1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.451 2007/01/08 11:56:41 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
10 fixed by 4.65/MH/01 (is this a record?) are fixed:
12 (i) An empty string was always treated as zero by the numeric comparison
13 operators. This behaviour has been restored.
15 (ii) It is documented that the numeric comparison operators always treat
16 their arguments as decimal numbers. This was broken in that numbers
17 starting with 0 were being interpreted as octal.
19 While fixing these problems I realized that there was another issue that
20 hadn't been noticed. Values of message_size_limit (both the global option
21 and the transport option) were treated as octal if they started with 0.
22 The documentation was vague. These values are now always treated as
23 decimal, and I will make that clear in the documentation.
29 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
30 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
33 MH/01 Don't check that the operands of numeric comparison operators are
34 integers when their expansion is in "skipping" mode (fixes bug
35 introduced by 4.64-PH/07).
37 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
38 child addresses, Exim now panics and dies. Previously, because the count
39 is held in a short int, deliveries were likely to be lost. As such a
40 large number of recipients for a single message is ridiculous
41 (performance will be very, very poor), I have chosen to impose a limit
42 rather than extend the field.
48 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
49 leftover -K file (the existence of which was triggered by #402).
50 While we were at it, introduced process PID as part of the -K
51 filename. This should rule out race conditions when creating
54 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
55 processing considerably. Previous code took too long for large mails,
56 triggering a timeout which in turn triggers #401.
58 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
59 in the DK code in transports.c. sendfile() is not really portable,
60 hence the _LINUX specificness.
62 TF/01 In the add_headers option to the mail command in an Exim filter,
63 there was a bug that Exim would claim a syntax error in any
64 header after the first one which had an odd number of characters
67 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
68 callout verification, Exim cached a "reject" for the entire domain. This
69 is correct for most verifications, but it is not correct for a recipient
70 verification with use_sender or use_postmaster set, because in that case
71 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
72 case of MAIL FROM:<> rejection from other early rejections (e.g.
73 rejection of HELO). When verifying a recipient using a non-null MAIL
74 address, the cache is ignored if it shows MAIL FROM:<> rejection.
75 Whatever the result of the callout, the value of the domain cache is
76 left unchanged (for any other kind of callout, getting as far as trying
77 RCPT means that the domain itself is ok).
79 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
82 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
83 manifest itself as EPIPE rather than ECONNECT. When tidying away a
84 session, the daemon ignores ECONNECT errors and logs others; it now
85 ignores EPIPE as well.
87 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
88 (quoted-printable decoding).
90 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
91 later the small subsequent patch to fix an introduced bug.
93 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
95 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
97 PH/08 An error is now given if message_size_limit is specified negative.
99 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
100 to be given (somewhat) arbitrary names.
102 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
105 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
106 miscellaneous code fixes
108 PH/10 Added the log_reject_target ACL modifier to specify where to log
111 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
112 hostname. This is wrong, because it relates to the incoming message (and
113 probably the interface on which it is arriving) and not to the outgoing
114 callout (which could be using a different interface). This has been
115 changed to use the value of the helo_data option from the smtp transport
116 instead - this is what is used when a message is actually being sent. If
117 there is no remote transport (possible with a router that sets up host
118 addresses), $smtp_active_hostname is used.
120 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
121 tweaks were necessary in order to get it to work (see also 21 below):
122 (a) The code assumed that strncpy() returns a negative number on buffer
123 overflow, which isn't the case. Replaced with Exim's string_format()
125 (b) There were several signed/unsigned issues. I just did the minimum
126 hacking in of casts. There is scope for a larger refactoring.
127 (c) The code used strcasecmp() which is not a standard C function.
128 Replaced with Exim's strcmpic() function.
129 (d) The code set only $1; it now sets $auth1 as well.
130 (e) A simple test gave the error "authentication client didn't specify
131 service in request". It would seem that Dovecot has changed its
132 interface. Fortunately there's a specification; I followed it and
133 changed what the client sends and it appears to be working now.
135 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
138 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
139 address A is aliased to B and C, where B exists and C does not. Without
140 -v the output is "A verified" because verification stops after a
141 successful redirection if more than one address is generated. However,
142 with -v the child addresses are also verified. Exim was outputting "A
143 failed to verify" and then showing the successful verification for C,
144 with its parentage. It now outputs "B failed to verify", showing B's
145 parentage before showing the successful verification of C.
147 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
148 look up a TXT record in a specific list after matching in a combined
151 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
152 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
153 they consult the DNS. I had assumed they would set it the way they
154 wanted; and indeed my experiments on Linux seem to show that in some
155 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
156 To be on the safe side, however, I have now made the interface to
157 host_find_byname() similar to host_find_bydns(), with an argument
158 containing the DNS resolver options. The host_find_byname() function now
159 sets these options at its start, just as host_find_bydns() does. The smtp
160 transport options dns_qualify_single and dns_search_parents are passed to
161 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
162 of host_find_byname() use the default settings of RES_DEFNAMES
163 (qualify_single) but not RES_DNSRCH (search_parents).
165 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
166 spool_read_header() do less string testing, by means of a preliminary
167 switch on the second character of optional "-foo" lines. (This is
168 overdue, caused by the large number of possibilities that now exist.
169 Originally there were few.) While I was there, I also converted the
170 str(n)cmp tests so they don't re-test the leading "-" and the first
171 character, in the hope this might squeeze out yet more improvement.
173 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
174 flag allowing group syntax was set by the header_syntax check but not
175 turned off, possible causing trouble later; (2) The flag was not being
176 set at all for the header_verify test, causing "group"-style headers to
177 be rejected. I have now set it in this case, and also caused header_
178 verify to ignore an empty address taken from a group. While doing this, I
179 came across some other cases where the code for allowing group syntax
180 while scanning a header line wasn't quite right (mostly, not resetting
181 the flag correctly in the right place). These bugs could have caused
182 trouble for malformed header lines. I hope it is now all correct.
184 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
185 with the "reply" argument non-NULL. The code, however (which originally
186 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
187 but it didn't always do it. This confused somebody who was copying the
188 code for some other use. I have removed all the tests.
190 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
191 feature that was used to support insecure browsers during the U.S. crypto
192 embargo. It requires special client support, and Exim is probably the
193 only MTA that supported it -- and would never use it because real RSA is
194 always available. This code has been removed, because it had the bad
195 effect of slowing Exim down by computing (never used) parameters for the
196 RSA_EXPORT functionality.
198 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
199 authenticator to fail if there's a tab character in the incoming data
200 (there should never be unless someone is messing about, as it's supposed
201 to be base64-encoded). Also added, on Timo's advice, the "secured" option
202 if the connection is using TLS or if the remote IP is the same as the
203 local IP, and the "valid-client-cert option" if a client certificate has
206 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
207 authenticators. This can be used for authorization after authentication
208 succeeds. (In the case of plaintext, it servers for both authentication
211 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
212 if any retry times were supplied.
214 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
215 connection, where there is no client IP address to check. In this
216 situation, the verify now always succeeds.
218 PH/25 Applied John Jetmore's -Mset patch.
220 PH/26 Added -bem to be like -Mset, but loading a message from a file.
222 PH/27 In a string expansion for a processed (not raw) header when multiple
223 headers of the same name were present, leading whitespace was being
224 removed from all of them, but trailing whitespace was being removed only
225 from the last one. Now trailing whitespace is removed from each header
226 before concatenation. Completely empty headers in a concatenation (as
229 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
230 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
232 PH/29 [Removed. This was a change that I later backed out, and forgot to
233 correct the ChangeLog entry (that I had efficiently created) before
234 committing the later change.]
236 PH/30 Exim was sometimes attempting to deliver messages that had suffered
237 address errors (4xx response to RCPT) over the same connection as other
238 messages routed to the same hosts. Such deliveries are always "forced",
239 so retry times are not inspected. This resulted in far too many retries
240 for the affected addresses. The effect occurred only when there were more
241 hosts than the hosts_max_try setting in the smtp transport when it had
242 the 4xx errors. Those hosts that it had tried were not added to the list
243 of hosts for which the message was waiting, so if all were tried, there
244 was no problem. Two fixes have been applied:
246 (i) If there are any address or message errors in an SMTP delivery, none
247 of the hosts (tried or untried) are now added to the list of hosts
248 for which the message is waiting, so the message should not be a
249 candidate for sending over the same connection that was used for a
250 successful delivery of some other message. This seems entirely
251 reasonable: after all the message is NOT "waiting for some host".
252 This is so "obvious" that I'm not sure why it wasn't done
253 previously. Hope I haven't missed anything, but it can't do any
254 harm, as the worst effect is to miss an optimization.
256 (ii) If, despite (i), such a delivery is accidentally attempted, the
257 routing retry time is respected, so at least it doesn't keep
258 hammering the server.
260 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
261 in ${readsocket because some servers need this prod.
263 PH/32 Added some extra debug output when updating a wait-xxx database.
265 PH/33 The hint "could be header name not terminated by colon", which has been
266 given for certain expansion errors for a long time, was not being given
267 for the ${if def:h_colon_omitted{... case.
269 PH/34 The spec says: "With one important exception, whenever a domain list is
270 being scanned, $domain contains the subject domain." There was at least
271 one case where this was not true.
273 PH/35 The error "getsockname() failed: connection reset by peer" was being
274 written to the panic log as well as the main log, but it isn't really
275 panic-worthy as it just means the connection died rather early on. I have
276 removed the panic log writing for the ECONNRESET error when getsockname()
279 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
280 runs only) independently of the message's sender address. This meant
281 that, if the 4xx error was in fact related to the sender, a different
282 message to the same recipient with a different sender could confuse
283 things. In particualar, this can happen when sending to a greylisting
284 server, but other circumstances could also provoke similar problems.
285 I have changed the default so that the retry time for these errors is now
286 based a combination of the sender and recipient addresses. This change
287 can be overridden by setting address_retry_include_sender=false in the
290 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
291 remote server are returned as part of bounce messages. This was not
292 happening for LMTP over a pipe (the lmtp transport), but now it is the
293 same for both kinds of LMTP.
295 PH/38 Despite being documented as not happening, Exim was rewriting addresses
296 in header lines that were in fact CNAMEs. This is no longer the case.
298 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
299 and queue runs started by the daemon processed all messages. This has
300 been fixed so that -R and -S can now usefully be given with -q<time>.
302 PH/40 Import PCRE release 6.7 (fixes some bugs).
304 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
306 PH/42 Give an error if -q is specified more than once.
308 PH/43 Renamed the variables $interface_address and $interface_port as
309 $received_ip_address and $received_port, to make it clear that these
310 values apply to message reception, and not to the outgoing interface when
311 a message is delivered. (The old names remain recognized, of course.)
313 PH/44 There was no timeout on the connect() call when using a Unix domain
314 socket in the ${readsocket expansion. There now is.
316 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
317 be meaningful with "accept".
319 SC/01 Eximstats V1.43
320 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
322 SC/02 Eximstats V1.44
323 Use a glob alias rather than an array ref in the generated
324 parser. This improves both readability and performance.
326 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
327 Collect SpamAssassin and rejection statistics.
328 Don't display local sender or destination tables unless
329 there is data to show.
330 Added average volumes into the top table text output.
332 SC/04 Eximstats V1.46
333 Collect data on the number of addresses (recipients)
334 as well as the number of messages.
336 SC/05 Eximstats V1.47
337 Added 'Message too big' to the list of mail rejection
338 reasons (thanks to Marco Gaiarin).
340 SC/06 Eximstats V1.48
341 Mainlog lines which have GMT offsets and are too short to
342 have a flag are now skipped.
344 SC/07 Eximstats V1.49 (Alain Williams)
345 Added the -emptyok flag.
347 SC/08 Eximstats V1.50
348 Fixes for obtaining the IP address from reject messages.
350 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
351 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
352 whitesspace changes from 4.64-PH/27
354 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
357 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
358 are found, allow negative numbers in numeric criteria)
360 JJ/06 exipick.20061117.2, added new $message_body_missing variable
362 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
363 to match changes made in 4.64-PH/43
365 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
367 PH/47 Put in an explicit test for a DNS lookup of an address record where the
368 "domain" is actually an IP address, and force a failure. This locks out
369 those revolvers/nameservers that support "A-for-A" lookups, in
370 contravention of the specifications.
372 PH/48 When a host name was looked up from an IP address, and the subsequent
373 forward lookup of the name timed out, the host name was left in
374 $sender_host_name, contrary to the specification.
376 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
377 restricted to single-key lookups, Exim was not diagnosing an error if
378 * or *@ was used with a query-style lookup.
380 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
382 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
383 long ago, but noone interested enough thought of it. Let's just say that
384 the "1.1" means that there are some new functions that weren't there at
385 some point in the past.
387 PH/51 Error processing for expansion failure of helo_data from an smtp
388 transport during callout processing was broken.
390 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
391 tested/used via the -bh/-bhc/-bs options.
393 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
394 bug, fixed in subsequent PCRE releases).
396 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
397 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
399 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
405 SC/01 Use a glob alias rather than an array ref in eximstats generated
406 parser. This improves both readability and performance.
408 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
409 Don't display local sender or destination tables in eximstats unless
410 there is data to show.
411 Added average volumes into the eximstats top table text output.
413 SC/03 Collect data on the number of addresses (recipients) as well
414 as the number of messages in eximstats.
416 TF/01 Correct an error in the documentation for the redirect router. Exim
417 does (usually) call initgroups() when daemonizing.
419 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
420 with consistent privilege compared to when running as a daemon.
422 TF/03 Note in the spec that $authenticated_id is not set for local
423 submissions from trusted users.
425 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
426 Thanks to Dean Brooks <dean@iglou.com> for the patch.
428 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
429 by adding some example configuration directives to the default
430 configuration file. A little bit of work is required to uncomment the
431 directives and define how usernames and passwords are checked, but
432 there is now a framework to start from.
434 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
435 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
436 without this. I don't know how relevant this is to other LDAP libraries.
438 PH/02 Add the verb name to the "unknown ACL verb" error.
440 PH/03 Magnus Holmgren's patch for filter_prepend_home.
442 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
444 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
445 directory not expanded when it should be if an expanded home directory
446 was set for the address (which is overridden by the transport).
448 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
451 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
452 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
453 because it is too late at that time, and has no effect.
455 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
456 security issue with \' (bugzilla #107). I could not use the
457 PQescapeStringConn() function, because it needs a PGconn value as one of
460 PH/08 When testing addresses using -bt, indicate those final addresses that
461 are duplicates that would not cause an additional delivery. At least one
462 person was confused, thinking that -bt output corresponded to deliveries.
463 (Suppressing duplicates isn't a good idea as you lose the information
464 about possibly different redirections that led to the duplicates.)
466 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
467 systems where poll() doesn't work, in particular OS X.
469 PH/10 Added more information to debugging output for retry time not reached.
471 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
472 operations in malware.c.
474 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
477 PH/13 If write_rejectlog was set false when logging was sent to syslog with
478 syslog_duplication set false, log lines that would normally be written
479 both the the main log and to the reject log were not written to syslog at
482 PH/14 In the default configuration, change the use of "message" in ACL warn
483 statements to "add_header".
485 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
486 not followed by a command (e.g. "seen endif").
488 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
489 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
492 PH/17 Added extra conditions to the default value of delay_warning_condition
496 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
497 { match{$h_precedence:}{(?i)bulk|list|junk} } \
498 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
501 The Auto-Submitted: and various List- headers are standardised, whereas I
502 don't think Precedence: ever was.
504 PH/18 Refactored debugging code in route_finduser() to show more information,
505 in particular, the error code if getpwnam() issues one.
507 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
508 This is apparently needed in addition to the PH/07 change above to avoid
509 any possible encoding problems.
511 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
512 but not after initializing Perl.
514 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
515 output them only if debugging. By default they are written stderr,
516 apparently, which is not desirable.
518 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
521 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
524 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
526 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
527 authenticated or an ident call has been made. Suppress the default
528 values for $authenticated_id and $authenticated_sender (but permit -oMai
529 and -oMas) when testing with -bh.
531 PH/24 Re-jigged the order of the tests in the default configuration so that the
532 tests for valid domains and recipients precede the DNS black list and CSA
533 tests, on the grounds that those ones are more expensive.
535 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
536 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
537 This bug exists in every version of Exim that I still have, right back to
540 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
541 However, an attempt to turn on case-sensitivity in a regex key by
542 including (?-i) didn't work because the subject string was already
543 lowercased, and the effects were non-intuitive. It turns out that a
544 one-line patch can be used to allow (?-i) to work as expected.
550 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
551 other effects) broke the use of negated acl sub-conditions.
553 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
556 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
557 "Deny" causes Exim to reject the incoming connection with a 554 error.
558 Unfortunately, if there is a major crisis, such as a disk failure,
559 tcp-wrappers gives "deny", whereas what one would like would be some
560 kind of temporary error. A kludge has been added to help with this.
561 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
562 554 error is used if errno is still zero or contains ENOENT (which occurs
563 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
566 PH/03 Add -lutil to the default FreeBSD LIBS setting.
568 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
569 errors. Otherwise a message that provokes a temporary error (when other
570 messages do not) can cause a whole host to time out.
572 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
573 addresses were routed directly to files or pipes from a redirect router.
574 File deliveries just didn't batch; pipe deliveries might have suffered
577 PH/06 A failure to get a lock for a hints database would erroneously always say
578 "Failed to get write lock", even when it was really a read lock.
580 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
581 of 0600. This has been changed to use the value of the lockfile_mode
582 option (which defaults to 0600).
584 PH/08 Applied small patch from the Sieve maintainer.
586 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
587 folder from quota calculations, a direct delivery into this folder messed
588 up the contents of the maildirsize file. This was because the regex was
589 used only to exclude .Trash (or whatever) when the size of the mailbox
590 was calculated. There was no check that a delivery was happening into an
591 excluded directory. This bug has been fixed by ignoring all quota
592 processing for deliveries into excluded directories.
594 PH/10 Added the maildirfolder_create_regex option to appendfile.
600 PH/01 The code for finding all the local interface addresses on a FreeBSD
601 system running IPv6 was broken. This may well have applied to all BSD
602 systems, as well as to others that have similar system calls. The broken
603 code found IPv4 interfaces correctly, but gave incorrect values for the
604 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
605 that it would not match correctly against @[] and not recognize the IPv6
608 PH/02 The ipliteral router was not recognizing addresses of the form user@
609 [ipv6:....] because it didn't know about the "ipv6:" prefix.
