. Of course, in all cases the client gets an
3193 appropriate error code.
3195 15. If a lookup fails in an ACL condition, a message about the failure may
3196 be available; it is used if testing the ACL cannot continue, because most
3197 such messages specify what the cause of the deferral is. However, some
3198 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3199 that caused an old message to be retained and used if a later statement
3200 caused a defer, replacing the real cause of the deferral.
3202 16. If an IP address had so many PTR records that the DNS lookup buffer
3203 was not large enough to hold them, Exim could crash while trying to process
3204 the truncated data. It now detects and logs this case.
3206 17. Further to 4.21/58, another change has been made: if (and only if) the
3207 first line of a message (the first header line) ends with CRLF, a bare LF
3208 in a subsequent header line has a space inserted after it, so as not to
3209 terminate the header.
3211 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3212 unnecessarily, used atoi() instead of strtol(), and didn't check the
3213 termination when getting file sizes from file names by regex.
3215 19. Completely re-implemented the support for maildirsize files, in the light
3216 of a number of problems with the previous contributed implementation
3217 (4.30/29). In particular:
3219 . If the quota is zero, the maildirsize file is maintained, but no quota is
3222 . If the maildir directory does not exist, it is created before any attempt
3223 to write a maildirsize file.
3225 . The quota value in the file is just a cache; if the quota is changed in
3226 the transport, the new value overrides.
3228 . A regular expression is available for excluding directories from the
3231 20. The autoreply transport checks the characters in options that define the
3232 message's headers; it allows continued headers, but it was checking with
3233 isspace() after an embedded newline instead of explicitly looking for a
3236 21. If all the "regular" hosts to which an address was routed had passed their
3237 expiry times, and had not reached their retry times, the address was
3238 bounced, even if fallback hosts were defined. Now Exim should go on to try
3241 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3242 equivalent code in the SMTP transport. Some hosts send humungous responses
3243 to HELO/EHLO, more than 1024 it seems.
3245 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3246 gives compiler warnings in some environments. I've now done it "properly",
3249 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3250 (because of problems with the built-in one) was declared to return uschar *
3251 instead of char *, causing compiler failure.
3253 25. Fixed a file descriptor leak when processing alias/forward files.
3255 26. Fixed a minor format string issue in dbfn.c.
3257 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3259 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3260 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3263 29. When a sender address is verified, it is cached, to save repeating the test
3264 when there is more than one recipient in a message. However, when the
3265 verification involves a callout, it is possible for different callout
3266 options to be set for different recipients. It is too complicated to keep
3267 track of this in the cache, so now Exim always runs a verification when a
3268 callout is required, relying on the callout cache for the optimization.
3269 The overhead is duplication of the address routing, but this should not be
3272 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3273 to be verified with callout=postmaster, and the main callout worked but the
3274 postmaster check failed, the verification correctly failed. However, if a
3275 subsequent RCPT command asked for sender verification *without* the
3276 postmaster check, incorrect caching caused this verification also to fail,
3279 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3280 it was not caching the DNS options (qualify_single, search_parents) that
3281 were used when the lookup failed. A subsequent lookup with different
3282 options therefore always gave the same answer, though there were cases
3283 where it should not have. (Example: a "domains = !$mx_any" option on a
3284 dnslookup router: the "domains" option is always processed without any
3285 widening, but the router might have qualify_single set.) Now Exim uses the
3286 cached value only when the same options are set.
3288 32. Added John Jetmore's "exipick" utility to the distribution.
3290 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3291 than a timeout (e.g. a certificate is required, and is not provided), an
3292 Exim server now closes the connection immediately. Previously it waited for
3293 the client to close - but if the client is SSL, it seems that they each
3294 wait for each other, leading to a delay before one of them times out.
3296 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3297 maintained 0.8.x compatibility because I don't think many are using it, and
3298 it is clearly obsolete.
3300 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3303 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3304 client certificate was expired. A simple patch fixes this, though I don't
3305 understand the full logic of why the verify callback is called multiple
3308 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3309 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3310 which causes problems with some clients (such as the Certicom SSL Plus
3311 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3312 disables the coutermeasure allowing Eudora to connect."
3314 38. Exim was not checking that a write() to a log file succeeded. This could
3315 lead to Bad Things if a log got too big, in particular if it hit a file
3316 size limit. Exim now panics and dies if it cannot write to a log file, just
3317 as it does if it cannot open a log file.
3319 39. Modified OS/Makefile-Linux so that it now contains
3321 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3323 The two -D definitions ensure that Exim is compiled with large file
3324 support, which makes it possible to handle log files that are bigger than
3327 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3328 instance) a domain was checked against a named list that involved a lookup,
3329 causing $domain_data to be set, then another domain was checked against the
3330 same list, then the first domain was re-checked, the value of $domain_data
3331 after the final check could be wrong. In particular, if the second check
3332 failed, it could be set empty. This bug probably also applied to
3335 41. The strip_trailing_dot option was not being applied to the address given
3336 with the -f command-line option.
3338 42. The code for reading a message's header from the spool was incrementing
3339 $received_count, but never initializing it. This meant that the value was
3340 incorrect (doubled) while delivering a message in the same process in which
3341 it was received. In the most common configuration of Exim, this never
3342 happens - a fresh exec is done - but it can happen when
3343 deliver_drop_privilege is set.
3345 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3346 it now includes up to 150 characters of the unexpected data in the log
3349 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3350 and building data strings. The size of both of these buffers was 10 000
3351 bytes - far larger than anybody would *ever* want, thought I. Needless to
3352 say, somebody hit the limit. I have increased the maximum line length to
3353 20 000 and the maximum data length of concatenated lines to 100 000. I have
3354 also fixed two bugs, because there was no checking on these buffers. Tsk,
3355 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3356 buffer is too small.
3358 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3359 lsearch lookups. Now it does.