611 PH/03 Added disable_ipv6.
613 PH/04 Changed $reply_address to use the raw form of the headers instead of the
614 decoded form, because it is most often used to construct To: headers
615 lines in autoreplies, and the decoded form may well be syntactically
616 invalid. However, $reply_address has leading white space removed, and all
617 newlines turned into spaces so that the autoreply transport does not
620 PH/05 If group was specified without a user on a router, and no group or user
621 was specified on a transport, the group from the router was ignored.
623 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
624 for visible compile-time settings that can be used to change these
625 numbers, for those that want even more. Backwards compatibility with old
626 spool files has been maintained. However, going back to a previous Exim
627 release will lost any variables that are in spool files.
629 PH/07 Two small changes when running in the test harness: increase delay when
630 passing a TCP/IP connection to a new process, in case the original
631 process has to generate a bounce, and remove special handling of
632 127.0.0.2 (sic), which is no longer necessary.
634 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
635 be the same on different OS.
637 PH/09 Moved a debug statement in filter processing to avoid a race problem when
640 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
641 whether --show-vars was specified or not
643 JJ/02 exipick: Added support for new ACL variable spool format introduced
646 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
647 syntactically invalid From: or Reply-to: line, and a filter used this to
648 generate an autoreply, and therefore failed to obtain an address for the
649 autoreply, Exim could try to deliver to a non-existent relative file
650 name, causing unrelated and misleading errors. What now happens is that
651 it logs this as a hard delivery error, but does not attempt to create a
654 PH/11 The exinext utility has a -C option for testing purposes, but although
655 the given file was scanned by exinext itself; it wasn't being passed on
656 when Exim was called.
658 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
659 an end-of-file indication when reading a command response.
661 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
662 compiled. In many other places in Exim, IPv6 addresses are always
663 recognized, so I have changed this. It also means that IPv4 domain
664 literals of the form [IPV4:n.n.n.n] are now always recognized.
666 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
667 used if the router is not running as root, for example, when verifying at
668 ACL time, or when using -bh. The debugging output from this situation was
669 non-existent - all you got was a failure to exec. I have made two
672 (a) Failures to set uid/gid, the current directory, or a process leader
673 in a subprocess such as that created by queryprogram now generate
674 suitable debugging ouput when -d is set.
676 (b) The queryprogram router detects when it is not running as root,
677 outputs suitable debugging information if -d is set, and then runs
678 the subprocess without attempting to change uid/gid.
680 PH/15 Minor change to Makefile for building test_host (undocumented testing
683 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
684 additional section of a DNS packet that returns MX or SRV records.
685 Instead, it always explicitly searches for A/AAAA records. This avoids
686 major problems that occur when a DNS server includes only records of one
687 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
688 fixed another bug: if SRV records were looked up and the corresponding
689 address records were *not* found in the additional section, the port
690 values from the SRV records were lost.
692 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
693 using the correct key (the original address) when searching the retry
694 rules in order to find which one to use for generating the retry hint.
696 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
697 adding the default one. Similarly, if it contains a Reply-To: header, the
698 errors_reply_to option, if set, is not used.
700 PH/19 When calculating a retry time, Exim used to measure the "time since
701 failure" by looking at the "first failed" field in the retry record. Now
702 it does not use this if it is later than than the arrival time of the
703 message. Instead it uses the arrival time. This makes for better
704 behaviour in cases where some deliveries succeed, thus re-setting the
705 "first failed" field. An example is a quota failure for a huge message
706 when small messages continue to be delivered. Without this change, the
707 "time since failure" will always be short, possible causing more frequent
708 delivery attempts for the huge message than are intended.
709 [Note: This change was subsequently modified - see PH/04 for 4.62.]
711 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
712 $1, $2, $3) because the numerical variables can be reset during some
713 expansion items (e.g. "match"), thereby losing the authentication data.
715 PH/21 Make -bV show the size of off_t variables so that the test suite can
716 decide whether to run tests for quotas > 2G.
718 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
719 mailbox_size, and mailbox_filecount in the appendfile transport. If a
720 filecount value is greater than 2G or if a quota value is greater than 2G
721 on a system where the size of off_t is not greater than 4, a panic error
724 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
725 never match. The debug and -bh output now contains an explicit error
726 message indicating a malformed IPv4 address or mask.
728 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
729 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
732 PH/25 Do not write to syslog when running in the test harness. The only
733 occasion when this arises is a failure to open the main or panic logs
734 (for which there is an explicit test).
736 PH/26 Added the /no_tell option to "control=freeze".
738 PH/27 If a host name lookup failed very early in a connection, for example, if
739 the IP address matched host_lookup and the reverse lookup yielded a name
740 that did not have a forward lookup, an error message of the form "no IP
741 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
742 could be logged. Now it outputs the IP address instead of "NULL".
744 PH/28 An enabling patch from MH: add new function child_open_exim2() which
745 allows the sender and the authenticated sender to be set when
746 submitting a message from within Exim. Since child_open_exim() is
747 documented for local_scan(), the new function should be too.
749 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
750 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
751 results in an empty string is now treated as unset.
753 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
755 PH/31 Added sender_verify_fail logging option.
757 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
758 needed by RFC 822 but not by RFC 2822 was commented out. I have now
759 tidied the source and removed it altogether.
761 PH/33 When a queue run was abandoned because the load average was too high, a
762 log line was always written; now it is written only if the queue_run log
763 selector is set. In addition, the log line for abandonment now contains
764 information about the queue run such as the pid. This is always present
765 in "start" and "stop" lines but was omitted from the "abandon" line.
767 PH/34 Omit spaces between a header name and the colon in the error message that
768 is given when verify = headers_syntax fails (if there are lots of them,
769 the message gets confusing).
771 PH/35 Change the default for dns_check_names_pattern to allow slashes within
772 names, as there are now some PTR records that contain slashes. This check
773 is only to protect against broken name servers that fall over on strange
774 characters, so the fact that it applies to all lookups doesn't matter.
776 PH/36 Now that the new test suite is complete, we can remove some of the
777 special code in Exim that was needed for the old test suite. For example,
778 sorting DNS records because real resolvers return them in an arbitrary
779 order. The new test suite's fake resolver always returns records in the
782 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
783 bounces) except when queue_only is set, to avoid logging races between
784 the different processes.
786 PH/38 Panic-die if .include specifies a non-absolute path.
788 PH/39 A tweak to the "H" retry rule from its user.
790 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
791 a label. They prevented compilation on older perls.
793 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
794 a warning to be raised on newish perls.
796 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
797 on queue. Changes to match documented behaviour of showing count of
798 messages matching specified criteria.
800 PH/40 Changed the default ident timeout from 30s to 5s.
802 PH/41 Added support for the use of login_cap features, on those BSD systems
803 that have them, for controlling the resources used by pipe deliveries.
805 PH/42 The content-scanning code uses fopen() to create files in which to put
806 message data. Previously it was not paying any attention to the mode of
807 the files. Exim runs with umask(0) because the rest of the code creates
808 files with open(), and sets the required mode explicitly. Thus, these
809 files were ending up world-writeable. This was not a big issue, because,
810 being within the spool directory, they were not world-accessible. I have
811 created a function called modefopen, which takes an additional mode
812 argument. It sets umask(777), creates the file, chmods it to the required
813 mode, then resets the umask. All the relevant calls to fopen() in the
814 content scanning code have been changed to use this function.
816 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
817 to 24 hours. This avoids potential overflow problems when processing G
818 and H retry rules. I suspect nobody ever tinkers with this value.
820 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
822 PH/45 When the plaintext authenticator is running as a client, the server's
823 challenges are checked to ensure they are valid base64 strings. By
824 default, the authentication attempt is cancelled if an invalid string is
825 received. Setting client_ignore_invalid_base64 true ignores these errors.
826 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
827 they are received. Thus, the responses can be made to depend on the
828 challenges. If an invalid string is ignored, an empty string is placed in
831 PH/46 Messages that are created by the autoreply transport now contains a
832 References: header, in accordance with RFCs 2822 and 3834.
834 PH/47 Added authenticated_sender_force to the smtp transport.
836 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
838 PH/49 Installed latest patch from the Sieve maintainer.
840 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
841 was also set, the appendfile transport was unnecessarily scanning a
842 directory of message files (e.g. for maildir delivery) to find the count
843 of files (along with the size), even though it did not need this
844 information. It now does the scan only if it needs to find either the
845 size of the count of files.
847 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
849 PH/52 Two bugs concerned with error handling when the smtp transport is
852 (i) Exim was not creating retry information for temporary errors given
853 for individual recipients after the DATA command when the smtp transport
854 was used in LMTP mode. This meant that they could be retried too
855 frequently, and not timed out correctly.
857 (ii) Exim was setting the flag that allows error details to be returned
858 for LMTP errors on RCPT commands, but not for LMTP errors for individual
859 recipients that were returned after the DATA command.
861 PH/53 This is related to PH/52, but is more general: for any failing address,
862 when detailed error information was permitted to be returned to the
863 sender, but the error was temporary, then after the final timeout, only
864 "retry timeout exceeded" was returned. Now it returns the full error as
865 well as "retry timeout exceeded".
867 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
868 do this, and (what is worse) MTAs that accept it.
870 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
871 will now be deprecated.
873 PH/56 New os.c-cygwin from the Cygwin maintainer.
875 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
876 formats (previously only available in exim formats via -bpr, -bpru,
877 and -bpra. Now also available in native and exiqgrep formats)
879 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
880 with very large, slow to parse queues
882 JJ/08 exipick: added ! as generic prefix to negate any criteria format
884 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
886 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
887 responses to authentication challenges, though it was showing the
888 challenges; (ii) I've removed the CR characters from the debug output for
891 PH/58 Allow for the insertion of a newline as well as a space when a string
892 is turned into more than one encoded-word during RFC 2047 encoding. The
893 Sieve code now uses this.
895 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
896 data_4xx, lost_connection, tls_required.
898 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
899 message was being sent as an SMTP response.
901 PH/61 Add -l and -k options to exicyclog.
903 PH/62 When verifying, if an address was redirected to one new address, so that
904 verification continued, and the new address failed or deferred after
905 having set something in $address_data, the value of $address_data was not
906 passed back to the ACL. This was different to the case when no
907 redirection occurred. The value is now passed back in both cases.
909 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
910 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
911 use login_cap.h, so on its own it isn't the distinguishing feature. The
912 new name refers directly to the setclassresources() function.
914 PH/65 Added configuration files for NetBSD3.
916 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
918 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
919 is preferred over IPv4.
921 PH/68 The bounce_return_message and bounce_return_body options were not being
922 honoured for bounces generated during the reception of non-SMTP messages.
923 In particular, this applied to messages rejected by the ACL. This bug has
924 been fixed. However, if bounce_return_message is true and bounce_return_
925 body is false, the headers that are returned for a non-SMTP message
926 include only those that have been read before the error was detected.
927 (In the case of an ACL rejection, they have all been read.)
929 PH/69 The HTML version of the specification is now built in a directory called
930 spec_html instead of spec.html, because the latter looks like a path with
931 a MIME-type, and this confuses some software.
933 PH/70 Catch two compiler warnings in sieve.c.
935 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
936 function verify_get_ident() calls ip_connect() to connect a socket, but
937 if the "connect()" function timed out, ip_connect() used to close the
938 socket. However, verify_get_ident() also closes the socket later, and in
939 between Exim writes to the log, which may get opened at this point. When
940 the socket was closed in ip_connect(), the log could get the same file
941 descriptor number as the socket. This naturally causes chaos. The fix is
942 not to close the socket in ip_connect(); the socket should be closed by
943 the function that creates it. There was only one place in the code where
944 this was missing, in the iplookup router, which I don't think anybody now
945 uses, but I've fixed it anyway.
947 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
948 well as to direct DNS lookups. Otherwise the handling of names in host
949 lists is inconsistent and therefore confusing.
955 PH/01 Two changes to the default runtime configuration:
957 (1) Move the checks for relay_from_hosts and authenticated clients from
958 after to before the (commented out) DNS black list checks.
960 (2) Add control=submission to the relay_from_hosts and authenticated
961 clients checks, on the grounds that messages accepted by these
962 statements are most likely to be submissions.
964 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
966 (1) Generate an error if the third argument for the ${prvs expansion is
969 (2) Treat a missing third argument of ${prvscheck as if it were an empty
972 (3) Reset the variables that are obtained from the first argument of
973 ${prvscheck and used in the second argument before leaving the code,
974 because their memory is reclaimed, so using them afterwards may do
977 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
978 one (it's much easier than Tom thought :-).
980 (5) Because of (4), we can now allow for the use of $prvscheck_result
981 inside the third argument.
983 PH/03 For some reason, the default setting of PATH when running a command from
984 a pipe transport was just "/usr/bin". I have changed it to
987 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
988 anything to be listed in the output from -bV.
990 PH/05 When a filter generated an autoreply, the entire To: header line was
991 quoted in the delivery log line, like this:
993 => >A.N.Other <ano@some.domain> <original@ddress> ...
995 This has been changed so that it extracts the operative address. There
996 may be more than one such address. If so, they are comma-separated, like
999 => >ano@some.domain,ona@other.domain <original@ddress> ...
1001 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1002 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1003 not being looked up in the rDNS to get a host name, Exim was showing the
1004 IP address twice in Received: lines, even though the IP addresses were
1005 identical. For example:
1007 Received: from [1.2.3.4] (helo=[1.2.3.4])
1009 However, if the real host name was known, it was omitting the HELO data
1010 if it matched the actual IP address. This has been tidied up so that it
1011 doesn't show the same IP address twice.
1013 PH/07 When both +timestamp and +memory debugging was on, the value given by
1014 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1015 called by the debug printing, thereby overwriting the timestamp buffer.
1016 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1018 PH/08 When the original message was included in an autoreply transport, it
1019 always said "this is a copy of the message, including all the headers",
1020 even if body_only or headers_only was set. It now gives an appropriate
1023 PH/09 Applied a patch from the Sieve maintainer which:
1025 o fixes some comments
1026 o adds the (disabled) notify extension core
1027 o adds some debug output for the result of if/elsif tests
1028 o points to the current vacation draft in the documentation
1029 and documents the missing references header update
1033 o fixes a bug in processing the envelope test (when testing
1034 multiple envelope elements, the last element determinted the
1037 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1038 Electronic Mail") by including:
1040 Auto-submitted: auto-generated
1042 in the messages that it generates (bounce messages and others, such as
1043 warnings). In the case of bounce messages for non-SMTP mesages, there was
1044 also a typo: it was using "Auto_submitted" (underscore instead of
1045 hyphen). Since every message generated by Exim is necessarily in response
1046 to another message, thes have all been changed to:
1048 Auto-Submitted: auto-replied
1050 in accordance with these statements in the RFC:
1052 The auto-replied keyword:
1054 - SHOULD be used on messages sent in direct response to another
1055 message by an automatic process,
1057 - MUST NOT be used on manually-generated messages,
1059 - MAY be used on Delivery Status Notifications (DSNs) and Message
1060 Disposition Notifications (MDNs),
1062 - MUST NOT be used on messages generated by automatic or periodic
1063 processes, except for messages which are automatic responses to
1066 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1067 to the default Received: header definition.
1069 PH/12 Added log selector acl_warn_skipped (default on).
1071 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1072 variables because (a) they are in the wrong storage pool and (b) even if
1073 they were copied, it wouldn't work properly because of the caching.
1075 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1076 checking when decoding. Apparently there are clients that generate
1077 overlong encoded strings. Why am I not surprised?
1079 PH/15 If the first argument of "${if match_address" was not empty, but did not
1080 contain an "@" character, Exim crashed. Now it writes a panic log message
1081 and treats the condition as false.
1083 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1085 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1086 extension "envelope-auth". The code is finished and in agreement with
1087 other implementations, but there is no documentation so far and in fact,
1088 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1089 not changing the active code.
1091 Print executed "if" and "elsif" statements when debugging is used. This
1092 helps a great deal to understand what a filter does.
1094 Document more things not specified clearly in RFC3028. I had all this
1095 sorted out, when out of a sudden new issues came to my mind. Oops."
1097 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1100 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1101 canonical form (as documented). However, after a host name lookup from
1102 the IP address, check_host() was doing a simple string comparison with
1103 addresses acquired from the DNS when checking that the found name did
1104 have the original IP as one of its addresses. Since any found IPv6
1105 addresses are likely to be in abbreviated form, the comparison could
1106 fail. Luckily, there already exists a function for doing the comparison
1107 by converting both addresses to binary, so now that is used instead of
1108 the text comparison.
1110 PH/20 There was another similar case to PH/19, when a complete host name was
1111 given in a host list; looking up its IP address could give an abbreviated
1112 form, whereas the current host's name might or might not be abbreviated.
1113 The same fix has been applied.
1119 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1120 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1123 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1124 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1126 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1128 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1129 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1130 Also, for "Auto-Submitted", treat anything other than "no" as
1131 non-personal, in accordance with RFC 3834. (Previously it treated
1132 anything starting "auto-" as non-personal.)
1134 TF/01 The control=submission/name=... option had a problem with syntax
1135 errors if the name included a slash character. The /name= option
1136 now slurps the rest of the string, so it can include any characters
1137 but it must come last in the list of options (after /sender_retain
1140 PH/05 Some modifications to the interface to the fake nameserver for the new
1148 TK/01 Added the "success_on_redirect" address verification option. See
1149 NewStuff for rationale and an example.
1151 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1153 PH/02 Patch to exigrep to allow it to work on syslog lines.
1155 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1156 fread() to skip over the body file's header line, because in Cygwin the
1157 header line is locked and is inaccessible.
1159 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1160 co-exist for some time) to make it clear that it is the Exim ID that is
1161 referenced, not the Message-ID: header line.
1163 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1164 string_format() function, because snprintf() does not exist on all
1167 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1168 ${stat: expansion item.
1170 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1171 protocol synchronization error", to keep the pedants happy.
1173 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1174 well as for IRIX systems, when gcc is being used. See the host.c source
1177 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1179 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1182 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1183 transport and to the smtp transport in LMTP mode.
1185 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1187 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1188 run a filter in a subprocess. This could lead to confusion in subsequent
1189 lookups in the parent process. There should also be a search_tidyup() at
1190 the end of the subprocess.
1192 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1193 only if the host matched helo_try_verify_hosts, which caused the
1194 verification to occur when the EHLO/HELO command was issued. The ACL just
1195 tested the remembered result. Now, if a previous verification attempt has
1196 not happened, "verify = helo" does it there and then.
1198 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1200 TK/03 Fix log output including CR from clamd.
1202 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1203 exist provoked a memory error which could cause a segfault.
1205 PH/15 Installed PCRE 6.2
1207 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1209 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1210 of the problem. Specifically, suggested +O2 rather than +O1 for the
1213 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1215 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1216 record, the port was not correctly logged when the outgoing_port log
1217 selector was set (it logged the transort's default port).
1219 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1220 fallback_hosts, and "hosts" in the smtp transport.
1222 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1223 host errors such as "Connection refused".
1225 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1226 authentication with radiusclient 0.4.9:
1228 - Error returned from rc_read_config was caught wrongly
1229 - Username/password not passed on to radius server due to wrong length.
1231 The presumption is that some radiusclient API changes for 4.51/PH/17
1232 were not taken care of correctly. The code is still untested by me (my
1233 Linux distribution still has 0.3.2 of radiusclient), but it was
1234 contributed by a Radius user.
1236 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1237 expanding the "port" option of the smtp transport.
1239 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1240 while reading a MIME header. Thanks to Tom Hughes for a patch.
1242 PH/24 Include config.h inside local_scan.h so that configuration settings are
1245 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1246 that in an ACL for RCPT (for example), you can examine exactly what was
1249 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1250 commands, but it was not correctly comparing the address with the actual
1251 client host address. Thus, it would show the EHLO address in Received:
1252 header lines when this was not necessary.
1254 PH/27 Added the % operator to ${eval:}.
1256 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1257 it should be ignoring failures (because with -C, for example, it has lost
1258 privilege). It wasn't ignoring creation failures other than "already
1261 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1264 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1265 because an input line was too long, either on its own, or by virtue of
1266 too many continuations, the temporary file was not being removed, and the
1267 return code was incorrect.
1269 PH/31 Missing "BOOL" in function definition in filtertest.c.
1271 PH/32 Applied Sieve patches from the maintainer.
1273 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1275 PH/33 Added "verify = not_blind".
1277 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1278 Local/Makefile (with some defaults set). These are used in built scripts
1279 such as exicyclog, but they have never been used in the exim_install
1280 script (though there are many overriding facilities there). I have
1281 arranged that the exim_install script now takes note of these two
1284 PH/35 Installed configuration files for Dragonfly.
1286 PH/36 When a locally submitted message by a trusted user did not contain a
1287 From: header, and the sender address was obtained from -f or from an SMTP
1288 MAIL command, and the trusted user did not use -F to supply a sender
1289 name, $originator_name was incorrectly used when constructing a From:
1290 header. Furthermore, $originator_name was used for submission mode
1291 messages from external hosts without From: headers in a similar way,
1292 which is clearly wrong.
1294 PH/37 Added control=suppress_local_fixups.
1296 PH/38 When log_selector = +received_sender was set, and the addition of the
1297 sender made the log line's construction buffer exactly full, or one byte
1298 less than full, an overflow happened when the terminating "\n" was
1301 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1302 when the result of a list match is failure because a DNS lookup failed.
1304 PH/40 RM_COMMAND is now used in the building process.
1306 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1307 the "build-* directories that it finds.
1309 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1310 address in a domain literal was a prefix of an interface address.
1312 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1313 when verifying a sender address, unless rewrite_headers is false.
1315 PH/44 Wrote a long comment about why errors_to addresses are verified as
1316 recipients, not senders.
1318 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1319 the ratelimit ACL was added.
1321 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1323 PH/46 Added extra information about PostgreSQL errors to the error string.
1325 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1326 avoiding the need to install special zones in a real server. This is
1327 backwards compatible; if it can't find the fake resolver, it drops back.
1328 Thus, both old and new test suites can be run.
1330 TF/02 Added util/ratelimit.pl
1332 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1333 clock is set back in time.
1335 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1336 Candler <B.Candler@pobox.com>.
1338 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1339 OK for addresses that are the result of redirections.
1341 PH/48 A number of further additions for the benefit of the new test suite,
1342 including a fake gethostbyname() that interfaces to the fake DNS resolver
1345 TF/06 The fix for widen_domains has also been applied to qualify_single and
1346 search_parents which are the other dnslookup options that can cause
1349 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1352 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1354 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1355 service" if it encountered an MX record with an empty target hostname.
1356 The message is now "an MX or SRV record indicated no SMTP service".
1358 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1359 if the DNS of the sending site is misconfigured. This is quite a
1360 common situation. This change restores the behaviour of treating a
1361 helo verification defer as a failure.
1363 PH/51 If self=fail was set on a router, the bounce message did not include the
1364 actual error message.
1370 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1372 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1373 command itself ended in error, the underlying message about the transport
1374 filter timeout was being overwritten with the pipe command error. Now the
1375 underlying error message should be appended to the second error message.
1377 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1379 PH/02 Reduce the amount of output that "make" produces by default. Full output
1380 can still be requested.
1382 PH/03 The warning log line about a condition test deferring for a "warn" verb
1383 was being output only once per connection, rather than after each
1384 occurrence (because it was using the same function as for successful
1385 "warn" verbs). This seems wrong, so I have changed it.
1387 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1388 it should not have, which might have caused a crash in the right
1389 circumstances, but probably never did.
1391 PH/04 Installed a modified version of Tony Finch's patch to make submission
1392 mode fix the return path as well as the Sender: header line, and to
1393 add a /name= option so that you can make the user's friendly name appear
1396 TF/03 Added the control = fakedefer ACL modifier.
1398 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1399 Mark Lowes for thorough testing.
1401 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1403 TK/03 Merged latest SRS patch from Miles Wilton.
1405 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1406 with the definition in sysexits.h (which is #included earlier).
1407 Fortunately, Exim does not actually use EX_OK. The code used to try to
1408 preserve the sysexits.h value, by assumimg that macro definitions were
1409 scanned for macro replacements. I have been disabused of this notion,
1410 so now the code just undefines EX_OK before #including unistd.h.
1412 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1413 in the smtp transport. When a block could not be written in a single
1414 write() function, the timeout was being re-applied to each part-write.
1415 This seems wrong - if the receiver was accepting one byte at a time it
1416 would take for ever. The timeout is now adjusted when this happens. It
1417 doesn't have to be particularly precise.
1419 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1420 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1422 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1423 without a domain if the check to <postmaster@domain> fails.
1425 SC/01 Eximstats: added -xls and the ability to specify output files
1426 (patch written by Frank Heydlauf).
1428 SC/02 Eximstats: use FileHandles for outputing results.
1430 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1432 SC/04 Eximstats: fixed display of large numbers with -nvr option
1434 SC/05 Eximstats: fixed merging of reports with empty tables.
1436 SC/06 Eximstats: added the -include_original_destination flag
1438 SC/07 Eximstats: removed tabs and trailing whitespace.
1440 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1442 TK/06 MBOX spool code: Add real "From " MBOX separator line
1443 so the .eml file is really in mbox format (even though
1444 most programs do not really care). Patch from Alex Miller.
1446 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1447 The latter is generated from $received_to and is only set if the
1448 message has one envelope recipient. SA can use these headers,
1449 obviously out-of-the-box. Patch from Alex Miller.
1451 PH/08 The ${def test on a variable was returning false if the variable's
1452 value was "0", contrary to what the specification has always said!
1453 The result should be true unless the variable is empty.
1455 PH/09 The syntax error of a character other than { following "${if
1456 def:variable_name" (after optional whitespace) was not being diagnosed.
1457 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1458 accidental colon was present, for example, could give incorrect results.
1460 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1461 result is used (not including appendfile, where other changes are about
1464 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1465 This involved changing a lot of size variables from int to off_t. It
1466 should work with maildirs and everything.
1468 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1469 spamd dying while we are connected to it.
1471 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1474 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1475 function for BDB 4.3.
1477 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1479 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1480 the sources for PCRE have been re-arranged and more files are now
1483 PH/15 The code I had for printing potentially long long variables in PH/11
1484 above was not the best (it lost precision). The length of off_t variables
1485 is now inspected at build time, and an appropriate printing format (%ld
1486 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1487 to be "long long int" or "long int". This is needed for the internal
1488 formatting function string_vformat().
1490 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1491 the configuration file to be ":syslog", then the script "guesses" where
1492 the logs files are, rather than using the compiled in default. In our
1493 case the guess is not the same as the compiled default, so the script
1494 suddenly stopped working when I started to use syslog. The patch checks
1495 to see if log_file_path is "". If so, it attempts to read it from exim
1496 with no configuration file to get the compiled in version, before it
1497 falls back to the previous guessing code."
1499 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1500 implementing BATV in an Exim configuration. See NewStuff for the gory
1503 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1504 Makefile that are specific to HP-UX.
1506 PH/18 If the "use_postmaster" option was set for a recipient callout together
1507 with the "random" option, the postmaster address was used as the MAIL
1508 FROM address for the random test, but not for the subsequent recipient
1509 test. It is now used for both.
1511 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1512 patch removes a few documentation additions to RFC 3028, because the
1513 latest draft now contains them. It adds the new en;ascii-case comparator
1514 and a new error check for 8bit text in MIME parts. Comparator and
1515 require names are now matched exactly. I enabled the subaddress
1516 extension, but it is not well tested yet (read: it works for me)."
1518 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1519 rework some of the code of TK/09 above to avoid the hardwired use of
1520 "%lld" and "long long". Replaced the call to snprintf() with a call to
1523 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1524 records point to non-existent hosts", "retry timeout exceeded", and
1525 "retry time not reached for any host after a long failure period".
1527 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1528 experimental DomainKeys support:
1530 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1531 (2) On an error such as an illegally used "control", the wrong name for
1532 the control was given.
1534 These problems did NOT occur unless DomainKeys support was compiled.
1536 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1538 PH/24 Added ${if match_ip condition.
1540 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1541 databases so that it will be absolutely obvious if a crash occurs in the
1542 DB library. This is a regular occurrence (often caused by mis-matched
1545 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1546 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1547 picked up on a user's system that detects such things. There doesn't seem
1548 to be a gcc warning option for this - only an attribute that has to be
1549 put on the function's prototype. It seems that in Fedora Core 4 they have
1550 set this on a number of new functions. No doubt there will be more in due
1553 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1554 specify a transport. However, if an address that was verified by such a
1555 router was the subject of a callout, Exim crashed because it tried to
1556 read the rcpt_include_affixes from the non-existent transport. Now it
1557 just assumes that the setting of that option is false. This bug was
1558 introduced by 4.51/PH/31.
1560 PH/28 Changed -d+all to exclude +memory, because that information is very
1561 rarely of interest, but it makes the output a lot bigger. People tend to
1562 do -d+all out of habit.
1564 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1565 code in os-type was giving problems when libc.so lives in lib64, like on
1568 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1569 aren't the modern standard, and it seems that some systems' include files
1570 don't always have them. Exim was already checking for some of the newer
1571 ones like T_AAAA, and defining it itself. I've added checks for all the
1572 record types that Exim uses.
1574 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1575 not automatically generating a new one, as it is supposed to. This
1576 prevented TLS from working. If the file did exist, but contained invalid
1577 data, a new version was generated, as expected. It was only the case of a
1578 non-existent file that was broken.
1580 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1581 with a change in libdomainkeys > 0.64.
1583 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1584 from DNS. If the selector record carries the flag, it now has
1585 precedence over the domain-wide flag.
1587 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1589 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1590 the use of a port name that isn't defined in /etc/services) occurred, the
1591 message was deferred as in a normal delivery, and thus remained on the
1592 spool, instead of being failed because of the mua_wrapper setting. This
1593 is now fixed, and I tidied up some of the mua_wrapper messages at the
1596 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1597 the messages in a hash of arrays rather than using individual hashes.
1598 This is a bit cleaner and results in dramatic memory savings, albeit
1599 at a slight CPU cost.
1601 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1602 as requested by Marc Sherman.
1604 SC/10 Eximstats: added histograms for user specified patterns as requested
1607 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1609 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1610 fopen() in the content-scanning modules that did not already have it.
1616 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1617 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1619 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1621 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1623 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1624 to test Sieve filters that use "vacation".
1626 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1627 that changes the way the GnuTLS parameters are stored in the cache file.
1628 The new format can be generated externally. For backward compatibility,
1629 if the data in the cache doesn't make sense, Exim assumes it has read an
1630 old-format file, and it generates new data and writes a new file. This
1631 means that you can't go back to an older release without removing the
1634 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1635 work if there are any delivery delays because "one_time" forces the
1636 parent to be marked "delivered", so its unseen clone is never tried
1637 again. For this reason, Exim now forbids the simultaneous setting of
1640 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1641 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1642 entry if you want to know the details. The fix, however, neglected to
1643 consider the case when local delivery batching is involved. The test for
1644 "previously delivered" was not happening when checking to see if an
1645 address could be batched with a previous (undelivered) one; under
1646 certain circumstances this could lead to multiple deliveries to the same
1649 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1650 in its include files, and this causes problems building Exim.
1652 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1653 header_syntax) but Exim was just ignoring anything given after a slash.
1654 In particular, this caused confusion with an attempt to use "verify =
1655 reverse_host_lookup/defer_ok". An error is now given when options are
1656 supplied for verify items that do not have them. (Maybe reverse_host_
1657 lookup should have a defer_ok option, but that's a different point.)
1659 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1660 defined by RFC 821) to 2048, because there were problems with some AUTH
1661 commands, and RFC 1869 says the size should be increased for extended
1662 SMTP commands that take arguments.
1664 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1667 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1668 "unknown" error; now it says that the functionality isn't in the binary.
1670 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1671 an address' error message when a string expansion fails (syntax or
1672 whatever). Otherwise the password may appear in the log. Following change
1673 PH/42 below, there is no longer a chance of it appearing in a bounce
1676 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1678 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1679 of its addresses was ever tried. (Bugzilla bug #2.)
1681 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1682 the result incorrectly in the debug output. (It correctly added a newline
1683 to what was transported.)
1685 TF/01 Added $received_time.
1687 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1688 commented out examples of how to interface to a virus scanner and to
1689 SpamAssassin. Also added commented examples of av_scanner and
1690 spamd_address settings.
1692 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1693 and controls are allowed in which ACLs. There were a couple of minor
1694 errors. Some of the entries in the conditions table (which is a table of
1695 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1696 negation of where the condition IS allowed.
1698 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1700 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1701 header file does not have a version number, so I've had to invent a new
1702 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1703 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1704 radiusclient), but it was contributed by a Radius user.
1706 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1707 files or directories, for OpenSSL.
1709 PH/19 When an Exim process that is running as root has to create an Exim log
1710 file, it does so in a subprocess that runs as exim:exim so as to get the
1711 ownership right at creation (otherwise, other Exim processes might see
1712 the file with the wrong ownership). There was no test for failure of this
1713 fork() call, which would lead to the process getting stuck as it waited
1714 for a non-existent subprocess. Forks do occasionally fail when resources
1715 run out. I reviewed all the other calls to fork(); they all seem to check
1718 PH/20 When checking for unexpected SMTP input at connect time (before writing
1719 the banner), Exim was not dealing correctly with a non-positive return
1720 from the read() function. If the client had disconnected by this time,
1721 the result was a log entry for a synchronization error with an empty
1722 string after "input=" when read() returned zero. If read() returned -1
1723 (an event I could not check), uninitialized data bytes were printed.
1724 There were reports of junk text (parts of files, etc) appearing after
1727 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1729 PH/22 Added support for macro redefinition, and (re)definition in between
1730 driver and ACL definitions.
1732 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1733 forgetting to use the resulting value; it was using the unexpanded value.
1735 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1736 hadn't been configured. The fix is from Juergen Kreileder, who
1737 understands it better than I do:
1739 "Here's what I see happening with three configured cyrus_sasl
1740 authenticators configured (plain, login, cram-md5):
1742 On startup auth_cyrus_sasl_init() gets called for each of these.
1743 This means three calls to sasl_listmech() without a specified mech_list.
1744 => SASL tests which mechs of all available mechs actually work
1745 => three warnings about OTP not working
1746 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1748 With the patch, sasl_listmech() also gets called three times. But now
1749 SASL's mech_list option is set to the server_mech specified in the the
1750 authenticator. Or in other words, the answer from sasl_listmech()
1751 gets limited to just the mech you're testing for (which is different
1753 => the return list contains just 'plain' or 'login', 'cram-md5' or
1754 nothing depending on the value of ob->server_mech.
1756 I've just tested the patch: Authentication still works fine,
1757 unavailable mechs specified in the exim configuration are still
1758 caught, and the auth.log warnings about OTP are gone."
1760 PH/25 When debugging is enabled, the contents of the command line are added
1761 to the debugging output, even when log_selector=+arguments is not
1764 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1765 answer is "GNU", and only if the return is "GNU/something" is the answer
1768 PH/27 $acl_verify_message is now set immediately after the failure of a
1769 verification in an ACL, and so is available in subsequent modifiers. In
1770 particular, the message can be preserved by coding like this:
1772 warn !verify = sender
1773 set acl_m0 = $acl_verify_message
1775 Previously, $acl_verify_message was set only while expanding "message"
1776 and "log_message" when a very denied access.
1778 PH/28 Modified OS/os.c-Linux with
1780 -#ifndef OS_LOAD_AVERAGE
1781 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1783 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1784 nomenclature these days.)
1786 PH/29 Installed patch from the Sieve maintainer that adds the options
1787 sieve_useraddress and sieve_subaddress to the redirect router.
1789 PH/30 In these circumstances:
1790 . Two addresses routed to the same list of hosts;
1791 . First host does not offer TLS;
1792 . First host accepts first address;
1793 . First host gives temporary error to second address;
1794 . Second host offers TLS and a TLS session is established;
1795 . Second host accepts second address.
1796 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1797 and peerdn, if requested) that were in fact used only for the second
1800 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1801 attention to any local part prefix or suffix that was matched by the
1802 router that accepted the address. It now behaves in the same way as it
1803 does for delivery: the affixes are removed from the local part unless
1804 rcpt_include_affixes is set on the transport.
1806 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1807 timeout during the DATA phase of an incoming message.
1809 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1810 applied a patch sanctioned by the Sieve maintainer.
1812 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1813 the uid or gid is negative. A case of a negative gid caused this to be
1814 noticed. The fix allows for either to be negative.
1816 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1817 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1820 PH/36 The change PH/12 above was broken. Fixed it.
1822 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1823 the grounds that routing the same address twice would always produce the
1824 same answer. This might have been true once, but it is certainly no
1825 longer true now. Routing a child address may depend on the previous
1826 routing that produced that child. Some complicated redirection strategies
1827 went wrong when messages had multiple recipients, and made Exim's
1828 behaviour dependent on the order in which the addresses were given.