3361 46. When parsing a route_list item in a manualroute router, a fixed-length
3362 buffer was used for the list of hosts. I made this 1024 bytes long,
3363 thinking that nobody would ever have a list of hosts that long. Wrong.
3364 Somebody had a whole pile of complicated expansion conditions, and the
3365 string was silently truncated, leading to an expansion error. It turns out
3366 that it is easier to change to an unlimited length (owing to other changes
3367 that have happened since this code was originally written) than to build
3368 structure for giving a limitation error. The length of the item that
3369 expands into the list of hosts is now unlimited.
3371 47. The lsearch lookup could not handle data where the length of text line was
3372 more than 4095 characters. Such lines were truncated, leading to shortened
3373 data being returned. It should now handle lines of any length.
3375 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3376 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3379 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3380 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3381 no longer happens in comment lines. A list of replacements is now placed
3382 at the head of all of the source files, except those whose only change is
3383 to replace PERL_COMMAND in the very first #! line.
3385 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3386 messages on the queue, with a bottom-up merge sort, using code contributed
3387 by Michael Haardt. This should make operations like -bp somewhat faster on
3388 large queues. It won't affect queue runners, except when queue_run_in_order
3391 51. Installed eximstats 1.31 in the distribution.
3393 52. Added support for SRV lookups to the dnslookup router.
3395 53. If an ACL referred to $message_body or $message_body_end, the value was not
3396 reset for any messages that followed in the same SMTP session.
3398 54. The store-handling optimization for building very long strings was not
3399 differentiating between the different store pools. I don't think this
3400 actually made any difference in practice, but I've tidied it.
3402 55. While running the routers to verify a sender address, $sender_address
3403 was still set to the sender address. This is wrong, because when routing to
3404 send a bounce to the sender, it would be empty. Therefore, I have changed
3405 it so that, while verifying a sender address, $sender_address is set to <>.
3406 (There is no change to what happens when verifying a recipient address.)
3408 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3409 target A or AAAA records (if not already returned) without resetting the
3410 qualify_single or search_parents options of the DNS resolver. These are
3411 inappropriate in this case because the targets of MX and SRV records must
3412 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3413 target that, when qualified, matched something in the local domain. These
3414 two options are now turned off when doing these lookups.
3416 57. It seems that at least some releases of Reiserfs (which does not have the
3417 concept of a fixed number of inodes) returns zero and not -1 for the
3418 number of available inodes. This interacted badly with check_spool_inodes,
3419 which assumed that -1 was the "no such thing" setting. What I have done is
3420 to check that the total number of inodes is greater than zero before doing
3421 the test of how many are available.
3423 58. When a "warn" ACL statement has a log_message modifier, the message is
3424 remembered, and not repeated. This is to avoid a lot of repetition when a
3425 message has many recipients that cause the same warning to be written.
3426 Howewer, Exim was preserving the list of already written lines for an
3427 entire SMTP session, which doesn't seem right. The memory is now reset if a
3428 new message is started.
3430 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3431 debugging output unless log_rewrite was also set.
3433 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3434 of (int)(handle) when we know that handle contains (void *)(-1).
3436 61. The Exim daemon panic-logs an error return when it closes the incoming
3437 connection. However "connection reset by peer" seems to be common, and
3438 isn't really an error worthy of noting specially, so that particular error
3441 62. When Exim is trying to find all the local interfaces, it used to panic and
3442 die if the ioctl to get the interface flags failed. However, it seems that
3443 on at least one OS (Solaris 9) it is possible to have an interface that is
3444 included in the list of interfaces, but for which you get a failure error
3445 for this call. This happens when the interface is not "plumbed" into a
3446 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3447 failure of the "get flags" call assumes that the interface is down.
3449 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3450 makes life easier for people who are doing arithmetic on fields extracted
3451 from dates, where you often get leading zeros that should not be
3452 interpreted as octal.
3454 64. Added qualify_domain to the redirect router, to override the global
3457 65. If a pathologically long header line contained very many addresses (the
3458 report of this problem mentioned 10 000) and each of them was rewritten,
3459 Exim could use up a very large amount of memory. (It kept on making new
3460 copies of the header line as it rewrote, and never released the old ones.)
3461 At the expense of a bit more processing, the header rewriting function has
3462 been changed so that it no longer eats memory in this way.
3464 66. The generation of the Received: header has been moved from the time that a
3465 message starts to be received, to the time that it finishes. The timestamp
3466 in the Received: header should now be very close to that of the <= log
3467 line. There are two side-effects of this change:
3469 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3470 logged header lines no longer include the local Received: line, because
3471 it has not yet been created. The same applies to a copy of the message
3472 that is returned to a non-SMTP sender when a message is rejected.
3474 (b) When a filter file is tested using -bf, no additional Received: header
3475 is added to the test message. After some thought, I decided that this
3478 This change does not affect the value of $received_for. It is still set
3479 after address rewriting, but before local_scan() is called.
3481 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3483 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3484 gave an unhelpful panic error message, and a defer error. I have managed to
3485 change this behaviour so that it now rejects any supplied certificate,
3486 which seems right, as the list of acceptable certificates is empty.
3488 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3489 gave an unhelpful defer error. I have not managed to make this reject any
3490 supplied certificates, but the error message it gives is "no certificate
3491 supplied", which is not helpful.
3493 70. exigrep's output now also includes lines that are not associated with any
3494 message, but which match the given pattern. Implemented by a patch from
3495 Martin Sluka, which also tidied up the Perl a bit.
3497 71. Recipient callout verification, like sender verification, was using <> in
3498 the MAIL FROM command. This isn't really the right thing, since the actual
3499 sender may affect whether the remote host accepts the recipient or not. I
3500 have changed it to use the actual sender in the callout; this means that
3501 the cache record is now keyed on a recipient/sender pair, not just the
3502 recipient address. There doesn't seem to be a real danger of callout loops,
3503 since a callout by the remote host to check the sender would use <>.