1830 I have moved the duplicate checking until after the routing is complete.
1831 Exim scans the addresses that are assigned to local and remote
1832 transports, and removes any duplicates. This means that more work will be
1833 done, as duplicates will always all be routed, but duplicates are
1834 presumably rare, so I don't expect this is of any significance.
1836 For deliveries to pipes, files, and autoreplies, the duplicate checking
1837 still happens during the routing process, since they are not going to be
1840 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1841 It corrects a timeout issue with spamd. This is Ian's comment: "The
1842 background is that sometimes spamd either never reads data from a
1843 connection it has accepted, or it never writes response data. The exiscan
1844 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1845 blindly assumes that writes won't block so it may never time out."
1847 PH/39 Allow G after quota size as well as K and M.
1849 PH/40 The value set for $authenticated_id in an authenticator may not contain
1850 binary zeroes or newlines because the value is written to log lines and
1851 to spool files. There was no check on this. Now the value is run through
1852 the string_printing() function so that such characters are converted to
1853 printable escape sequences.
1855 PH/41 $message_linecount is a new variable that contains the total number of
1856 lines in the message. Compare $body_linecount, which is the count for the
1859 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1860 bounce and delay warning messages, except in certain special cases, which
1863 (a) An SMTP error message from a remote host;
1864 (b) A message specified in a :fail: redirection;
1865 (c) A message specified in a "fail" command in a system filter;
1866 (d) A message specified in a FAIL return from the queryprogram router;
1867 (e) A message specified by the cannot_route_message router option.
1869 In these cases only, Exim does include the error details in bounce and
1870 warning messages. There are also a few cases where bland messages such
1871 as "unrouteable address" or "local delivery error" are given.
1873 PH/43 $value is now also set for the "else" part of a ${run expansion.
1875 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1876 being worked on, but at least Exim now implements the latest version to
1879 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1880 process to complete was treated as a delivery failure, a timeout while
1881 writing the message to the pipe was logged, but erroneously treated as a
1882 successful delivery. Such timeouts include transport filter timeouts. For
1883 consistency with the overall process timeout, these timeouts are now
1884 treated as errors, giving rise to delivery failures by default. However,
1885 there is now a new Boolean option for the pipe transport called
1886 timeout_defer, which, if set TRUE, converts the failures into defers for
1887 both kinds of timeout. A transport filter timeout is now identified in
1890 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1891 systems where "make" and "gmake" are different, calling "gmake" at top
1892 level broke things. I've arranged for the value of $(MAKE) to be passed
1893 from the Makefile to this script so that it can call the same version of
1897 A note about Exim versions 4.44 and 4.50
1898 ----------------------------------------
1900 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1901 changes of various kinds. As a consequence, a big documentation update was
1902 needed. This delayed the release for rather longer than seemed good, especially
1903 in the light of a couple of (minor) security issues. Therefore, the changes
1904 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1905 release. So 4.44 and 4.50 are in effect two different branches that both start
1908 I have left the 4.50 change log unchanged; it contains all the changes since
1909 4.43. The change log for 4.44 is below; many of its items are identical to
1910 those for 4.50. This seems to be the most sensible way to preserve the
1911 historical information.
1917 1. Minor wording change to the doc/README.SIEVE file.
1919 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1920 computation of the current number of files was incorrect.
1922 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1923 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1926 4. Give more explanation in the error message when the command for a transport
1927 filter fails to execute.
1929 5. There are several places where Exim runs a non-Exim command in a
1930 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1931 was being done only for the command run by the queryprogram router. It is
1932 now done for all such subprocesses. The other cases are: ${run, transport
1933 filters, and the commands run by the lmtp and pipe transports.
1935 6. Added CONFIGURE_GROUP build-time option.
1937 7. Some older OS have a limit of 256 on the maximum number of file
1938 descriptors. Exim was using setrlimit() to set 1000 as a large value
1939 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1940 systems. I've change it so that if it can't get 1000, it tries for 256.
1942 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1943 was an oversight, and furthermore, ever since the addition of extra
1944 controls (e.g. 4.43/32), the checks on when to allow different forms of
1945 "control" were broken. There should now be diagnostics for all cases when a
1946 control that does not make sense is encountered.
1948 9. Added the /retain_sender option to "control=submission".
1950 10. $recipients is now available in the predata ACL (oversight).
1952 11. Tidy the search cache before the fork to do a delivery from a message
1953 received from the command line. Otherwise the child will trigger a lookup
1954 failure and thereby defer the delivery if it tries to use (for example) a
1955 cached ldap connection that the parent has called unbind on.
1957 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1958 of $address_data from the recipient verification was clobbered by the
1959 sender verification.
1961 13. The value of address_data from a sender verification is now available in
1962 $sender_address_data in subsequent conditions in the ACL statement.
1964 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1966 15. Added a new option "connect=<time>" to callout options, to set a different
1969 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1970 was its contents. (It was OK if the option was not defined at all.)
1972 17. A "Completed" log line is now written for messages that are removed from
1973 the spool by the -Mrm option.
1975 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1976 information about exactly what failed.
1978 19. Added -dd to debug only the daemon process.
1980 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1981 handles timeouts, both on the server side and network timeouts. Renamed the
1982 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1984 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1985 It is now set to "smtps".
1987 22. $host_address is now set to the target address during the checking of
1988 ignore_target_hosts.
1990 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1991 being passed; this would have caused $sender_host_name to have been used if
1992 matching the list had actually called for a host name (not very likely,
1993 since this list is usually IP addresses). A host name is now passed as
1996 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1997 code that specifies a non-restarting handler (typically sigaction() in
1998 modern systems) in an attempt to fix a rare and obscure crash bug.
2000 25. Narrowed the window for a race in the daemon that could cause it to ignore
2001 SIGCHLD signals. This is not a major problem, because they are used only to
2002 wake it up if nothing else does.
2004 26. A malformed maildirsize file could cause Exim to calculate negative values
2005 for the mailbox size or file count. Odd effects could occur as a result.
2006 The maildirsize information is now recalculated if the size or filecount
2009 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2010 support for a long time. Removed HAVE_SYS_VFS_H.
2012 28. Installed the latest version of exipick from John Jetmore.
2014 29. In an address list, if the pattern was not a regular expression, an empty
2015 subject address (from a bounce message) matched only if the pattern was an
2016 empty string. Non-empty patterns were not even tested. This was the wrong
2017 because it is perfectly reasonable to use an empty address as part of a
2018 database query. An empty address is now tested by patterns that are
2019 lookups. However, all the other forms of pattern expect the subject to
2020 contain a local part and a domain, and therefore, for them, an empty
2021 address still always fails if the pattern is not itself empty.
2023 30. Exim went into a mad DNS loop when attempting to do a callout where the
2024 host was specified on an smtp transport, and looking it up yielded more
2025 than one IP address.
2027 31. Re-factored the code for checking spool and log partition space into a
2028 function that finds that data and another that does the check. The former
2029 is then used to implement four new variables: $spool_space, $log_space,
2030 $spool_inodes, and $log_inodes.
2032 32. The RFC2047 encoding function was originally intended for short strings
2033 such as real names; it was not keeping to the 75-character limit for
2034 encoded words that the RFC imposes. It now respects the limit, and
2035 generates multiple encoded words if necessary. To be on the safe side, I
2036 have increased the buffer size for the ${rfc2047: expansion operator from
2039 33. It is now permitted to omit both strings after an "if" condition; if the
2040 condition is true, the result is "true". As before, when the second string
2041 is omitted, a false condition yields an empty string. This makes it less
2042 cumbersome to write custom ACL and router conditions.
2044 34. Failure to deliver a bounce message always caused it to be frozen, even if
2045 there was an errors_to setting on the router. The errors_to setting is now
2048 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2049 canonical form (fully expanded) before being placed in
2050 $sender_host_address.
2052 36. The table in the code that translates DNS record types into text (T_A to
2053 "A" for instance) was missing entries for NS and CNAME. It is just possible
2054 that this could have caused confusion if both these types were looked up
2055 for the same domain, because the text type is used as part of Exim's
2056 per-process caching. But the chance of anyone hitting this buglet seems
2059 37. The dnsdb lookup has been extended in a number of ways.
2061 (1) There is a new type, "zns", which walks up the domain tree until it
2062 finds some nameserver records. It should be used with care.
2064 (2) There is a new type, "mxh", which is like "mx" except that it returns
2065 just the host names, not the priorities.
2067 (3) It is now possible to give a list of domains (or IP addresses) to be
2068 looked up. The behaviour when one of the lookups defers can be
2069 controlled by a keyword.
2071 (4) It is now possible to specify the separator character for use when
2072 multiple records are returned.
2074 38. The dnslists ACL condition has been extended: it is now possible to supply
2075 a list of IP addresses and/or domains to be looked up in a particular DNS
2078 39. Added log_selector=+queue_time_overall.
2080 40. When running the queue in the test harness, wait just a tad after forking a
2081 delivery process, to get repeatability of debugging output.
2083 41. Include certificate and key file names in error message when GnuTLS fails
2084 to set them up, because the GnuTLS error message doesn't include the name
2085 of the failing file when there is a problem reading it.
2087 42. Allow both -bf and -bF in the same test run.
2089 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2091 44. The "Exiscan patch" is now merged into the mainline Exim source.
2093 45. Sometimes the final signoff response after QUIT could fail to get
2094 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2095 < 0 before doing a fflush(). This bug looks as though it goes back to the
2096 introduction of TLS in release 3.20, but "sometimes" must have been rare
2097 because the tests only now provoked it.
2099 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2100 (this can affect the format of dates).
2102 47. exim_tidydb, when checking for the continued existence of a message for
2103 which it has found a message-specific retry record, was not finding
2104 messages that were in split spool directories. Consequently, it was
2105 deleting retry records that should have stayed in existence.
2107 48. Steve fixed some bugs in eximstats.
2109 49. The SPA authentication driver was not abandoning authentication and moving
2110 on to the next authenticator when an expansion was forced to fail,
2111 contradicting the general specification for all authenticators. Instead it
2112 was generating a temporary error. It now behaves as specified.
2114 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2115 (the order specifies the preference for clients). The order is now AES256,
2116 AES128, 3DES, ARCFOUR128.
2118 51. Small patch to Sieve code - explicitly set From: when generating an
2121 52. Exim crashed if a remote delivery caused a very long error message to be
2122 recorded - for instance if somebody sent an entire SpamAssassin report back
2123 as a large number of 550 error lines. This bug was coincidentally fixed by
2124 increasing the size of one of Exim's internal buffers (big_buffer) that
2125 happened as part of the Exiscan merge. However, to be on the safe side, I
2126 have made the code more robust (and fixed the comments that describe what
2129 53. Now that there can be additional text after "Completed" in log lines (if
2130 the queue_time_overall log selector is set), a one-byte patch to exigrep
2131 was needed to allow it to recognize "Completed" as not the last thing in
2134 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2135 patch that reportedly fixes this has been added. I am not expert enough to
2136 create a test for it. This is what the patch creator wrote:
2138 "I found a little strange behaviour of ldap code when working with
2139 Windows 2003 AD Domain, where users was placed in more than one
2140 Organization Units. When I tried to give exim partial DN, the exit code
2141 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2142 But simultaneously result of request was absolutely normal ldap result,
2143 so I produce this patch..."
2145 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2146 REFERENCE, so I have modified the code to exclude the patch when that macro
2149 55. Some experimental protocols are using DNS PTR records for new purposes. The
2150 keys for these records are domain names, not reversed IP addresses. The
2151 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2152 leaves it alone. Component reversal etc. now happens only for IP addresses.
2155 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2157 57. Double the size of the debug message buffer (to 2048) so that more of very
2158 long debug lines gets shown.
2160 58. The exicyclog utility now does better if the number of log files to keep
2161 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2163 59. Two changes related to the smtp_active_hostname option:
2165 (1) $smtp_active_hostname is now available as a variable.
2166 (2) The default for smtp_banner uses $smtp_active_hostname instead
2167 of $primary_hostname.
2169 60. The host_aton() function is supposed to be passed a string that is known
2170 to be a valid IP address. However, in the case of IPv6 addresses, it was
2171 not checking this. This is a hostage to fortune. Exim now panics and dies
2172 if the condition is not met. A case was found where this could be provoked
2173 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2174 components; fortuitously, this particular loophole had already been fixed
2175 by change 4.50/55 above.
2177 If there are any other similar loopholes, the new check in host_aton()
2178 itself should stop them being exploited. The report I received stated that
2179 data on the command line could provoke the exploit when Exim was running as
2180 exim, but did not say which command line option was involved. All I could
2181 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2182 running as the user.
2185 61. There was a buffer overflow vulnerability in the SPA authentication code
2186 (which came originally from the Samba project). I have added a test to the
2187 spa_base64_to_bits() function which I hope fixes it.
2190 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2191 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2193 63. The daemon start-up calls getloadavg() while still root for those OS that
2194 need the first call to be done as root, but it missed one case: when
2195 deliver_queue_load_max is set with deliver_drop_privilege. This is
2196 necessary for the benefit of the queue runner, because there is no re-exec
2197 when deliver_drop_privilege is set.
2199 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2200 This has been fixed.
2202 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2203 was in use, was not putting the data itself into the right store pool;
2204 consequently, it could be overwritten for a subsequent message in the same
2205 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2208 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2210 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2211 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2212 the function were treating the return as a boolean value, which happened to
2213 work because 0=false and not-0=true, but is not correct code.
2215 68. The host_aton() function was not handling scoped IPv6 addresses (those
2216 with, for example, "%eth0" on the end) correctly.
2218 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2219 negated items (that is, ~something) in unsigned ints. Some compilers
2220 apparently mutter when there is no cast.
2222 70. If an address verification called from an ACL failed, and did not produce a
2223 user-specific message (i.e. there was only a "system" message), nothing was
2224 put in $acl_verify_message. In this situation, it now puts the system
2227 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2228 session; change 4.31/43 added the unwanted input to the log line - except
2229 that it did not do this in the start of session case. It now does.
2231 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2232 This is wrong and can cause the other end to generate a synchronization
2233 error if it is another Exim or anything else that does the synchronization
2234 check. A QUIT command is no longer sent after a timeout.
2236 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2237 during host lookups.
2239 74. The defer_ok option of callout verification was not working if it was used
2240 when verifying addresses in header lines, that is, for this case:
2242 verify = header_sender/callout=defer_ok
2244 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2245 those file descriptors could be used for SMTP connections. If anything
2246 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2247 could be sent to the SMTP client, causing chaos. The daemon now opens
2248 stdin, stdout, and stderr to /dev/null when it puts itself into the
2251 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2252 log by default. The user can override this with suitable Perl magic.
2254 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2255 the log message when discard triggers, was not working for the DATA ACL or
2256 for the non-SMTP ACL.
2258 78. Error message wording change in sieve.c.
2260 79. If smtp_accept_max_per_host was set, the number of connections could be
2261 restricted to fewer than expected, because the daemon was trying to set up
2262 a new connection before checking whether the processes handling previous
2263 connections had finished. The check for completed processes is now done
2264 earlier. On busy systems, this bug wouldn't be noticed because something
2265 else would have woken the daemon, and it would have reaped the completed
2268 80. If a message was submitted locally by a user whose login name contained one
2269 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2270 It caused a spool format error. I have fixed the spool reading code. A
2271 related problem was that the "from" clause in the Received: line became
2272 illegal because of the space(s). It is now covered by ${quote_local_part.
2274 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2277 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2279 83. Patch from Sieve maintainer to fix unterminated string problem in
2280 "vacation" handling.
2282 84. Some minor changes to the Linux configuration files to help with other
2283 OS variants using glibc.
2285 85. One more patch for Sieve to update vacation handling to latest spec.
2288 ----------------------------------------------------
2289 See the note above about the 4.44 and 4.50 releases.
2290 ----------------------------------------------------
2296 1. Change 4.43/35 introduced a bug that caused file counts to be
2297 incorrectly computed when quota_filecount was set in an appendfile
2300 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2301 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2304 3. Give more explanation in the error message when the command for a transport
2305 filter fails to execute.
2307 4. There are several places where Exim runs a non-Exim command in a
2308 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2309 was being done only for the command run by the queryprogram router. It is
2310 now done for all such subprocesses. The other cases are: ${run, transport
2311 filters, and the commands run by the lmtp and pipe transports.
2313 5. Some older OS have a limit of 256 on the maximum number of file
2314 descriptors. Exim was using setrlimit() to set 1000 as a large value
2315 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2316 systems. I've change it so that if it can't get 1000, it tries for 256.
2318 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2319 was an oversight, and furthermore, ever since the addition of extra
2320 controls (e.g. 4.43/32), the checks on when to allow different forms of
2321 "control" were broken. There should now be diagnostics for all cases when a
2322 control that does not make sense is encountered.
2324 7. $recipients is now available in the predata ACL (oversight).
2326 8. Tidy the search cache before the fork to do a delivery from a message
2327 received from the command line. Otherwise the child will trigger a lookup
2328 failure and thereby defer the delivery if it tries to use (for example) a
2329 cached ldap connection that the parent has called unbind on.
2331 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2332 of $address_data from the recipient verification was clobbered by the
2333 sender verification.
2335 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2336 was its contents. (It was OK if the option was not defined at all.)
2338 11. A "Completed" log line is now written for messages that are removed from
2339 the spool by the -Mrm option.
2341 12. $host_address is now set to the target address during the checking of
2342 ignore_target_hosts.
2344 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2345 being passed; this would have caused $sender_host_name to have been used if
2346 matching the list had actually called for a host name (not very likely,
2347 since this list is usually IP addresses). A host name is now passed as
2350 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2351 code that specifies a non-restarting handler (typically sigaction() in
2352 modern systems) in an attempt to fix a rare and obscure crash bug.
2354 15. Narrowed the window for a race in the daemon that could cause it to ignore
2355 SIGCHLD signals. This is not a major problem, because they are used only to
2356 wake it up if nothing else does.
2358 16. A malformed maildirsize file could cause Exim to calculate negative values
2359 for the mailbox size or file count. Odd effects could occur as a result.
2360 The maildirsize information is now recalculated if the size or filecount
2363 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2364 support for a long time. Removed HAVE_SYS_VFS_H.
2366 18. Updated exipick to current release from John Jetmore.
2368 19. Allow an empty sender to be matched against a lookup in an address list.