3504 [SEE ABOVE: changed after hitting problems.]
3506 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3507 temporary errors. However, in the case of such a code being given after
3508 the end of a data transmission (i.e. after ".") Exim was failing to write
3509 a retry record for the message. (Yes, there was some broken host that was
3510 actually sending 8xx at this point.)
3512 73. An unknown lookup type in a host list could cause Exim to panic-die when
3513 the list was checked. (An example that provoked this was putting <; in the
3514 middle of a list instead of at the start.) If this happened during a DATA
3515 ACL check, a -D file could be left lying around. This kind of configuration
3516 error no longer causes Exim to die; instead it causes a defer errror. The
3517 incident is still logged to the main and panic logs.
3519 74. Buglet left over from Exim 3 conversion. The message "too many messages
3520 in one connection" was written to the rejectlog but not the mainlog, except
3521 when address rewriting (yes!) was being logged.
3523 75. Added write_rejectlog option.
3525 76. When a system filter was run not as root (that is, when system_filter_user
3526 was set), the values of the $n variables were not being returned to the
3527 main process; thus, they were not subsequently available in the $sn
3530 77. Added +return_path_on_delivery log selector.
3532 78. A connection timeout was being treated differently from recipients deferred
3533 when testing hosts_max_try with a message that was older than the host's
3534 retry timeout. (The host should not be counted, thus allowing all hosts to
3535 be tried at least once before bouncing.) This may have been the cause of an
3536 occasionally reported bug whereby a message would remain on the queue
3537 longer than the retry timeout, but would be bounced if a delivery was
3538 forced. I say "may" because I never totally pinned down the problem;
3539 setting up timeout/retry tests is difficult. See also the next item.
3541 79. The ultimate address timeout was not being applied to errors that involved
3542 a combination of host plus message (for example, a timeout on a MAIL
3543 command). When an address resolved to a number of possible hosts, and they
3544 were not all tried for each delivery (e.g. because of hosts_max_try), a
3545 message could remain on the queue longer than the retry timeout.
3547 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3550 81. Fixed an obscure SMTP outgoing bug which required at least the following
3551 conditions: (a) there was another message waiting for the same server;
3552 (b) the server returned 5xx to all RCPT commands in the first message so
3553 that the message was not completed; (c) the server dropped the connection
3554 or gave a negative response to the RSET that Exim sends to abort the
3555 transaction. The observed case was a dropped connection after DATA that had
3556 been sent in pipelining mode. That is, the server had advertised PIPELINING
3557 but was not implementing it correctly. The effect of the bug was incorrect
3558 behaviour, such as trying another host, and this could lead to a crash.
3564 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3565 and daemon.c were passed as pointers to ints; they should have been
3566 pointers to socklen_t variables (which are typically unsigned ints).
3568 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3571 3. Fixed a really odd bug that affected only the testing scheme; patching a
3572 certain fixed string in the binary changed the value of another string that
3573 happened to be identical to the end of the original first string.
3575 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3576 name", it returns that address as the IP address. On some operating
3577 systems (e.g. Solaris), it also passes back the IP address string as the
3578 "host name". However, on others (e.g. Linux), it passes back an empty
3579 string. Exim wasn't checking for this, and was changing the host name to an
3580 empty string, assuming it had been canonicized.
3582 5. Although rare, it is permitted to have more than one PTR record for a given
3583 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3584 all the names associated with an address, because they do in Solaris.
3585 However, it seems that they do not in Linux for data that comes from the
3586 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3587 I found this out when I moved to a new Linux workstation and tried to run
3588 the Exim test suite.
3590 To get round this problem I have changed the code so that it now does its
3591 own call to the DNS to look up PTR records when searching for a host name.
3592 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3593 addresses that are only in /etc/hosts are still found.
3595 This behaviour is, however, controlled by an option called host_lookup_
3596 order, which defaults to "bydns:byaddr". If people want to use the other
3597 order, or indeed, just use one or the other means of lookup, they can
3598 specify it in this variable.
3600 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3601 some operating systems, this comes back from gethostbyaddr() as an empty
3602 string, and this is what Exim used to test for. However, it seems that in
3603 other systems, "." is yielded. Exim now tests for this case too.
3605 7. The values of check_spool_space and check_log_space are now held internally
3606 as a number of kilobytes instead of an absolute number of bytes. If a
3607 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3608 kilobyte. This means that much larger values can be stored.
3610 8. Exim monitor: an attempt to get the action menu when not actually pointing
3611 at a message produces an empty menu entitled "No message selected". This
3612 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3613 no entries in it ("Shell widget menu has zero width and/or height"). So I
3614 have added a single, blank menu entry in this case.
3616 9. Added ${quote_local_part.
3618 10. MIME decoding is now applied to the contents of Subject: header lines when
3621 11. Now that a reference to $sender_host_address automatically causes a reverse
3622 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3623 host lookup before query-style lookups in lists that might use this
3624 variable. This has therefore been abolished, and the "net-" prefix is no
3625 longer necessary for query-style lookups.
3627 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3628 have been a typo for LFLAGS, so it has been changed.
3630 13. The install script calls Exim with "-C /dev/null" in order to find the
3631 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3632 to be output. Howeve, since Exim outputs its version number before the
3633 error, it didn't break the script. It just looked ugly. I fixed this by
3634 always allowing "-C /dev/null" if the caller is root.
3636 14. Ignore overlarge ACL variable number when reading spool file - insurance
3637 against a later release with more variables having written the file.
3639 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3640 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3641 this, as well as the form without the "IPv6" on the front.
3643 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3644 OS/Makefile-Darwin file.
3646 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3647 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3649 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3650 variable. However, some OS use ints for inodes. I've added cast to long int
3651 to get rid of the compiler warning.
3653 19. I had forgotten to lock out "/../" in configuration file names when
3654 ALT_CONFIG_PREFIX was set.
3656 20. Routers used for verification do not need to specify transports. However,
3657 if such a router generated a host list, and callout was configured, Exim
3658 crashed, because it could not find a port number from the (non-existent)
3659 transport. It now assumes port 25 in this circumstance.
3661 21. Added the -t option to exigrep.
3663 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3664 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3665 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3666 them is set, LOOKUP_LSEARCH is forced.
3668 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3670 24. Added sender and host information to the "rejected by local_scan()" log
3671 line; previously there was no indication of these.
3673 25. Added .include_if_exists.
3675 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3676 Linux. It turns out that not all file systems support this. Apparently some
3677 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3678 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3679 not supported on the file descriptor, is now ignored when Exim is trying to
3680 sync a directory. This applies only to Linux.
3682 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3684 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3685 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3688 29. Added support for maildirsize files from supplied patch (modified a bit).
3690 30. The use of :fail: followed by an empty string could lead Exim to respond to
3691 sender verification failures with (e.g.):
3693 550 Verification failed for <xxx>
3694 550 Sender verify failed
3696 where the first response line was missing the '-' that indicates it is not
3697 the final line of the response.
3699 31. The loop for finding the name of the user that called Exim had a hardwired
3700 limit of 10; it now uses the value of finduser_retries, which is used for
3701 all other user lookups.
3703 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3706 33. Exim was neglecting to zero errno before one call of strtol() when
3707 expanding a string and expecting an integer value. On some systems this
3708 resulted in spurious "integer overflow" errors. Also, it was casting the
3709 result into an int without checking.
3711 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3712 did not work. The code looks as if it has *never* worked, though it appears
3713 to have been documented since at least releast 1.62. I have made it work.
3715 35. The "timeout_DNS" error in retry rules, also documented since at least
3716 1.62, also never worked. As it isn't clear exactly what this means, and
3717 clearly it isn't a major issue, I have abolished the feature by treating it
3718 as "timeout", and writing a warning to the main and panic logs.
3720 36. The display of retry rules for -brt wasn't always showing the error code
3723 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3724 timeout_connect_A, timeout_connect_MX.
3726 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3727 to the empty sender.
3729 39. The daemon was not analysing the content of -oX till after it had closed
3730 stderr and disconnected from the controlling terminal. This meant that any
3731 syntax errors were only noted on the panic log, and the return code from
3732 the command was 0. By re-arranging the code a little, I've made the
3733 decoding happen first, so such errors now appear on stderr, and the return
3734 code is 1. However, the actual setting up of the sockets still happens in
3735 the disconnected process, so errors there are still only recorded on the
3738 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3739 connections (as happens on some IP stacks) was logged at start up time as
3740 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3741 it from "IPv6 and IPv4", which means that two separate sockets are being
3744 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3745 says whether AF_INET or AF_INET6 was passed as an argument.
3747 42. Exiwhat output was messed up when time zones were included in log
3750 43. Exiwhat now gives more information about the daemon's listening ports,
3751 and whether -tls-on-connect was used.
3753 44. The "port" option of the smtp transport is now expanded.
3755 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3756 silently ignored. Now an error message is written to the main and panic
3759 46. There's a new ACL modifier called "logwrite" which writes to a log file
3760 as soon as it is encountered.
3762 47. Added $local_user_uid and $local_user_gid at routing time.
3764 48. Exim crashed when trying to verify a sender address that was being
3767 49. Exim was recognizing only a space character after ".include". It now also
3768 recognizes a tab character.
3770 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3771 extracting the relevant information from the specification. The man page no
3772 longer contains scrambled data for the -d option, and I've added a section
3773 at the front about calling Exim under different names.
3775 51. Added "extra_headers" argument to the "mail" command in filter files.
3777 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3780 53. Installed eximstats 1.29.
3782 54. Added transport_filter_timeout as a generic transport option.
3784 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3785 Cc: header lines. This was required by RFC 822, but it not required by RFC
3788 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3789 incoming messages that did not have them. Now it does so only if the
3790 message originates locally, that is, if there is no associated remote host
3791 address. When Resent- header lines are present, this applies to the Resent-
3792 lines rather than the non-Resent- lines.
3794 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3795 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3797 58. Messages for configuration errors now include the name of the main
3798 configuration file - useful now that there may be more than one file in a
3799 list (.included file names were always shown).
3801 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3802 for those rare installations that do not start the daemon as root or run it
3803 setuid root. I've cut out the call to initgroups() if the daemon is not
3806 60. The Exim user and group can now be bound into the binary as text strings
3807 that are looked up at the start of Exim's processing.
3809 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3811 62. Added $mailstore_basename variable.
3813 63. Installed patch to sieve.c from Michael Haardt.
3815 64. When Exim failed to open the panic log after failing to open the main log,
3816 the original message it was trying to log was written to stderr and debug
3817 output, but if they were not available (the usual case in production), it
3818 was lost. Now it is written to syslog before the two lines that record the
3819 failures to open the logs.
3821 65. Users' Exim filters run in subprocesses under the user's uid. It is
3822 possible for a "deliver" command or an alias in a "personal" command to
3823 provoke an address rewrite. If logging of address rewriting is configured,
3824 this fails because the process is not running as root or exim. There may be
3825 a better way of dealing with this, but for the moment (because 4.30 needs
3826 to be released), I have disabled address rewrite logging when running a
3827 filter in a non-root, non-exim process.
3833 1. The buildconfig auxiliary program wasn't quoting the value set for
3834 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3835 not defined. This bug was masked in 4.22 by the effect that was fixed in
3838 2. Some messages that were rejected after a message id was allocated were
3839 shown as "incomplete" by exigrep. It no longer does this for messages that
3840 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3842 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3843 have allow_domain_literals set, the ID did not get logged in the <= line.