2369 Previously the only cases considered were a regular expression, or an
2372 20. Exim went into a mad DNS lookup loop when doing a callout where the
2373 host was specified on the transport, if the DNS lookup yielded more than
2376 21. The RFC2047 encoding function was originally intended for short strings
2377 such as real names; it was not keeping to the 75-character limit for
2378 encoded words that the RFC imposes. It now respects the limit, and
2379 generates multiple encoded words if necessary. To be on the safe side, I
2380 have increased the buffer size for the ${rfc2047: expansion operator from
2383 22. Failure to deliver a bounce message always caused it to be frozen, even if
2384 there was an errors_to setting on the router. The errors_to setting is now
2387 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2388 canonical form (fully expanded) before being placed in
2389 $sender_host_address.
2391 24. Updated eximstats to version 1.33
2393 25. Include certificate and key file names in error message when GnuTLS fails
2394 to set them up, because the GnuTLS error message doesn't include the name
2395 of the failing file when there is a problem reading it.
2397 26. Expand error message when OpenSSL has problems setting up cert/key files.
2400 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2401 (this can affect the format of dates).
2403 28. exim_tidydb, when checking for the continued existence of a message for
2404 which it has found a message-specific retry record, was not finding
2405 messages that were in split spool directories. Consequently, it was
2406 deleting retry records that should have stayed in existence.
2408 29. eximstats updated to version 1.35
2409 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2410 1.35 - bugfix such that pie charts by volume are generated correctly
2412 30. The SPA authentication driver was not abandoning authentication and moving
2413 on to the next authenticator when an expansion was forced to fail,
2414 contradicting the general specification for all authenticators. Instead it
2415 was generating a temporary error. It now behaves as specified.
2417 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2418 (the order specifies the preference for clients). The order is now AES256,
2419 AES128, 3DES, ARCFOUR128.
2421 31. Small patch to Sieve code - explicitly set From: when generating an
2424 32. Exim crashed if a remote delivery caused a very long error message to be
2425 recorded - for instance if somebody sent an entire SpamAssassin report back
2426 as a large number of 550 error lines. This bug was coincidentally fixed by
2427 increasing the size of one of Exim's internal buffers (big_buffer) that
2428 happened as part of the Exiscan merge. However, to be on the safe side, I
2429 have made the code more robust (and fixed the comments that describe what
2432 33. Some experimental protocols are using DNS PTR records for new purposes. The
2433 keys for these records are domain names, not reversed IP addresses. The
2434 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2435 leaves it alone. Component reversal etc. now happens only for IP addresses.
2438 34. The host_aton() function is supposed to be passed a string that is known
2439 to be a valid IP address. However, in the case of IPv6 addresses, it was
2440 not checking this. This is a hostage to fortune. Exim now panics and dies
2441 if the condition is not met. A case was found where this could be provoked
2442 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2443 components; fortuitously, this particular loophole had already been fixed
2444 by change 4.50/55 or 4.44/33 above.
2446 If there are any other similar loopholes, the new check in host_aton()
2447 itself should stop them being exploited. The report I received stated that
2448 data on the command line could provoke the exploit when Exim was running as
2449 exim, but did not say which command line option was involved. All I could
2450 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2451 running as the user.
2454 35. There was a buffer overflow vulnerability in the SPA authentication code
2455 (which came originally from the Samba project). I have added a test to the
2456 spa_base64_to_bits() function which I hope fixes it.
2459 36. The daemon start-up calls getloadavg() while still root for those OS that
2460 need the first call to be done as root, but it missed one case: when
2461 deliver_queue_load_max is set with deliver_drop_privilege. This is
2462 necessary for the benefit of the queue runner, because there is no re-exec
2463 when deliver_drop_privilege is set.
2465 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2466 was in use, was not putting the data itself into the right store pool;
2467 consequently, it could be overwritten for a subsequent message in the same
2468 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2471 38. Sometimes the final signoff response after QUIT could fail to get
2472 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2473 < 0 before doing a fflush(). This bug looks as though it goes back to the
2474 introduction of TLS in release 3.20, but "sometimes" must have been rare
2475 because the tests only now provoked it.
2481 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2482 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2483 Now it returns an integer. A number of calls were still expecting a T/F
2484 return. Fortuitously, in all cases, the tests worked in OK situations,
2485 which is the norm. However, things would have gone wrong on any write
2486 failures on the smtp file descriptor. This function is used when sending
2487 messages over SMTP and also when doing verify callouts.
2489 2. When Exim is called to do synchronous delivery of a locally submitted
2490 message (the -odf or -odi options), it no longer closes stderr before doing
2493 3. Implemented the mua_wrapper option.
2495 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2497 5. Implemented the functions header_remove(), header_testname(),
2498 header_add_at_position(), and receive_remove_recipient(), and exported them
2501 6. If an ACL "warn" statement specified the addition of headers, Exim already
2502 inserted X-ACL-Warn: at the start if there was no header name. However, it
2503 was not making this test for the second and subsequent header lines if
2504 there were newlines in the string. This meant that an invalid header could
2505 be inserted if Exim was badly configured.
2507 7. Allow an ACL "warn" statement to add header lines at the start or after all
2508 the Received: headers, as well as at the end.
2510 8. Added the rcpt_4xx retry error code.
2512 9. Added postmaster_mailfrom=xxx to callout verification option.
2514 10. Added mailfrom=xxxx to the callout verification option, for verify=
2517 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2518 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2520 12. Inserted some casts to stop certain compilers warning when using pointer
2521 differences as field lengths or precisions in printf-type calls (mostly
2522 affecting debugging statements).
2524 13. Added optional readline() support for -be (dynamically loaded).
2526 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2527 same clock tick as a message's arrival, so that its received time was the
2528 same as the "first fail" time on the retry record, and that message
2529 remained on the queue past the ultimate address timeout, every queue runner
2530 would try a delivery (because it was past the ultimate address timeout) but
2531 after another failure, the ultimate address timeout, which should have then
2532 bounced the address, did not kick in. This was a "< instead of <=" error;
2533 in most cases the first failure would have been in the next clock tick
2534 after the received time, and all would be well.
2536 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2537 being recognized when the domain list was tested by the match_domain
2538 condition in an expansion string.
2540 16. Added the ${str2b64: operator.
2542 17. Exim was always calling setrlimit() to set a large limit for the number of
2543 processes, without checking whether the existing limit was already
2544 adequate. (It did check for the limit on file descriptors.) Furthermore,
2545 errors from getrlimit() and setrlimit() were being ignored. Now they are
2546 logged to the main and panic logs, but Exim does carry on, to try to do its
2547 job under whatever limits there are.
2549 18. Imported PCRE 5.0.
2551 19. Trivial typo in log message " temporarily refused connection" (the leading
2554 20. If the log selector return_path_on_delivery was set and an address was
2555 redirected to /dev/null, the delivery process crashed because it assumed
2556 that a return path would always be set for a "successful" delivery. In this
2557 case, the whole delivery is bypassed as an optimization, and therefore no
2560 21. Internal re-arrangement: the function for sending a challenge and reading
2561 a response while authentication was assuming a zero-terminated challenge
2562 string. It's now changed to take a pointer and a length, to allow for
2563 binary data in such strings.
2565 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2567 23. Exim was not respecting finduser_retries when seeking the login of the
2568 uid under which it was called; it was always trying 10 times. (The default
2569 setting of finduser_retries is zero.) Also, it was sleeping after the final
2570 failure, which is pointless.
2572 24. Implemented tls_on_connect_ports.
2574 25. Implemented acl_smtp_predata.
2576 26. If the domain in control=submission is set empty, Exim assumes that the
2577 authenticated id is a complete email address when it generates From: or
2578 Sender: header lines.
2580 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2581 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2582 chown and chgrp in /bin and hostname in /usr/bin.
2584 28. Exim was keeping the "process log" file open after each use, just as it
2585 does for the main log. This opens the possibility of it remaining open for
2586 long periods when the USR1 signal hits a daemon. Occasional processlog
2587 errors were reported, that could have been caused by this. Anyway, it seems
2588 much more sensible not to leave this file open at all, so that is what now
2591 29. The long-running daemon process does not normally write to the log once it
2592 has entered its main loop, and it closes the log before doing so. This is
2593 so that log files can straightforwardly be renamed and moved. However,
2594 there are a couple of unusual error situations where the daemon does write
2595 log entries, and I had neglected to close the log afterwards.
2597 30. The text of an SMTP error response that was received during a remote
2598 delivery was being truncated at 512 bytes. This is too short for some of
2599 the long messages that one sometimes sees. I've increased the limit to
2602 31. It is now possible to make retry rules that apply only when a message has a
2603 specific sender, in particular, an empty sender.
2605 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2606 it possible to be selective about when SMTP synchronization is enforced.
2608 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2610 32. Implemented hosts_connection_nolog.
2612 33. Added an ACL for QUIT.
2614 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2617 35. Added mailbox_size and mailbox_filecount to appendfile.
2619 36. Added control = no_multiline_responses to ACLs.
2621 37. There was a bug in the logic of the code that waits for the clock to tick
2622 in the case where the clock went backwards by a substantial amount such
2623 that the microsecond fraction of "now" was more than the microsecond
2624 fraction of "then" (but the whole seconds number was less).
2626 38. Added support for the libradius Radius client library this is found on
2627 FreeBSD (previously only the radiusclient library was supported).
2633 1. When certain lookups returned multiple values in the form name=value, the
2634 quoting of the values was not always being done properly. Specifically:
2635 (a) If the value started with a double quote, but contained no whitespace,
2637 (b) If the value contained whitespace other than a space character (i.e.
2638 tabs or newlines or carriage returns) it was not quoted.
2639 This fix has been applied to the mysql and pgsql lookups by writing a
2640 separate quoting function and calling it from the lookup code. The fix
2641 should probably also be applied to nisplus, ibase and oracle lookups, but
2642 since I cannot test any of those, I have not disturbed their existing code.
2644 2. A hit in the callout cache for a specific address caused a log line with no
2645 reason for rejecting RCPT. Now it says "Previous (cached) callout
2646 verification failure".
2648 3. There was an off-by-one bug in the queryprogram router. An over-long
2649 return line was truncated at 256 instead of 255 characters, thereby
2650 overflowing its buffer with the terminating zero. As well as fixing this, I
2651 have increased the buffer size to 1024 (and made a note to document this).
2653 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2654 when Exim is waiting for an SMTP response from a remote server, Exim
2655 restarts its select() call on the socket, thereby resetting its timeout.
2656 This is not a problem when such interrupts are rare. Somebody set up a cron
2657 job to run exiwhat every 2 minutes, which is less than the normal select()
2658 timeout (5 or 10 minutes). This meant that the select() timeout never
2659 kicked in because it was always reset. I have fixed this by comparing the
2660 time when an interrupt arrives with the time at the start of the first call
2661 to select(). If more time than the timeout has elapsed, the interrupt is
2662 treated as a timeout.
2664 5. Some internal re-factoring in preparation for the addition of Sieve
2665 extensions (by MH). In particular, the "personal" test is moved to a
2666 separate function, and given an option for scanning Cc: and Bcc: (which is
2667 not set for Exim filters).
2669 6. When Exim created an email address using the login of the caller as the
2670 local part (e.g. when creating a From: or Sender: header line), it was not
2671 quoting the local part when it contained special characters such as @.
2673 7. Installed new OpenBSD configuration files.
2675 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2676 try to make them clearer.
2678 9. Callout options, other than the timeout value, were being ignored when
2679 verifying sender addresses in header lines. For example, when using
2681 verify = header_sender/callout=no_cache
2683 the cache was (incorrectly) being used.
2685 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2686 only the Cygwin environment.
2688 11. When return_path_on_delivery was set as a log selector, if different remote
2689 addresses in the same message used different return paths and parallel
2690 remote delivery occurred, the wrong values would sometimes be logged.
2691 (Whenever a remote delivery process finished, the return path value from
2692 the most recently started remote delivery process was logged.)
2694 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2695 lines when AUTH and/or TLS are in use. This is the "received protocol"
2696 field. Exim used to use "asmtp" for authenticated SMTP, without any
2697 indication (in the protocol name) for TLS use. Now it follows the RFC and
2698 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2699 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2700 names appear in log lines as well as in Received: header lines.
2702 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2703 extensions, and comparison tests, and to fix some bugs.
2705 14. Changes to the "personal" filter test:
2707 (1) The test was buggy in that it was just doing the equivalent of
2708 "contains" tests on header lines. For example, if a user's address was
2709 anne@some.where, the "personal" test would incorrectly be true for
2711 To: susanne@some.where
2713 This test is now done by extracting each address from the header in turn,
2714 and checking the entire address. Other tests that are part of "personal"
2715 are now done using regular expressions (for example, to check local parts
2716 of addresses in From: header lines).
2718 (2) The list of non-personal local parts in From: addresses has been
2719 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2720 taken from the Sieve specification recommendations.
2722 (3) If the message contains any header line starting with "List-" it is
2723 treated as non-personal.
2725 (4) The test for "circular" in the Subject: header line has been removed
2726 because it now seems ill-conceived.
2728 15. Minor typos in src/EDITME comments corrected.
2730 16. Installed latest exipick from John Jetmore.
2732 17. If headers_add on a router specified a text string that was too long for
2733 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2734 of string_sprintf() is now avoided.
2736 18. $message_body_size was not set (it was always zero) when running the DATA
2737 ACL and the local_scan() function.
2739 19. For the "mail" command in an Exim filter, no default was being set for
2740 the once_repeat time, causing a random time value to be used if "once" was
2741 specified. (If the value happened to be <= 0, no repeat happened.) The
2742 default is now 0s, meaning "never repeat". The "vacation" command was OK
2743 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2744 (I found it when inspecting the code).
2746 20. There is now an overall timeout for performing a callout verification. It
2747 defaults to 4 times the callout timeout, which applies to individual SMTP
2748 commands during the callout. The overall timeout applies when there is more
2749 than one host that can be tried. The timeout is checked before trying the
2750 next host. This prevents very long delays if there are a large number of
2751 hosts and all are timing out (e.g. when the network connections are timing
2752 out). The value of the overall timeout can be changed by specifying an
2753 additional sub-option for "callout", called "maxwait". For example:
2755 verify = sender/callout=5s,maxwait=20s
2757 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2758 to the end before writing, but this should make it even safer).
2760 22. Exim was forgetting that it had advertised PIPELINING for the second and
2761 subsequent messages on an SMTP connection. It was also not resetting its
2762 memory on STARTTLS and an internal HELO.
2764 23. When Exim logs an SMTP synchronization error within a session, it now
2765 records whether PIPELINING has been advertised or not.
2767 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2768 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2769 rather than long int.
2771 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2773 26. Added the never_mail option to autoreply.
2779 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2780 crash if the getsockname() call failed; this can happen if a connection is
2781 closed very soon after it is established. The problem was simply in the
2782 order in which certain operations were done, causing Exim to try to write
2783 to the SMTP stream before it had set up the file descriptor. The bug has
2784 been fixed by making things happen in the correct order.
2790 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2791 before the connection was closed, thus losing the rejection response.
2793 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2794 some early Solaris releases, but causes trouble in current releases where
2795 socklen_t is defined.
2797 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2800 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2803 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2804 that releasing the top of it at the end releases what was used for sub-
2805 expansions (unless the block got too big). However, discard this block if
2806 the first thing is a variable or header, so that we can use its block when
2807 it is dynamic (useful for very large $message_headers, for example).
2809 6. Lookups now cache *every* query, not just the most recent. A new, separate
2810 store pool is used for this. It can be recovered when all lookup caches are
2811 flushed. Lookups now release memory at the end of their result strings.
2812 This has involved some general refactoring of the lookup sources.
2814 7. Some code has been added to the store_xxx() functions to reduce the amount
2815 of flapping under certain conditions.
2817 8. log_incoming_interface used to affect only the <= reception log lines. Now
2818 it causes the local interface and port to be added to several more SMTP log
2819 lines, for example "SMTP connection from", and rejection lines.
2821 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2823 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2825 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2826 could be overwritten at the end of the current message (or the start of a
2827 new message if it was set in a HELO ACL). The value is now preserved for
2828 the duration of the SMTP connection.
2830 12. If a transport had a headers_rewrite setting, and a matching header line
2831 contained an unqualified address, that address was qualified, even if it
2832 did not match any rewriting rules. The underlying bug was that the values
2833 of the flags that permit the existence of unqualified sender and recipient
2834 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2835 non-local messages, and by -bnq for local messages) were not being
2836 preserved with the message after it was received.
2838 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2839 "next input=" as part of the text comprising the host identity instead of
2840 the correct text. The code was using the same buffer for two different
2841 strings. However, depending on which order the printing function evaluated
2842 its arguments, the bug did not always show up. Under Linux, for example, my
2843 test suite worked just fine.
2845 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2846 doesn't work with some older versions of Perl. It has been changed to "my",
2847 which in any case is probably the better facility to use.
2849 15. A really picky compiler found some instances of statements for creating
2850 error messages that either had too many or two few arguments for the format
2853 16. The size of the buffer for calls to the DNS resolver has been increased
2854 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2855 for addresses that have a lot of PTR records. This alleviates a problem; it
2856 does not fully solve it.
2858 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2859 buffer now truncates the list and logs the incident, which is the same
2860 action as happens when Exim is looking up a host name and its aliases.
2861 Previously in this situation something unpredictable would happen;
2862 sometimes it was "internal error: store_reset failed".
2864 18. If a server dropped the connection unexpectedly when an Exim client was
2865 using GnuTLS and trying to read a response, the client delivery process
2866 crashed while trying to generate an error log message.
2868 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2869 string, for example:
2871 warn message = H1: something\nH2: something
2873 the text was added as a single header line from Exim's point of view
2874 though it ended up OK in the delivered message. However, searching for the
2875 second and subsequent header lines using $h_h2: did not work. This has been
2876 fixed. Similarly, if a system filter added multiple headers in this way,
2877 the routers could not see them.
2879 20. Expanded the error message when iplsearch is called with an invalid key to
2880 suggest using net-iplsearch in a host list.
2882 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2883 ACLs are no longer actually imposed (and a message to that effect is
2886 22. If a "gecos" field in a passwd entry contained escaped characters, in
2887 particular, if it contained a \" sequence, Exim got it wrong when building
2888 a From: or a Sender: header from that name. A second bug also caused
2889 incorrect handling when an unquoted " was present following a character
2890 that needed quoting.
2892 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2893 was not being matched caselessly.
2895 24. Arranged for all hyphens in the exim.8 source to be escaped with
2898 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2899 properly. Recipient callout cache records were still being keyed to include
2900 the sender, even when use_sender was set false. This led to far more
2901 callouts that were necessary. The sender is no longer included in the key
2902 when use_sender is false.
2904 26. Added "control = submission" modifier to ACLs.
2906 27. Added the ${base62d: operator to decode base 62 numbers.
2908 28. dnsdb lookups can now access SRV records.