3844 Domain literals are now always recognized in Message-ID: header lines.
3846 4. The first argument for a ${extract expansion item is the key name or field
3847 number. Leading and trailing spaces in this item were not being ignored,
3848 causing some misleading effects.
3850 5. When deliver_drop_privilege was set, single queue runner processes started
3851 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3852 same command in the process it spins off) were not dropping privilege.
3854 6. When the daemon running as "exim" started a queue runner, it always
3855 re-executed Exim in the spun-off process. This is a waste of effort when
3856 deliver_drop_privilege is set. The new process now just calls the
3857 queue-runner function directly.
3863 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3866 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3867 ignored. Though the use of -oP was forcing the writing of a pid file, it
3868 was always written to the default place.
3870 3. If the message "no IP address found for host xxxx" is generated during
3871 incoming verification, it is now followed by identification of the incoming
3872 connection (so you can more easily find what provoked it).
3874 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3876 5. Added some features to "harden" Exim a bit more against certain attacks:
3878 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3879 be put in Local/Makefile. This is like the never_users runtime option,
3880 but it cannot be overridden. The default setting is "root".
3882 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3883 prefix string with which any file named in a -C command line option
3886 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3887 is retained for -C and -D only if the caller of Exim is root. Without
3888 it, the exim user may also use -C and -D and retain privilege.
3890 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3891 command line option is disabled.
3893 6. Macro names set by the -D option must start with an upper case letter, just
3894 like macro names defined in the configuration file.
3896 7. Added "dereference=" facility to LDAP.
3898 8. Two instances of the typo "uknown" in the source files are fixed.
3900 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3901 the Configure-Makefile script screwed up while processing it.
3903 10. Incorporated PCRE 4.4.
3905 11. The SMTP synchronization check was not operating right at the start of an
3906 SMTP session. For example, it could not catch a HELO sent before the client
3907 waited for the greeting. There is now a check for outstanding input at the
3908 point when the greeting is written. Because of the duplex, asynchronous
3909 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3910 way, but not yet received, when the check is performed.
3912 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3913 on TCP/IP sockets, because this apparently causes some broken clients to
3916 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3917 unchanged) from the Cygwin maintainer.
3919 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3920 was supported instead of testing for mbx. Effectively a typo.
3922 15. The spa authenticator server code was not checking that the input it
3923 received was valid base64.
3925 16. The debug output line for the "set" modifier in ACLs was not showing the
3926 name of the variable that was being set.
3928 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3930 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3931 lookup on its own. Something else was needed to trigger the lookup. For
3932 example, a match in host_lookup or the need for a host name in a host list.
3933 Now, if $sender_host_name is referenced and the host name has not yet been
3934 looked up, a lookup is performed. If the lookup fails, the variable remains
3935 empty, and $host_lookup_failed is set to "1".
3937 19. Added "eqi" as a case-independent comparison operator.
3939 20. The saslauthd authentication condition could segfault if neither service
3940 nor realm was specified.
3942 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3943 error message that was logged was misleading, and incoming SMTP
3944 connections were dropped. The message is now more accurate, and temporary
3945 errors are given to SMTP connections.
3947 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3948 (except RSET) with a 421 error, until QUIT is received. However, it was
3949 failing to send a response to QUIT.
3951 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3952 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3953 ACL is now run after the helo_try_verify_hosts code.
3955 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3956 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3957 case-independent, so other case variants are also recognized). Apparently
3958 some systems use these upper case variants.
3960 25. If more than two messages were waiting for the same host, and a transport
3961 filter was specified for the transport, Exim sent two messages over the
3962 same TCP/IP connection, and then failed with "socket operation on non-
3963 socket" when it tried to send the third.
3965 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3967 27. The extern definition of crypt16() in expand.c was not being excluded when
3968 the OS had its own crypt16() function.
3970 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3971 as a preferred synonym for return_size_limit, both as an option and as an
3974 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3976 30. Changed the default configuration ACL to relax the local part checking rule
3977 for addresses that are not in any local domains. For these addresses,
3978 slashes and pipe symbols are allowed within local parts, but the sequence
3979 /../ is explicitly forbidden.
3981 31. SPA server authentication was not clearing the challenge buffer before
3984 32. log_message in a "warn" ACL statement was writing to the reject log as
3985 well as to the main log, which contradicts the documentation and doesn't
3986 seem right (because no rejection is happening). So I have stopped it.
3988 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3989 However, I am unable to do any testing of this.
3991 34. Fixed an infelicity in the appendfile transport. When checking directories
3992 for a mailbox, to see if any needed to be created, it was accidentally
3993 using path names with one or more superfluous leading slashes; tracing
3994 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3996 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3997 contents are added to the log line that is written for every discarded
3998 recipient. (Previously a log_message setting was ignored.)
4000 36. The ${quote: operator now quotes the string if it is empty.
4002 37. The install script runs exim in order to find its version number. If for
4003 some reason other than non-existence or emptiness, which it checks, it
4004 could not run './exim', it was installing it with an empty version number,
4005 i.e. as "exim-". This error state is now caught, and the installation is
4008 38. An argument was missing from the function that creates an error message
4009 when Exim fails to connect to the socket for saslauthd authentication.
4010 This could cause Exim to crash, or give a corrupted message.
4012 39. Added isip, isip4, and isip6 to ${if conditions.
4014 40. The ACL variables $acl_xx are now saved with the message, and can be
4015 accessed later in routers, transports, and filters.
4017 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4018 strings in the file are not string-expanded.
4020 42. If a MAIL command specified a SIZE value that was too large to fit into an
4021 int variable, the check against message_size_limit failed. Such values are
4022 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4023 day this will have to be increased, but I don't think I want to be around
4024 when emails are that large.
4031 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4032 iconv() is not standard in FreeBSD.
4034 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4035 IPv6 enabled. The observed symptom was a segmentation fault on return from
4036 the function os_common_find_running_interfaces() in src/os.c.
4038 3. In the check_special_case() function in daemon.c I had used "errno" as an
4039 argument name, which causes warnings on some systems. This was basically a
4040 typo, since it was named "eno" in the comments!
4042 4. The code that waits for the clock to tick (at a resolution of some fraction
4043 of a second) so as to ensure message-id uniqueness was always waiting for
4044 at least one whole tick, when it could have waited for less. [This is
4045 almost certainly not relevant at current processor speeds, where it is
4046 unlikely to ever wait at all. But we try to future-proof.]
4048 5. The function that sleeps for a time interval that includes fractions of a
4049 second contained a race. It did not block SIGALRM between setting the
4050 timer, and suspending (a couple of lines later). If the interval was short
4051 and the sigsuspend() was delayed until after it had expired, the suspension
4052 never ended. On busy systems this could lead to processes getting stuck for
4055 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4056 process, before it forks any delivery processes. The open lookup caching
4057 mechanism meant that the open file or database connection was passed into
4058 the delivery process. The problem was that delivery processes always tidy
4059 up cached lookup data. This could cause a problem for the next delivery
4060 process started by the queue runner, because the external queue runner
4061 process does not know about the closure. So the next delivery process
4062 still has data in the lookup cache. In the case of a file lookup, there was
4063 no problem because closing a file descriptor in a subprocess doesn't affect
4064 the parent. However, if the lookup was caching a connection to a database,
4065 the connection was closed, and the second delivery process was likely to
4066 see errors such as "PGSQL: query failed: server closed the connection
4067 unexpectedly". The problem has been fixed by closing all cached lookups
4068 in a queue runner before running a delivery process.
4070 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4071 seem to have the "const" qualifier which it has on other OS. I've
4074 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4075 *of the same type* (client or server) with the same public name that an
4076 error should be diagnosed.
4078 9. When Exim looked up a host name for an IP address, but failed to find the
4079 original IP address when looking up the host name (a safety check), it
4080 output the message "<ip address> does not match any IP for NULL", which was
4081 confusing, to say the least. The bug was that the host name should have
4082 appeared instead of "NULL".
4084 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4085 user that is built into the binary, and the -C or -D options is used, root
4086 privilege is dropped before the configuration file is read. In addition,
4087 logging is switched to stderr instead of the normal log files. If the
4088 configuration then re-defines the Exim user, the unprivileged environment
4089 is probably not what is expected, so Exim logs a panic warning message (but
4092 However, if deliver_drop_privilege is set, the unprivileged state may well
4093 be exactly what is intended, so the warning has been cut out in that case,
4094 and Exim is allowed to try to write to its normal log files.
4100 1. smtp_return_error_details was not giving details for temporary sender
4101 or receiver verification errors.
4103 2. Diagnose a configuration error if two authenticators have the same public
4106 3. Exim used not to create the message log file for a message until the first
4107 delivery attempt. This could be confusing when incoming messages were held
4108 for policy or load reasons. The message log file is now created at the time
4109 the message is received, and an initial "Received" line is written to it.
4111 4. The automatically generated man page for command line options had a minor
4112 bug that caused no ill effects; however, a more serious problem was that
4113 the procedure for building the man page automatically didn't always
4114 operate. Consequently, release 4.20 contains an out-of-date version. This
4115 shouldn't happen again.
4117 5. When building Exim with embedded Perl support, the script that builds the
4118 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4119 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4121 6. The freeze_tell option was not being used for messages that were frozen on
4122 arrival, either by an ACL or by local_scan().
4124 7. Added the smtp_incomplete_transaction log selector.
4126 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4127 was accepting AUTH without a new EHLO.
4129 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4130 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4133 10. Logging of TCP/IP connections (when configured) now happens in the main
4134 daemon process instead of the child process, so that the TCP/IP connection
4135 count is more accurate (but it can never be perfect).
4137 11. The use of "drop" in a nested ACL was not being handled correctly in the
4138 outer ACL. Now, if condition failure induced by the nested "drop" causes
4139 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4140 or "require"), the connection is dropped.
4142 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4143 that yield "discard" can now be used with an "accept" or a "discard" verb,
4144 but an error is generated for any others (because I can't see a useful way
4145 to define what should happen).
4147 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4148 are now processed in the same way as lines in the Exim configuration file.
4149 In particular, continuation lines are supported.
4151 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4153 15. Added -ti meaning -t -i.
4155 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4156 domains, and warn by logging if others are found.
4158 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4159 returned from GIFCONF when Exim is trying to find the list of interfaces on
4160 a host. The code in os.c has been modified to copy each ifreq to an aligned
4161 structure in all cases.
4163 Also, in some cases, the returned ifreq's were being copied to a 'struct
4164 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4165 means the last couple of bytes of an IPv6 address could be chopped if the
4166 ifreq contained only a normal sockaddr (14 bytes storage).
4168 18. Named domain lists were not supported in the hosts_treat_as_local option.
4169 An entry such as +xxxx was not recognized, and was treated as a literal
4172 19. Ensure that header lines added by a DATA ACL are included in the reject log
4173 if the ACL subsequently rejects the message.
4175 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4176 MD5 (which is deprecated).
4178 21. When testing a filter file using -bf, Exim was writing a message when it
4179 took the sender from a "From " line in the message, but it was not doing so
4180 when it took $return_path from a Return-Path: header line. It now does.
4182 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4183 with a valid header line field name (a series of printing characters
4184 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4186 23. Changed "disc" in the source to "disk" to conform to the documentation and
4187 the book and for uniformity.
4189 24. Ignore Sendmail's -Ooption=value command line item.
4191 25. When execve() failed while trying to run a command in a pipe transport,
4192 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4193 could be confused with a return value of 69 from the command itself. This
4194 has been changed to 127, the value the shell returns if it is asked to run
4195 a non-existent command. The wording for the related log line suggests a
4196 non-existent command as the problem.