2910 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2911 the configuration file.
2913 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2914 (-v) mode. This makes the output for a verbose queue run more intelligible.
2916 31. Added a use_postmaster feature to recipient callouts.
2918 32. Added the $body_zerocount variable, containing the number of binary zero
2919 bytes in the message body.
2921 33. The time of last modification of the "new" subdirectory is now used as the
2922 "mailbox time last read" when there is a quota error for a maildir
2925 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2927 35. Added +ignore_unknown as a special item in host lists.
2929 36. Code for decoding IPv6 addresses in host lists is now included, even if
2930 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2931 address was recognized as an IP address, but was then not correctly decoded
2932 into binary, causing unexpected and incorrect effects when compared with
2939 1. Very minor rewording of debugging text in manualroute to say "list of
2940 hosts" instead of "hostlist".
2942 2. If verify=header_syntax was set, and a header line with an unqualified
2943 address (no domain) and a large number of spaces between the end of the
2944 name and the colon was received, the reception process suffered a buffer
2945 overflow, and (when I tested it) crashed. This was caused by some obsolete
2946 code that should have been removed. The fix is to remove it!
2948 3. When running in the test harness, delay a bit after writing a bounce
2949 message to get a bit more predictability in the log output.
2951 4. Added a call to search_tidyup() just before forking a reception process. In
2952 theory, someone could use a lookup in the expansion of smtp_accept_max_
2953 per_host which, without the tidyup, could leave open a database connection.
2955 5. Added the variables $recipient_data and $sender_data which get set from a
2956 lookup success in an ACL "recipients" or "senders" condition, or a router
2957 "senders" option, similar to $domain_data and $local_part_data.
2959 6. Moved the writing of debug_print from before to after the "senders" test
2962 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2963 problems for message scanning, either using a data ACL, or using
2964 local_scan() because the Received: header was not generated till after they
2965 were called (in order to set the time as the time of reception completion).
2966 I have revised the way this works. The header is now generated after the
2967 body is received, but before the ACL or local_scan() are called. After they
2968 are run, the timestamp in the header is updated.
2974 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2975 before starting a queue runner without re-exec. This happened only when
2976 deliver_drop_privilege was set or when the Exim user was set to root. The
2977 effect of the bug was that timeouts during subsequent deliveries caused
2978 crashes instead of being properly handled. The handler is now left at its
2979 default (and expected) setting.
2981 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2982 message, again when deliver_drop_privilege is set or Exim is run as root.
2983 The bug described in (1) was not present in this case, but the tidying up
2984 of the other signals was missing. I have made the two cases consistent.
2986 3. The ignore_target_hosts setting on a manualroute router was being ignored
2987 for hosts that were looked up using the /MX notation.
2989 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2992 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2993 operated on the sender address. After changing the $sender_address to <>
2994 for the sender address verify, Exim was re-instated it as the original
2995 (before rewriting) address, but remembering that it had rewritten it, so it
2996 wasn't rewriting it again. This bug also had the effect of breaking the
2997 sender address verification caching when the sender address was rewritten.
2999 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3000 This has been changed so that if the ip literal address matches
3001 ignore_target_hosts, the router declines.
3003 7. Added expansion conditions match_domain, match_address, and match_local_
3004 part (NOT match_host).
3006 8. The placeholder for the Received: header didn't have a length field set.
3008 9. Added code to Exim itself and to exim_lock to test for a specific race
3009 condition that could lead to file corruption when using MBX delivery. The
3010 issue is with the lockfile that is created in /tmp. If this file is removed
3011 after a process has opened it but before that process has acquired a lock,
3012 there is the potential for a second process to recreate the file and also
3013 acquire a lock. This could lead to two Exim processes writing to the file
3014 at the same time. The added code performs the same test as UW imapd; it
3015 checks after acquiring the lock that its file descriptor still refers to
3016 the same named file.
3018 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3019 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3022 11. Added the smtp_active_hostname option. If used, this will typically be made
3023 to depend on the incoming interface address. Because $interface_address is
3024 not set up until the daemon has forked a reception process, error responses
3025 that can happen earlier (such as "too many connections") no longer contain
3028 12. If an expansion in a condition on a "warn" statement fails because a lookup
3029 defers, the "warn" statement is abandoned, and the next ACL statement is
3030 processed. Previously this caused the whole ACL to be aborted.
3032 13. Added the iplsearch lookup type.
3034 14. Added ident_timeout as a log selector.
3036 15. Added tls_certificate_verified as a log selector.
3038 16. Added a global option tls_require_ciphers (compare the smtp transport
3039 option of the same name). This controls incoming TLS connections.
3041 17. I finally figured out how to make tls_require_ciphers do a similar thing
3042 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3043 before starting the TLS session.
3045 18. Tabs are now shown as \t in -bP output.
3047 19. If the log selector return_path_on_delivery was set, Exim crashed when
3048 bouncing a message because it had too many Received: header lines.
3050 20. If two routers both had headers_remove settings, and the first one included
3051 a superfluous trailing colon, the final name in the first list and the
3052 first name in the second list were incorrectly joined into one item (with a
3053 colon in the middle).
3059 1. Added -C and -D options to the exinext utility, mainly to make it easier
3060 to include in the automated testing, but these could be helpful when
3061 multiple configurations are in use.
3063 2. The exinext utility was not formatting the output nicely when there was
3064 an alternate port involved in the retry record key, nor when there was a
3065 message id as well (for retries that were specific to a specific message
3066 and a specific host). It was also confused by IPv6 addresses, because of
3067 the additional colons they contain. I have fixed the IPv4 problem, and
3068 patched it up to do a reasonable job for IPv6.
3070 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3071 delivery, the log line now contains "pipelined" if PIPELINING was used.
3073 4. An SMTP transport process used to panic and die if the bind() call to set
3074 an explicit outgoing interface failed. This has been changed; it is now
3075 treated in the same way as a connect() failure.
3077 5. A reference to $sender_host_name in the part of a conditional expansion
3078 that was being skipped was still causing a DNS lookup. This no longer
3081 6. The def: expansion condition was not recognizing references to header lines
3082 that used bh_ and bheader_.
3084 7. Added the _cache feature to named lists.
3086 8. The code for checking quota_filecount in the appendfile transport was
3087 allowing one more file than it should have been.
3089 9. For compatibility with Sendmail, the command line option
3097 and sets the incoming protocol and host name (for trusted callers). The
3098 host name and its colon can be omitted when only the protocol is to be set.
3099 Note the Exim already has two private options, -pd and -ps, that refer to
3100 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3101 "s", but I don't think that's a major issue.
3103 10. A number of refactoring changes to the code, none of which should affect
3106 (a) The number of logging options was getting close to filling up the
3107 32-bit word that was used as a bit map. I have split them into two classes:
3108 those that are passed in the argument to log_write(), and those that are
3109 only ever tested independently outside of that function. These are now in
3110 separate 32-bit words, so there is plenty of room for expansion again.
3111 There is no change in the user interface or the logging behaviour.
3113 (b) When building, for example, log lines, the code previously used a
3114 macro that called string_cat() twice, in order to add two strings. This is
3115 not really sufficiently general. Furthermore, there was one instance where
3116 it was actually wrong because one of the argument was used twice, and in
3117 one call a function was used. (As it happened, calling the function twice
3118 did not affect the overall behaviour.) The macro has been replaced by a
3119 function that can join an arbitrary number of extra strings onto a growing
3122 (c) The code for expansion conditions now uses a table and a binary chop
3123 instead of a serial search (which was left over from when there were very
3124 few conditions). Also, it now recognizes conditions like "pam" even when
3125 the relevant support is not compiled in: a suitably worded error message is
3126 given if an attempt is made to use such a condition.
3128 11. Added ${time_interval:xxxxx}.
3130 12. A bug was causing one of the ddress fields not to be passed back correctly
3131 from remote delivery subprocesses. The field in question was not being
3132 subsequently used, so this caused to problems in practice.
3134 13. Added new log selectors queue_time and deliver_time.
3136 14. Might have fixed a bug in maildirsizefile handling that threw up
3137 "unexpected character" debug warnings, and recalculated the data
3138 unnecessarily. In any case, I expanded the warning message to give more
3141 15. Added the message "Restricted characters in address" to the statements in
3142 the default ACL that block characters like @ and % in local parts.
3144 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3145 Three changes have been made:
3147 (a) There was a serious bug; a negative response to MAIL caused the whole
3148 recipient domain to be cached as invalid, thereby blocking all messages
3149 to all local parts at the same domain, from all senders. This bug has
3150 been fixed. The domain is no longer cached after a negative response to
3151 MAIL if the sender used is not empty.
3153 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3156 (c) A new callout option, "use_sender" has been added for people who want
3157 the modified behaviour.
3163 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3166 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3167 indeed breaks things for older releases.
3169 3. Added additional logging to the case where there is a problem reading data
3170 from a filter that is running in a subprocess using a pipe, in order to
3171 try to track down a specific problem.
3173 4. Testing facility fudge: when running in the test harness and attempting
3174 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3175 getting "No route to host". Convert this to a timeout.
3177 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3180 6. Some OS don't have socklen_t but use size_t instead. This affects the
3181 fifth argument of getsockopt() amongst other things. This is now
3182 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3183 can be set for individual OS. I have set it for SunOS5, OSF1, and
3184 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3185 some earlier ones do not.
3187 7. Change 4.30/15 was not doing the test caselessly.
3189 8. The standard form for an IPv6 address literal was being rejected by address
3190 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3191 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3192 this, as well as the form without the "IPv6" on the front (but only when
3193 address literals are enabled, of course).
3195 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3197 10. Exim crashed if a message with an empty sender address specified by -f
3198 encountered a router with an errors_to setting. This could be provoked only
3199 by a command such as
3203 where an empty string was supplied; "<>" did not hit this bug.
3205 11. Installed PCRE release 4.5.
3207 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3208 remained set. It is now erased.
3210 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3211 times from message ids (which are base 36 rather than the normal 62).
3213 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3214 were being counted as actual protocol errors, and logged if the log
3215 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3216 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3217 and DATA following a set of rejected RCPTs do not count as protocol errors.
3218 In other words, Exim assumes they were pipelined, though this may not
3219 actually be the case. Of course, in all cases the client gets an
3220 appropriate error code.
3222 15. If a lookup fails in an ACL condition, a message about the failure may
3223 be available; it is used if testing the ACL cannot continue, because most
3224 such messages specify what the cause of the deferral is. However, some
3225 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3226 that caused an old message to be retained and used if a later statement
3227 caused a defer, replacing the real cause of the deferral.
3229 16. If an IP address had so many PTR records that the DNS lookup buffer
3230 was not large enough to hold them, Exim could crash while trying to process
3231 the truncated data. It now detects and logs this case.
3233 17. Further to 4.21/58, another change has been made: if (and only if) the
3234 first line of a message (the first header line) ends with CRLF, a bare LF
3235 in a subsequent header line has a space inserted after it, so as not to
3236 terminate the header.
3238 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3239 unnecessarily, used atoi() instead of strtol(), and didn't check the
3240 termination when getting file sizes from file names by regex.
3242 19. Completely re-implemented the support for maildirsize files, in the light
3243 of a number of problems with the previous contributed implementation
3244 (4.30/29). In particular:
3246 . If the quota is zero, the maildirsize file is maintained, but no quota is
3249 . If the maildir directory does not exist, it is created before any attempt
3250 to write a maildirsize file.
3252 . The quota value in the file is just a cache; if the quota is changed in
3253 the transport, the new value overrides.
3255 . A regular expression is available for excluding directories from the
3258 20. The autoreply transport checks the characters in options that define the
3259 message's headers; it allows continued headers, but it was checking with
3260 isspace() after an embedded newline instead of explicitly looking for a
3263 21. If all the "regular" hosts to which an address was routed had passed their
3264 expiry times, and had not reached their retry times, the address was
3265 bounced, even if fallback hosts were defined. Now Exim should go on to try
3268 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3269 equivalent code in the SMTP transport. Some hosts send humungous responses
3270 to HELO/EHLO, more than 1024 it seems.
3272 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3273 gives compiler warnings in some environments. I've now done it "properly",
3276 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3277 (because of problems with the built-in one) was declared to return uschar *
3278 instead of char *, causing compiler failure.
3280 25. Fixed a file descriptor leak when processing alias/forward files.
3282 26. Fixed a minor format string issue in dbfn.c.
3284 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3286 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3287 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3290 29. When a sender address is verified, it is cached, to save repeating the test
3291 when there is more than one recipient in a message. However, when the
3292 verification involves a callout, it is possible for different callout
3293 options to be set for different recipients. It is too complicated to keep
3294 track of this in the cache, so now Exim always runs a verification when a
3295 callout is required, relying on the callout cache for the optimization.
3296 The overhead is duplication of the address routing, but this should not be
3299 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3300 to be verified with callout=postmaster, and the main callout worked but the
3301 postmaster check failed, the verification correctly failed. However, if a
3302 subsequent RCPT command asked for sender verification *without* the
3303 postmaster check, incorrect caching caused this verification also to fail,
3306 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3307 it was not caching the DNS options (qualify_single, search_parents) that
3308 were used when the lookup failed. A subsequent lookup with different
3309 options therefore always gave the same answer, though there were cases
3310 where it should not have. (Example: a "domains = !$mx_any" option on a
3311 dnslookup router: the "domains" option is always processed without any
3312 widening, but the router might have qualify_single set.) Now Exim uses the
3313 cached value only when the same options are set.
3315 32. Added John Jetmore's "exipick" utility to the distribution.
3317 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3318 than a timeout (e.g. a certificate is required, and is not provided), an
3319 Exim server now closes the connection immediately. Previously it waited for
3320 the client to close - but if the client is SSL, it seems that they each
3321 wait for each other, leading to a delay before one of them times out.
3323 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3324 maintained 0.8.x compatibility because I don't think many are using it, and
3325 it is clearly obsolete.
3327 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3330 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3331 client certificate was expired. A simple patch fixes this, though I don't
3332 understand the full logic of why the verify callback is called multiple
3335 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3336 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3337 which causes problems with some clients (such as the Certicom SSL Plus
3338 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3339 disables the coutermeasure allowing Eudora to connect."
3341 38. Exim was not checking that a write() to a log file succeeded. This could
3342 lead to Bad Things if a log got too big, in particular if it hit a file
3343 size limit. Exim now panics and dies if it cannot write to a log file, just
3344 as it does if it cannot open a log file.
3346 39. Modified OS/Makefile-Linux so that it now contains
3348 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3350 The two -D definitions ensure that Exim is compiled with large file
3351 support, which makes it possible to handle log files that are bigger than
3354 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3355 instance) a domain was checked against a named list that involved a lookup,
3356 causing $domain_data to be set, then another domain was checked against the
3357 same list, then the first domain was re-checked, the value of $domain_data
3358 after the final check could be wrong. In particular, if the second check
3359 failed, it could be set empty. This bug probably also applied to
3362 41. The strip_trailing_dot option was not being applied to the address given
3363 with the -f command-line option.
3365 42. The code for reading a message's header from the spool was incrementing
3366 $received_count, but never initializing it. This meant that the value was
3367 incorrect (doubled) while delivering a message in the same process in which
3368 it was received. In the most common configuration of Exim, this never
3369 happens - a fresh exec is done - but it can happen when
3370 deliver_drop_privilege is set.
3372 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3373 it now includes up to 150 characters of the unexpected data in the log
3376 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3377 and building data strings. The size of both of these buffers was 10 000
3378 bytes - far larger than anybody would *ever* want, thought I. Needless to
3379 say, somebody hit the limit. I have increased the maximum line length to
3380 20 000 and the maximum data length of concatenated lines to 100 000. I have
3381 also fixed two bugs, because there was no checking on these buffers. Tsk,
3382 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3383 buffer is too small.
3385 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3386 lsearch lookups. Now it does.
3388 46. When parsing a route_list item in a manualroute router, a fixed-length
3389 buffer was used for the list of hosts. I made this 1024 bytes long,
3390 thinking that nobody would ever have a list of hosts that long. Wrong.
3391 Somebody had a whole pile of complicated expansion conditions, and the
3392 string was silently truncated, leading to an expansion error. It turns out
3393 that it is easier to change to an unlimited length (owing to other changes
3394 that have happened since this code was originally written) than to build
3395 structure for giving a limitation error. The length of the item that
3396 expands into the list of hosts is now unlimited.
3398 47. The lsearch lookup could not handle data where the length of text line was
3399 more than 4095 characters. Such lines were truncated, leading to shortened
3400 data being returned. It should now handle lines of any length.
3402 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3403 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3406 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3407 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3408 no longer happens in comment lines. A list of replacements is now placed
3409 at the head of all of the source files, except those whose only change is
3410 to replace PERL_COMMAND in the very first #! line.
3412 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3413 messages on the queue, with a bottom-up merge sort, using code contributed
3414 by Michael Haardt. This should make operations like -bp somewhat faster on
3415 large queues. It won't affect queue runners, except when queue_run_in_order
3418 51. Installed eximstats 1.31 in the distribution.
3420 52. Added support for SRV lookups to the dnslookup router.
3422 53. If an ACL referred to $message_body or $message_body_end, the value was not
3423 reset for any messages that followed in the same SMTP session.
3425 54. The store-handling optimization for building very long strings was not
3426 differentiating between the different store pools. I don't think this
3427 actually made any difference in practice, but I've tidied it.
3429 55. While running the routers to verify a sender address, $sender_address
3430 was still set to the sender address. This is wrong, because when routing to
3431 send a bounce to the sender, it would be empty. Therefore, I have changed
3432 it so that, while verifying a sender address, $sender_address is set to <>.
3433 (There is no change to what happens when verifying a recipient address.)
3435 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3436 target A or AAAA records (if not already returned) without resetting the
3437 qualify_single or search_parents options of the DNS resolver. These are
3438 inappropriate in this case because the targets of MX and SRV records must
3439 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3440 target that, when qualified, matched something in the local domain. These
3441 two options are now turned off when doing these lookups.
3443 57. It seems that at least some releases of Reiserfs (which does not have the
3444 concept of a fixed number of inodes) returns zero and not -1 for the
3445 number of available inodes. This interacted badly with check_spool_inodes,
3446 which assumed that -1 was the "no such thing" setting. What I have done is
3447 to check that the total number of inodes is greater than zero before doing
3448 the test of how many are available.
3450 58. When a "warn" ACL statement has a log_message modifier, the message is
3451 remembered, and not repeated. This is to avoid a lot of repetition when a
3452 message has many recipients that cause the same warning to be written.
3453 Howewer, Exim was preserving the list of already written lines for an
3454 entire SMTP session, which doesn't seem right. The memory is now reset if a
3455 new message is started.
3457 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3458 debugging output unless log_rewrite was also set.
3460 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3461 of (int)(handle) when we know that handle contains (void *)(-1).
3463 61. The Exim daemon panic-logs an error return when it closes the incoming
3464 connection. However "connection reset by peer" seems to be common, and
3465 isn't really an error worthy of noting specially, so that particular error
3468 62. When Exim is trying to find all the local interfaces, it used to panic and
3469 die if the ioctl to get the interface flags failed. However, it seems that
3470 on at least one OS (Solaris 9) it is possible to have an interface that is
3471 included in the list of interfaces, but for which you get a failure error
3472 for this call. This happens when the interface is not "plumbed" into a
3473 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3474 failure of the "get flags" call assumes that the interface is down.
3476 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3477 makes life easier for people who are doing arithmetic on fields extracted
3478 from dates, where you often get leading zeros that should not be
3479 interpreted as octal.
3481 64. Added qualify_domain to the redirect router, to override the global
3484 65. If a pathologically long header line contained very many addresses (the
3485 report of this problem mentioned 10 000) and each of them was rewritten,
3486 Exim could use up a very large amount of memory. (It kept on making new
3487 copies of the header line as it rewrote, and never released the old ones.)
3488 At the expense of a bit more processing, the header rewriting function has
3489 been changed so that it no longer eats memory in this way.
3491 66. The generation of the Received: header has been moved from the time that a
3492 message starts to be received, to the time that it finishes. The timestamp
3493 in the Received: header should now be very close to that of the <= log
3494 line. There are two side-effects of this change:
3496 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3497 logged header lines no longer include the local Received: line, because
3498 it has not yet been created. The same applies to a copy of the message
3499 that is returned to a non-SMTP sender when a message is rejected.
3501 (b) When a filter file is tested using -bf, no additional Received: header
3502 is added to the test message. After some thought, I decided that this
3505 This change does not affect the value of $received_for. It is still set
3506 after address rewriting, but before local_scan() is called.
3508 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3510 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3511 gave an unhelpful panic error message, and a defer error. I have managed to
3512 change this behaviour so that it now rejects any supplied certificate,
3513 which seems right, as the list of acceptable certificates is empty.
3515 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3516 gave an unhelpful defer error. I have not managed to make this reject any
3517 supplied certificates, but the error message it gives is "no certificate
3518 supplied", which is not helpful.
3520 70. exigrep's output now also includes lines that are not associated with any
3521 message, but which match the given pattern. Implemented by a patch from
3522 Martin Sluka, which also tidied up the Perl a bit.
3524 71. Recipient callout verification, like sender verification, was using <> in
3525 the MAIL FROM command. This isn't really the right thing, since the actual
3526 sender may affect whether the remote host accepts the recipient or not. I
3527 have changed it to use the actual sender in the callout; this means that
3528 the cache record is now keyed on a recipient/sender pair, not just the
3529 recipient address. There doesn't seem to be a real danger of callout loops,
3530 since a callout by the remote host to check the sender would use <>.
3531 [SEE ABOVE: changed after hitting problems.]
3533 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3534 temporary errors. However, in the case of such a code being given after
3535 the end of a data transmission (i.e. after ".") Exim was failing to write
3536 a retry record for the message. (Yes, there was some broken host that was
3537 actually sending 8xx at this point.)
3539 73. An unknown lookup type in a host list could cause Exim to panic-die when
3540 the list was checked. (An example that provoked this was putting <; in the
3541 middle of a list instead of at the start.) If this happened during a DATA
3542 ACL check, a -D file could be left lying around. This kind of configuration
3543 error no longer causes Exim to die; instead it causes a defer errror. The
3544 incident is still logged to the main and panic logs.
3546 74. Buglet left over from Exim 3 conversion. The message "too many messages
3547 in one connection" was written to the rejectlog but not the mainlog, except
3548 when address rewriting (yes!) was being logged.
3550 75. Added write_rejectlog option.
3552 76. When a system filter was run not as root (that is, when system_filter_user
3553 was set), the values of the $n variables were not being returned to the
3554 main process; thus, they were not subsequently available in the $sn
3557 77. Added +return_path_on_delivery log selector.
3559 78. A connection timeout was being treated differently from recipients deferred
3560 when testing hosts_max_try with a message that was older than the host's
3561 retry timeout. (The host should not be counted, thus allowing all hosts to
3562 be tried at least once before bouncing.) This may have been the cause of an
3563 occasionally reported bug whereby a message would remain on the queue
3564 longer than the retry timeout, but would be bounced if a delivery was
3565 forced. I say "may" because I never totally pinned down the problem;
3566 setting up timeout/retry tests is difficult. See also the next item.
3568 79. The ultimate address timeout was not being applied to errors that involved
3569 a combination of host plus message (for example, a timeout on a MAIL
3570 command). When an address resolved to a number of possible hosts, and they
3571 were not all tried for each delivery (e.g. because of hosts_max_try), a
3572 message could remain on the queue longer than the retry timeout.
3574 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3577 81. Fixed an obscure SMTP outgoing bug which required at least the following
3578 conditions: (a) there was another message waiting for the same server;
3579 (b) the server returned 5xx to all RCPT commands in the first message so
3580 that the message was not completed; (c) the server dropped the connection
3581 or gave a negative response to the RSET that Exim sends to abort the
3582 transaction. The observed case was a dropped connection after DATA that had
3583 been sent in pipelining mode. That is, the server had advertised PIPELINING
3584 but was not implementing it correctly. The effect of the bug was incorrect
3585 behaviour, such as trying another host, and this could lead to a crash.
3591 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3592 and daemon.c were passed as pointers to ints; they should have been
3593 pointers to socklen_t variables (which are typically unsigned ints).
3595 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3598 3. Fixed a really odd bug that affected only the testing scheme; patching a
3599 certain fixed string in the binary changed the value of another string that
3600 happened to be identical to the end of the original first string.
3602 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3603 name", it returns that address as the IP address. On some operating
3604 systems (e.g. Solaris), it also passes back the IP address string as the
3605 "host name". However, on others (e.g. Linux), it passes back an empty
3606 string. Exim wasn't checking for this, and was changing the host name to an
3607 empty string, assuming it had been canonicized.
3609 5. Although rare, it is permitted to have more than one PTR record for a given
3610 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3611 all the names associated with an address, because they do in Solaris.
3612 However, it seems that they do not in Linux for data that comes from the
3613 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3614 I found this out when I moved to a new Linux workstation and tried to run
3615 the Exim test suite.
3617 To get round this problem I have changed the code so that it now does its
3618 own call to the DNS to look up PTR records when searching for a host name.
3619 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3620 addresses that are only in /etc/hosts are still found.
3622 This behaviour is, however, controlled by an option called host_lookup_
3623 order, which defaults to "bydns:byaddr". If people want to use the other
3624 order, or indeed, just use one or the other means of lookup, they can
3625 specify it in this variable.
3627 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3628 some operating systems, this comes back from gethostbyaddr() as an empty
3629 string, and this is what Exim used to test for. However, it seems that in
3630 other systems, "." is yielded. Exim now tests for this case too.
3632 7. The values of check_spool_space and check_log_space are now held internally
3633 as a number of kilobytes instead of an absolute number of bytes. If a
3634 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3635 kilobyte. This means that much larger values can be stored.
3637 8. Exim monitor: an attempt to get the action menu when not actually pointing
3638 at a message produces an empty menu entitled "No message selected". This
3639 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3640 no entries in it ("Shell widget menu has zero width and/or height"). So I
3641 have added a single, blank menu entry in this case.
3643 9. Added ${quote_local_part.
3645 10. MIME decoding is now applied to the contents of Subject: header lines when
3648 11. Now that a reference to $sender_host_address automatically causes a reverse
3649 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3650 host lookup before query-style lookups in lists that might use this
3651 variable. This has therefore been abolished, and the "net-" prefix is no
3652 longer necessary for query-style lookups.
3654 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3655 have been a typo for LFLAGS, so it has been changed.
3657 13. The install script calls Exim with "-C /dev/null" in order to find the
3658 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3659 to be output. Howeve, since Exim outputs its version number before the
3660 error, it didn't break the script. It just looked ugly. I fixed this by
3661 always allowing "-C /dev/null" if the caller is root.
3663 14. Ignore overlarge ACL variable number when reading spool file - insurance
3664 against a later release with more variables having written the file.
3666 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3667 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3668 this, as well as the form without the "IPv6" on the front.
3670 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3671 OS/Makefile-Darwin file.
3673 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3674 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3676 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3677 variable. However, some OS use ints for inodes. I've added cast to long int
3678 to get rid of the compiler warning.
3680 19. I had forgotten to lock out "/../" in configuration file names when
3681 ALT_CONFIG_PREFIX was set.
3683 20. Routers used for verification do not need to specify transports. However,
3684 if such a router generated a host list, and callout was configured, Exim
3685 crashed, because it could not find a port number from the (non-existent)
3686 transport. It now assumes port 25 in this circumstance.
3688 21. Added the -t option to exigrep.
3690 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3691 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3692 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3693 them is set, LOOKUP_LSEARCH is forced.
3695 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3697 24. Added sender and host information to the "rejected by local_scan()" log
3698 line; previously there was no indication of these.
3700 25. Added .include_if_exists.
3702 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3703 Linux. It turns out that not all file systems support this. Apparently some
3704 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3705 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3706 not supported on the file descriptor, is now ignored when Exim is trying to
3707 sync a directory. This applies only to Linux.
3709 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3711 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3712 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3715 29. Added support for maildirsize files from supplied patch (modified a bit).
3717 30. The use of :fail: followed by an empty string could lead Exim to respond to
3718 sender verification failures with (e.g.):
3720 550 Verification failed for <xxx>
3721 550 Sender verify failed
3723 where the first response line was missing the '-' that indicates it is not
3724 the final line of the response.
3726 31. The loop for finding the name of the user that called Exim had a hardwired
3727 limit of 10; it now uses the value of finduser_retries, which is used for
3728 all other user lookups.
3730 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3733 33. Exim was neglecting to zero errno before one call of strtol() when
3734 expanding a string and expecting an integer value. On some systems this
3735 resulted in spurious "integer overflow" errors. Also, it was casting the
3736 result into an int without checking.
3738 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3739 did not work. The code looks as if it has *never* worked, though it appears
3740 to have been documented since at least releast 1.62. I have made it work.
3742 35. The "timeout_DNS" error in retry rules, also documented since at least
3743 1.62, also never worked. As it isn't clear exactly what this means, and
3744 clearly it isn't a major issue, I have abolished the feature by treating it
3745 as "timeout", and writing a warning to the main and panic logs.
3747 36. The display of retry rules for -brt wasn't always showing the error code
3750 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3751 timeout_connect_A, timeout_connect_MX.
3753 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3754 to the empty sender.
3756 39. The daemon was not analysing the content of -oX till after it had closed
3757 stderr and disconnected from the controlling terminal. This meant that any
3758 syntax errors were only noted on the panic log, and the return code from
3759 the command was 0. By re-arranging the code a little, I've made the
3760 decoding happen first, so such errors now appear on stderr, and the return
3761 code is 1. However, the actual setting up of the sockets still happens in
3762 the disconnected process, so errors there are still only recorded on the
3765 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3766 connections (as happens on some IP stacks) was logged at start up time as
3767 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3768 it from "IPv6 and IPv4", which means that two separate sockets are being
3771 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3772 says whether AF_INET or AF_INET6 was passed as an argument.
3774 42. Exiwhat output was messed up when time zones were included in log
3777 43. Exiwhat now gives more information about the daemon's listening ports,
3778 and whether -tls-on-connect was used.
3780 44. The "port" option of the smtp transport is now expanded.
3782 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3783 silently ignored. Now an error message is written to the main and panic
3786 46. There's a new ACL modifier called "logwrite" which writes to a log file
3787 as soon as it is encountered.
3789 47. Added $local_user_uid and $local_user_gid at routing time.
3791 48. Exim crashed when trying to verify a sender address that was being
3794 49. Exim was recognizing only a space character after ".include". It now also
3795 recognizes a tab character.
3797 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3798 extracting the relevant information from the specification. The man page no
3799 longer contains scrambled data for the -d option, and I've added a section
3800 at the front about calling Exim under different names.
3802 51. Added "extra_headers" argument to the "mail" command in filter files.
3804 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3807 53. Installed eximstats 1.29.
3809 54. Added transport_filter_timeout as a generic transport option.
3811 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3812 Cc: header lines. This was required by RFC 822, but it not required by RFC
3815 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3816 incoming messages that did not have them. Now it does so only if the
3817 message originates locally, that is, if there is no associated remote host
3818 address. When Resent- header lines are present, this applies to the Resent-
3819 lines rather than the non-Resent- lines.
3821 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3822 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3824 58. Messages for configuration errors now include the name of the main
3825 configuration file - useful now that there may be more than one file in a
3826 list (.included file names were always shown).
3828 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3829 for those rare installations that do not start the daemon as root or run it
3830 setuid root. I've cut out the call to initgroups() if the daemon is not
3833 60. The Exim user and group can now be bound into the binary as text strings
3834 that are looked up at the start of Exim's processing.
3836 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3838 62. Added $mailstore_basename variable.
3840 63. Installed patch to sieve.c from Michael Haardt.
3842 64. When Exim failed to open the panic log after failing to open the main log,
3843 the original message it was trying to log was written to stderr and debug
3844 output, but if they were not available (the usual case in production), it
3845 was lost. Now it is written to syslog before the two lines that record the
3846 failures to open the logs.
3848 65. Users' Exim filters run in subprocesses under the user's uid. It is
3849 possible for a "deliver" command or an alias in a "personal" command to
3850 provoke an address rewrite. If logging of address rewriting is configured,
3851 this fails because the process is not running as root or exim. There may be
3852 a better way of dealing with this, but for the moment (because 4.30 needs
3853 to be released), I have disabled address rewrite logging when running a
3854 filter in a non-root, non-exim process.
3860 1. The buildconfig auxiliary program wasn't quoting the value set for
3861 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3862 not defined. This bug was masked in 4.22 by the effect that was fixed in
3865 2. Some messages that were rejected after a message id was allocated were
3866 shown as "incomplete" by exigrep. It no longer does this for messages that
3867 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3869 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3870 have allow_domain_literals set, the ID did not get logged in the <= line.
3871 Domain literals are now always recognized in Message-ID: header lines.
3873 4. The first argument for a ${extract expansion item is the key name or field
3874 number. Leading and trailing spaces in this item were not being ignored,
3875 causing some misleading effects.
3877 5. When deliver_drop_privilege was set, single queue runner processes started
3878 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3879 same command in the process it spins off) were not dropping privilege.
3881 6. When the daemon running as "exim" started a queue runner, it always
3882 re-executed Exim in the spun-off process. This is a waste of effort when
3883 deliver_drop_privilege is set. The new process now just calls the
3884 queue-runner function directly.
3890 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3893 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3894 ignored. Though the use of -oP was forcing the writing of a pid file, it
3895 was always written to the default place.
3897 3. If the message "no IP address found for host xxxx" is generated during
3898 incoming verification, it is now followed by identification of the incoming
3899 connection (so you can more easily find what provoked it).
3901 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3903 5. Added some features to "harden" Exim a bit more against certain attacks:
3905 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3906 be put in Local/Makefile. This is like the never_users runtime option,
3907 but it cannot be overridden. The default setting is "root".
3909 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3910 prefix string with which any file named in a -C command line option
3913 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3914 is retained for -C and -D only if the caller of Exim is root. Without
3915 it, the exim user may also use -C and -D and retain privilege.
3917 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3918 command line option is disabled.
3920 6. Macro names set by the -D option must start with an upper case letter, just
3921 like macro names defined in the configuration file.
3923 7. Added "dereference=" facility to LDAP.
3925 8. Two instances of the typo "uknown" in the source files are fixed.
3927 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3928 the Configure-Makefile script screwed up while processing it.
3930 10. Incorporated PCRE 4.4.
3932 11. The SMTP synchronization check was not operating right at the start of an
3933 SMTP session. For example, it could not catch a HELO sent before the client
3934 waited for the greeting. There is now a check for outstanding input at the
3935 point when the greeting is written. Because of the duplex, asynchronous
3936 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3937 way, but not yet received, when the check is performed.
3939 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3940 on TCP/IP sockets, because this apparently causes some broken clients to
3943 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3944 unchanged) from the Cygwin maintainer.
3946 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3947 was supported instead of testing for mbx. Effectively a typo.
3949 15. The spa authenticator server code was not checking that the input it
3950 received was valid base64.
3952 16. The debug output line for the "set" modifier in ACLs was not showing the
3953 name of the variable that was being set.
3955 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3957 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3958 lookup on its own. Something else was needed to trigger the lookup. For
3959 example, a match in host_lookup or the need for a host name in a host list.
3960 Now, if $sender_host_name is referenced and the host name has not yet been
3961 looked up, a lookup is performed. If the lookup fails, the variable remains
3962 empty, and $host_lookup_failed is set to "1".
3964 19. Added "eqi" as a case-independent comparison operator.
3966 20. The saslauthd authentication condition could segfault if neither service
3967 nor realm was specified.
3969 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3970 error message that was logged was misleading, and incoming SMTP
3971 connections were dropped. The message is now more accurate, and temporary
3972 errors are given to SMTP connections.
3974 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3975 (except RSET) with a 421 error, until QUIT is received. However, it was
3976 failing to send a response to QUIT.
3978 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3979 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3980 ACL is now run after the helo_try_verify_hosts code.
3982 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3983 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3984 case-independent, so other case variants are also recognized). Apparently
3985 some systems use these upper case variants.
3987 25. If more than two messages were waiting for the same host, and a transport
3988 filter was specified for the transport, Exim sent two messages over the
3989 same TCP/IP connection, and then failed with "socket operation on non-
3990 socket" when it tried to send the third.
3992 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3994 27. The extern definition of crypt16() in expand.c was not being excluded when
3995 the OS had its own crypt16() function.
3997 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3998 as a preferred synonym for return_size_limit, both as an option and as an
4001 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4003 30. Changed the default configuration ACL to relax the local part checking rule
4004 for addresses that are not in any local domains. For these addresses,
4005 slashes and pipe symbols are allowed within local parts, but the sequence
4006 /../ is explicitly forbidden.
4008 31. SPA server authentication was not clearing the challenge buffer before
4011 32. log_message in a "warn" ACL statement was writing to the reject log as
4012 well as to the main log, which contradicts the documentation and doesn't
4013 seem right (because no rejection is happening). So I have stopped it.