4198 26. If received_header_text expands to an empty string, do not add a Received:
4199 header line to the message. (Well, it adds a token one on the spool, but
4200 marks it "old" so that it doesn't get used or transmitted.)
4202 27. Installed eximstats 1.28 (addition of -nt option).
4204 28. There was no check for failure on the call to getsockname() in the daemon
4205 code. This can fail if there is a shortage of resources on the system, with
4206 ENOMEM, for example. A temporary error is now given on failure.
4208 29. Contrary to the C standard, it seems that in some environments, the
4209 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4210 program. Exim now does this explicitly; it affects the formatting of
4211 timestamps using strftime().
4213 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4214 complaints. I've now initialized all the variables, to avoid this.
4216 32. Header lines added by a system filter were not being "seen" during
4217 transport-time rewrites.
4219 33. The info_callback() function passed to OpenSSL is set up with type void
4220 (*)(SSL *, int, int), as described somewhere. However, when calling the
4221 function (actually a macro) that sets it up, the type void(*)() is
4222 expected. I've put in a cast to prevent warnings from picky compilers.
4224 34. If a DNS black list lookup found a CNAME record, but there were no A
4225 records associated with the domain it pointed at, Exim crashed.
4227 35. If a DNS black list lookup returned more than one A record, Exim ignored
4228 all but the first. It now scans all returned addresses if a particular IP
4229 value is being sought. In this situation, the contents of the
4230 $dnslist_value variable are a list of all the addresses, separated by a
4233 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4234 to accept a host name and all its aliases if the forward lookup for any of
4235 them yielded the IP address of the incoming connection. Now it accepts only
4236 those names whose forward lookup yields the correct IP address. Any other
4237 names are discarded. This closes a loophole whereby a rogue DNS
4238 administrator could create reverse DNS records to break through a
4239 wildcarded host restriction in an ACL.
4241 37. If a user filter or a system filter that ran in a subprocess used any of
4242 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4243 the wrong values were passed to the pipe command ($thisaddress had the
4244 value of $0, $0 had the value of $1, etc). This bug was introduced by
4245 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4247 38. Improved the line breaking for long SMTP error messages from ACLs.
4248 Previously, if there was no break point between 40 and 75 characters, Exim
4249 left the rest of the message alone. Two changes have been made: (a) I've
4250 reduced the minimum length to 35 characters; (b) if it can't find a break
4251 point between 35 and 75 characters, it looks ahead and uses the first one
4252 that it finds. This may give the occasional overlong line, but at least the
4253 remaining text gets split now.
4255 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4256 file descriptors might be low, and that setting 1000 would always raise it.
4257 It turns out that in some environments, the limit is already over 1000 and
4258 that lowering it causes trouble. So now Exim takes care not to decrease it.
4260 40. When delivering a message, the value of $return_path is set to $sender_
4261 address at the start of routing (routers may change the value). By an
4262 oversight, this default was not being set up when an address was tested by
4263 -bt or -bv, which affected the outcome if any router or filter referred to
4266 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4267 log only when "message" or "log_message" are set. However, if one of the
4268 conditions was an address verification, or a call to a nested ACL, the
4269 messages generated by the underlying test were being passed through. This
4270 no longer happens. The underlying message is available in $acl_verify_
4271 message for both "message" and "log_message" expansions, so it can be
4272 passed through if needed.
4274 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4275 new expansion $bh_ to give the encoded byte string without charset
4276 translation. Translation happens only if iconv() is available; HAVE_ICONV
4277 indicates this at build time. HEADERS_CHARSET gives the charset to
4278 translate to; headers_charset can change it in the configuration, and
4279 "headers charset" can change it in an individual filter file.
4281 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4282 that creates RFC 2047 encoded "words" labels them as that charset instead
4283 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4284 expansion operator; (ii) when Exim creates a From: line for a local
4285 message; (iii) when a header line is rewritten to include a "phrase" part.
4287 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4288 buggy, causing it to skip the first lines of messages whose message ID
4289 ended in 'D'. This would not have bitten before Exim release 4.14, because
4290 message IDs were unlikely to end in 'D' before then. The effect was to have
4291 incorrect size information for certain domains.
4293 45. #include "config.h" was missing at the start of the crypt16.c module. This
4294 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4297 46. If there was a timeout during a "random" callout check, Exim treated it as
4298 a failure of the random address, and carried on sending RSET and the real
4299 address. If the delay was just some slowness somewhere, the response to the
4300 original RCPT would be taken as a response to RSET and so on, causing
4301 mayhem of various kinds.
4303 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4304 when I implemented it. It didn't allow for the fact that some option values
4305 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4306 the right test for positive values.
4308 48. Domain names in DNS records are case-independent. Exim always looks them up
4309 in lower case. Some resolvers return domain names in exactly the case they
4310 appear in the zone file, that is, they may contain uppercase letters. Not
4311 all resolvers do this - some return always lower case. Exim was treating a
4312 change of case by a resolver as a change of domain, similar to a widening
4313 of a domain abbreviation. This triggered its re-routing code and so it was
4314 trying to route what was effectively the same domain again. This normally
4315 caused routing to fail (because the router wouldn't handle the domain
4316 twice). Now Exim checks for this case specially, and just changes the
4317 casing of the domain that it ultimately uses when it transmits the message
4320 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4323 50. If a filter generated a file delivery with a non-absolute name (possible if
4324 no home directory exists for the router), the forbid_file option was not
4327 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4328 the existing equality matching.
4330 52. Exim was using ints instead of ino_t variables in some places where it was
4331 dealing with inode numbers.
4333 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4334 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4335 called TMPDIR, and if it finds it is different, it changes its value.