4015 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4016 However, I am unable to do any testing of this.
4018 34. Fixed an infelicity in the appendfile transport. When checking directories
4019 for a mailbox, to see if any needed to be created, it was accidentally
4020 using path names with one or more superfluous leading slashes; tracing
4021 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4023 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4024 contents are added to the log line that is written for every discarded
4025 recipient. (Previously a log_message setting was ignored.)
4027 36. The ${quote: operator now quotes the string if it is empty.
4029 37. The install script runs exim in order to find its version number. If for
4030 some reason other than non-existence or emptiness, which it checks, it
4031 could not run './exim', it was installing it with an empty version number,
4032 i.e. as "exim-". This error state is now caught, and the installation is
4035 38. An argument was missing from the function that creates an error message
4036 when Exim fails to connect to the socket for saslauthd authentication.
4037 This could cause Exim to crash, or give a corrupted message.
4039 39. Added isip, isip4, and isip6 to ${if conditions.
4041 40. The ACL variables $acl_xx are now saved with the message, and can be
4042 accessed later in routers, transports, and filters.
4044 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4045 strings in the file are not string-expanded.
4047 42. If a MAIL command specified a SIZE value that was too large to fit into an
4048 int variable, the check against message_size_limit failed. Such values are
4049 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4050 day this will have to be increased, but I don't think I want to be around
4051 when emails are that large.
4058 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4059 iconv() is not standard in FreeBSD.
4061 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4062 IPv6 enabled. The observed symptom was a segmentation fault on return from
4063 the function os_common_find_running_interfaces() in src/os.c.
4065 3. In the check_special_case() function in daemon.c I had used "errno" as an
4066 argument name, which causes warnings on some systems. This was basically a
4067 typo, since it was named "eno" in the comments!
4069 4. The code that waits for the clock to tick (at a resolution of some fraction
4070 of a second) so as to ensure message-id uniqueness was always waiting for
4071 at least one whole tick, when it could have waited for less. [This is
4072 almost certainly not relevant at current processor speeds, where it is
4073 unlikely to ever wait at all. But we try to future-proof.]
4075 5. The function that sleeps for a time interval that includes fractions of a
4076 second contained a race. It did not block SIGALRM between setting the
4077 timer, and suspending (a couple of lines later). If the interval was short
4078 and the sigsuspend() was delayed until after it had expired, the suspension
4079 never ended. On busy systems this could lead to processes getting stuck for
4082 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4083 process, before it forks any delivery processes. The open lookup caching
4084 mechanism meant that the open file or database connection was passed into
4085 the delivery process. The problem was that delivery processes always tidy
4086 up cached lookup data. This could cause a problem for the next delivery
4087 process started by the queue runner, because the external queue runner
4088 process does not know about the closure. So the next delivery process
4089 still has data in the lookup cache. In the case of a file lookup, there was
4090 no problem because closing a file descriptor in a subprocess doesn't affect
4091 the parent. However, if the lookup was caching a connection to a database,
4092 the connection was closed, and the second delivery process was likely to
4093 see errors such as "PGSQL: query failed: server closed the connection
4094 unexpectedly". The problem has been fixed by closing all cached lookups
4095 in a queue runner before running a delivery process.
4097 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4098 seem to have the "const" qualifier which it has on other OS. I've
4101 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4102 *of the same type* (client or server) with the same public name that an
4103 error should be diagnosed.
4105 9. When Exim looked up a host name for an IP address, but failed to find the
4106 original IP address when looking up the host name (a safety check), it
4107 output the message "<ip address> does not match any IP for NULL", which was
4108 confusing, to say the least. The bug was that the host name should have
4109 appeared instead of "NULL".
4111 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4112 user that is built into the binary, and the -C or -D options is used, root
4113 privilege is dropped before the configuration file is read. In addition,
4114 logging is switched to stderr instead of the normal log files. If the
4115 configuration then re-defines the Exim user, the unprivileged environment
4116 is probably not what is expected, so Exim logs a panic warning message (but
4119 However, if deliver_drop_privilege is set, the unprivileged state may well
4120 be exactly what is intended, so the warning has been cut out in that case,
4121 and Exim is allowed to try to write to its normal log files.
4127 1. smtp_return_error_details was not giving details for temporary sender
4128 or receiver verification errors.
4130 2. Diagnose a configuration error if two authenticators have the same public
4133 3. Exim used not to create the message log file for a message until the first
4134 delivery attempt. This could be confusing when incoming messages were held
4135 for policy or load reasons. The message log file is now created at the time
4136 the message is received, and an initial "Received" line is written to it.
4138 4. The automatically generated man page for command line options had a minor
4139 bug that caused no ill effects; however, a more serious problem was that
4140 the procedure for building the man page automatically didn't always
4141 operate. Consequently, release 4.20 contains an out-of-date version. This
4142 shouldn't happen again.
4144 5. When building Exim with embedded Perl support, the script that builds the
4145 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4146 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4148 6. The freeze_tell option was not being used for messages that were frozen on
4149 arrival, either by an ACL or by local_scan().
4151 7. Added the smtp_incomplete_transaction log selector.
4153 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4154 was accepting AUTH without a new EHLO.
4156 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4157 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4160 10. Logging of TCP/IP connections (when configured) now happens in the main
4161 daemon process instead of the child process, so that the TCP/IP connection
4162 count is more accurate (but it can never be perfect).
4164 11. The use of "drop" in a nested ACL was not being handled correctly in the
4165 outer ACL. Now, if condition failure induced by the nested "drop" causes
4166 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4167 or "require"), the connection is dropped.
4169 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4170 that yield "discard" can now be used with an "accept" or a "discard" verb,
4171 but an error is generated for any others (because I can't see a useful way
4172 to define what should happen).
4174 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4175 are now processed in the same way as lines in the Exim configuration file.
4176 In particular, continuation lines are supported.
4178 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4180 15. Added -ti meaning -t -i.
4182 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4183 domains, and warn by logging if others are found.
4185 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4186 returned from GIFCONF when Exim is trying to find the list of interfaces on
4187 a host. The code in os.c has been modified to copy each ifreq to an aligned
4188 structure in all cases.
4190 Also, in some cases, the returned ifreq's were being copied to a 'struct
4191 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4192 means the last couple of bytes of an IPv6 address could be chopped if the
4193 ifreq contained only a normal sockaddr (14 bytes storage).
4195 18. Named domain lists were not supported in the hosts_treat_as_local option.
4196 An entry such as +xxxx was not recognized, and was treated as a literal
4199 19. Ensure that header lines added by a DATA ACL are included in the reject log
4200 if the ACL subsequently rejects the message.
4202 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4203 MD5 (which is deprecated).
4205 21. When testing a filter file using -bf, Exim was writing a message when it
4206 took the sender from a "From " line in the message, but it was not doing so
4207 when it took $return_path from a Return-Path: header line. It now does.
4209 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4210 with a valid header line field name (a series of printing characters
4211 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4213 23. Changed "disc" in the source to "disk" to conform to the documentation and
4214 the book and for uniformity.
4216 24. Ignore Sendmail's -Ooption=value command line item.
4218 25. When execve() failed while trying to run a command in a pipe transport,
4219 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4220 could be confused with a return value of 69 from the command itself. This
4221 has been changed to 127, the value the shell returns if it is asked to run
4222 a non-existent command. The wording for the related log line suggests a
4223 non-existent command as the problem.
4225 26. If received_header_text expands to an empty string, do not add a Received:
4226 header line to the message. (Well, it adds a token one on the spool, but
4227 marks it "old" so that it doesn't get used or transmitted.)
4229 27. Installed eximstats 1.28 (addition of -nt option).
4231 28. There was no check for failure on the call to getsockname() in the daemon
4232 code. This can fail if there is a shortage of resources on the system, with
4233 ENOMEM, for example. A temporary error is now given on failure.
4235 29. Contrary to the C standard, it seems that in some environments, the
4236 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4237 program. Exim now does this explicitly; it affects the formatting of
4238 timestamps using strftime().
4240 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4241 complaints. I've now initialized all the variables, to avoid this.
4243 32. Header lines added by a system filter were not being "seen" during
4244 transport-time rewrites.
4246 33. The info_callback() function passed to OpenSSL is set up with type void
4247 (*)(SSL *, int, int), as described somewhere. However, when calling the
4248 function (actually a macro) that sets it up, the type void(*)() is
4249 expected. I've put in a cast to prevent warnings from picky compilers.
4251 34. If a DNS black list lookup found a CNAME record, but there were no A
4252 records associated with the domain it pointed at, Exim crashed.
4254 35. If a DNS black list lookup returned more than one A record, Exim ignored
4255 all but the first. It now scans all returned addresses if a particular IP
4256 value is being sought. In this situation, the contents of the
4257 $dnslist_value variable are a list of all the addresses, separated by a
4260 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4261 to accept a host name and all its aliases if the forward lookup for any of
4262 them yielded the IP address of the incoming connection. Now it accepts only
4263 those names whose forward lookup yields the correct IP address. Any other
4264 names are discarded. This closes a loophole whereby a rogue DNS
4265 administrator could create reverse DNS records to break through a
4266 wildcarded host restriction in an ACL.
4268 37. If a user filter or a system filter that ran in a subprocess used any of
4269 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4270 the wrong values were passed to the pipe command ($thisaddress had the
4271 value of $0, $0 had the value of $1, etc). This bug was introduced by
4272 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4274 38. Improved the line breaking for long SMTP error messages from ACLs.
4275 Previously, if there was no break point between 40 and 75 characters, Exim
4276 left the rest of the message alone. Two changes have been made: (a) I've
4277 reduced the minimum length to 35 characters; (b) if it can't find a break
4278 point between 35 and 75 characters, it looks ahead and uses the first one
4279 that it finds. This may give the occasional overlong line, but at least the
4280 remaining text gets split now.
4282 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4283 file descriptors might be low, and that setting 1000 would always raise it.
4284 It turns out that in some environments, the limit is already over 1000 and
4285 that lowering it causes trouble. So now Exim takes care not to decrease it.
4287 40. When delivering a message, the value of $return_path is set to $sender_
4288 address at the start of routing (routers may change the value). By an
4289 oversight, this default was not being set up when an address was tested by
4290 -bt or -bv, which affected the outcome if any router or filter referred to
4293 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4294 log only when "message" or "log_message" are set. However, if one of the
4295 conditions was an address verification, or a call to a nested ACL, the
4296 messages generated by the underlying test were being passed through. This
4297 no longer happens. The underlying message is available in $acl_verify_
4298 message for both "message" and "log_message" expansions, so it can be
4299 passed through if needed.
4301 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4302 new expansion $bh_ to give the encoded byte string without charset
4303 translation. Translation happens only if iconv() is available; HAVE_ICONV
4304 indicates this at build time. HEADERS_CHARSET gives the charset to
4305 translate to; headers_charset can change it in the configuration, and
4306 "headers charset" can change it in an individual filter file.
4308 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4309 that creates RFC 2047 encoded "words" labels them as that charset instead
4310 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4311 expansion operator; (ii) when Exim creates a From: line for a local
4312 message; (iii) when a header line is rewritten to include a "phrase" part.
4314 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4315 buggy, causing it to skip the first lines of messages whose message ID
4316 ended in 'D'. This would not have bitten before Exim release 4.14, because
4317 message IDs were unlikely to end in 'D' before then. The effect was to have
4318 incorrect size information for certain domains.
4320 45. #include "config.h" was missing at the start of the crypt16.c module. This
4321 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4324 46. If there was a timeout during a "random" callout check, Exim treated it as
4325 a failure of the random address, and carried on sending RSET and the real
4326 address. If the delay was just some slowness somewhere, the response to the
4327 original RCPT would be taken as a response to RSET and so on, causing
4328 mayhem of various kinds.
4330 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4331 when I implemented it. It didn't allow for the fact that some option values
4332 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4333 the right test for positive values.
4335 48. Domain names in DNS records are case-independent. Exim always looks them up
4336 in lower case. Some resolvers return domain names in exactly the case they
4337 appear in the zone file, that is, they may contain uppercase letters. Not
4338 all resolvers do this - some return always lower case. Exim was treating a
4339 change of case by a resolver as a change of domain, similar to a widening
4340 of a domain abbreviation. This triggered its re-routing code and so it was
4341 trying to route what was effectively the same domain again. This normally
4342 caused routing to fail (because the router wouldn't handle the domain
4343 twice). Now Exim checks for this case specially, and just changes the
4344 casing of the domain that it ultimately uses when it transmits the message
4347 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4350 50. If a filter generated a file delivery with a non-absolute name (possible if
4351 no home directory exists for the router), the forbid_file option was not
4354 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4355 the existing equality matching.
4357 52. Exim was using ints instead of ino_t variables in some places where it was
4358 dealing with inode numbers.
4360 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4361 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4362 called TMPDIR, and if it finds it is different, it changes its value.
4364 54. The smtp_printf() function is now made available to local_scan() so
4365 additional output lines can be written before returning. There is also an
4366 smtp_fflush() function to enable the detection of a dropped connection.
4367 The variables smtp_input and smtp_batched_input are exported to
4370 55. Changed the default runtime configuration: the message "Unknown user"
4371 has been removed from the ACL, and instead placed on the localuser router,
4372 using the cannot_route_message feature. This means that any verification
4373 failures that generate their own messages won't get overridden. Similarly,
4374 the "Unrouteable address" message that was in the ACL for unverifiable
4375 relay addresses has also been removed.
4377 56. Added hosts_avoid_esmtp to the smtp transport.
4379 57. The exicyclog script was not checking for the esoteric option
4380 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4381 will work only if exicyclog is run under the appropriate euid.
4383 58. Following a discussion on the list, the rules by which Exim recognises line
4384 endings on incoming messages have been changed. The -dropcr and drop_cr
4385 options are now no-ops, retained only for backwards compatibility. The
4386 following line terminators are recognized: LF CRLF CR. However, special
4387 processing applies to CR:
4389 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4390 nor a local message in the state where . is a terminator.
4392 (ii) If a bare CR is encountered in a header line, an extra space is added
4393 after the line terminator so as not to end the header. The reasoning
4394 behind this is that bare CRs in header lines are most likely either
4395 to be mistakes, or people trying to play silly games.
4397 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4398 was being incorrectly given as 18 bytes larger than it should have been.
4399 This is a VOB (very old bug).
4401 60. This may never have affected anything current, but just in case it has:
4402 When the local host is found other than at the start of a list of hosts,
4403 the local host, those with the same MX, and any that follow, are discarded.
4404 When the list in question was part of a longer list of hosts, the following
4405 hosts (not currently being processed) were also being discarded. This no
4406 longer happens. I'm not sure if this situation could ever has previously
4409 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4412 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4413 header. This is something that is recommended in a new Internet Draft, and
4414 is something that is documented as being done by Sendmail. There are two
4415 possible values. For messages generated by the autoreply transport, Exim
4418 Auto-Submitted: auto-replied
4420 whereas for all other generated messages (e.g. bounces) it adds
4422 Auto-Submitted: auto-generated
4424 63. The "personal" condition in filters now includes a test for the
4425 Auto-Submitted: header. If it contains the string "auto-" the message it
4426 not considered personal.
4428 64. Added rcpt_include_affixes as a generic transport option.
4430 65. Added queue_only_override (default true).
4432 66. Added the syslog_duplication option.
4434 67. If what should have been the first header line of a message consisted of
4435 a space followed by a colon, Exim was mis-interpreting it as a header line.
4436 It isn't of course - it is syntactically invalid and should therefore be
4437 treated as the start of the message body. The misbehaviour could have
4438 caused a number of strange effects, including loss of data in subsequent
4439 header lines, and spool format errors.
4441 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4442 client host had authenticated. This control can now be exercised by an ACL
4443 for more flexibility.
4445 69. By default, callouts do not happen when testing with -bh. There is now a
4446 variant, -bhc, which does actually run the callout code, including
4447 consulting and updating the callout cache.
4449 70. Added support for saslauthd authentication, courtesy of Alexander
4452 71. If statvfs() failed on the spool or log directories while checking their
4453 size for availability, Exim confusingly gave the error "space shortage".
4454 Furthermore, in debugging mode it crashed with a floating point exception.
4455 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4456 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4457 serious problem, Exim now writes to the main and panic logs when this
4458 happens, with details of the failure. It then refuses to accept the
4459 incoming message, giving the message "spool directory problem" or "log
4460 directory problem" with a 421 code for SMTP messages.
4462 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4463 0, 1, and 2 exist, because some OS complain for execs without them (see
4464 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4465 descriptors. However, the code omitted to check that the open succeeded,
4466 causing mysterious errors if for some reason the permissions on /dev/null
4467 got screwed. Now Exim writes a message to the main and panic logs, and
4468 bombs out if it can't open /dev/null.
4470 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4471 interact so that it is all more flexible. It is supposed to remain
4472 backwards compatible. Also added extra_local_interfaces.
4474 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4475 to bomb out with an assertion failure - to the client this appears as a
4476 connection drop. This problem occurs in the part of the code that was taken
4477 from the Samba project. Fortunately, the assertion is in a very simple
4478 function, so I have fixed this by reproducing the function inline in the
4479 one place where it is called, and arranging for authentication to fail
4480 instead of killing the process with assert().
4482 75. The SPA client code was not working when the server requested OEM rather
4483 than Unicode encoding.
4485 76. Added code to make require_files with a specific uid setting more usable in
4486 the case where statting the file as root fails - usually a non-root-mounted
4487 NFS file system. When this happens and the failure is EACCES, Exim now
4488 forks a subprocess and does the per-uid checking as the relevant uid.
4490 77. Added process_log_path.
4492 78. If log_file_path was not explicitly set, a setting of check_log_space or
4493 check_log_inodes was ignored.
4495 79. If a space check for the spool or log partitions fails, the incident is now
4496 logged. Of course, in the latter case the data may get lost...
4498 80. Added the %p formatting code to string_format() so that it can be used to
4499 print addresses in debug_print(). Adjusted all the address printing in the
4500 debugging in store.c to use %p rather than %d.
4502 81. There was a concern that a line of code in smtp_in.c could overflow a
4503 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4504 initially expressed, the concern was not well-founded, because trailing
4505 spaces are removed early. However, if the trailing spaces were followed by
4506 a NULL, they did not get removed, so the overflow was possible. Two fixes
4509 (a) I re-wrote the offending code in a cleaner fashion.
4510 (b) If an incoming SMTP command contains a NULL character, it is rejected
4513 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4514 runs initgroups(), so that if the