4337 54. The smtp_printf() function is now made available to local_scan() so
4338 additional output lines can be written before returning. There is also an
4339 smtp_fflush() function to enable the detection of a dropped connection.
4340 The variables smtp_input and smtp_batched_input are exported to
4343 55. Changed the default runtime configuration: the message "Unknown user"
4344 has been removed from the ACL, and instead placed on the localuser router,
4345 using the cannot_route_message feature. This means that any verification
4346 failures that generate their own messages won't get overridden. Similarly,
4347 the "Unrouteable address" message that was in the ACL for unverifiable
4348 relay addresses has also been removed.
4350 56. Added hosts_avoid_esmtp to the smtp transport.
4352 57. The exicyclog script was not checking for the esoteric option
4353 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4354 will work only if exicyclog is run under the appropriate euid.
4356 58. Following a discussion on the list, the rules by which Exim recognises line
4357 endings on incoming messages have been changed. The -dropcr and drop_cr
4358 options are now no-ops, retained only for backwards compatibility. The
4359 following line terminators are recognized: LF CRLF CR. However, special
4360 processing applies to CR:
4362 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4363 nor a local message in the state where . is a terminator.
4365 (ii) If a bare CR is encountered in a header line, an extra space is added
4366 after the line terminator so as not to end the header. The reasoning
4367 behind this is that bare CRs in header lines are most likely either
4368 to be mistakes, or people trying to play silly games.
4370 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4371 was being incorrectly given as 18 bytes larger than it should have been.
4372 This is a VOB (very old bug).
4374 60. This may never have affected anything current, but just in case it has:
4375 When the local host is found other than at the start of a list of hosts,
4376 the local host, those with the same MX, and any that follow, are discarded.
4377 When the list in question was part of a longer list of hosts, the following
4378 hosts (not currently being processed) were also being discarded. This no
4379 longer happens. I'm not sure if this situation could ever has previously
4382 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4385 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4386 header. This is something that is recommended in a new Internet Draft, and
4387 is something that is documented as being done by Sendmail. There are two
4388 possible values. For messages generated by the autoreply transport, Exim
4391 Auto-Submitted: auto-replied
4393 whereas for all other generated messages (e.g. bounces) it adds
4395 Auto-Submitted: auto-generated
4397 63. The "personal" condition in filters now includes a test for the
4398 Auto-Submitted: header. If it contains the string "auto-" the message it
4399 not considered personal.
4401 64. Added rcpt_include_affixes as a generic transport option.
4403 65. Added queue_only_override (default true).
4405 66. Added the syslog_duplication option.
4407 67. If what should have been the first header line of a message consisted of
4408 a space followed by a colon, Exim was mis-interpreting it as a header line.
4409 It isn't of course - it is syntactically invalid and should therefore be
4410 treated as the start of the message body. The misbehaviour could have
4411 caused a number of strange effects, including loss of data in subsequent
4412 header lines, and spool format errors.
4414 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4415 client host had authenticated. This control can now be exercised by an ACL
4416 for more flexibility.
4418 69. By default, callouts do not happen when testing with -bh. There is now a
4419 variant, -bhc, which does actually run the callout code, including
4420 consulting and updating the callout cache.
4422 70. Added support for saslauthd authentication, courtesy of Alexander
4425 71. If statvfs() failed on the spool or log directories while checking their
4426 size for availability, Exim confusingly gave the error "space shortage".
4427 Furthermore, in debugging mode it crashed with a floating point exception.
4428 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4429 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4430 serious problem, Exim now writes to the main and panic logs when this
4431 happens, with details of the failure. It then refuses to accept the
4432 incoming message, giving the message "spool directory problem" or "log
4433 directory problem" with a 421 code for SMTP messages.
4435 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4436 0, 1, and 2 exist, because some OS complain for execs without them (see
4437 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4438 descriptors. However, the code omitted to check that the open succeeded,
4439 causing mysterious errors if for some reason the permissions on /dev/null
4440 got screwed. Now Exim writes a message to the main and panic logs, and
4441 bombs out if it can't open /dev/null.
4443 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4444 interact so that it is all more flexible. It is supposed to remain
4445 backwards compatible. Also added extra_local_interfaces.
4447 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4448 to bomb out with an assertion failure - to the client this appears as a
4449 connection drop. This problem occurs in the part of the code that was taken
4450 from the Samba project. Fortunately, the assertion is in a very simple
4451 function, so I have fixed this by reproducing the function inline in the
4452 one place where it is called, and arranging for authentication to fail
4453 instead of killing the process with assert().
4455 75. The SPA client code was not working when the server requested OEM rather
4456 than Unicode encoding.
4458 76. Added code to make require_files with a specific uid setting more usable in
4459 the case where statting the file as root fails - usually a non-root-mounted
4460 NFS file system. When this happens and the failure is EACCES, Exim now
4461 forks a subprocess and does the per-uid checking as the relevant uid.
4463 77. Added process_log_path.
4465 78. If log_file_path was not explicitly set, a setting of check_log_space or
4466 check_log_inodes was ignored.
4468 79. If a space check for the spool or log partitions fails, the incident is now
4469 logged. Of course, in the latter case the data may get lost...
4471 80. Added the %p formatting code to string_format() so that it can be used to
4472 print addresses in debug_print(). Adjusted all the address printing in the
4473 debugging in store.c to use %p rather than %d.
4475 81. There was a concern that a line of code in smtp_in.c could overflow a
4476 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4477 initially expressed, the concern was not well-founded, because trailing
4478 spaces are removed early. However, if the trailing spaces were followed by
4479 a NULL, they did not get removed, so the overflow was possible. Two fixes
4482 (a) I re-wrote the offending code in a cleaner fashion.
4483 (b) If an incoming SMTP command contains a NULL character, it is rejected
4486 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4487 runs initgroups(), so that if the Exim user is in any additional groups,
4488 they will be used during message reception.
4494 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob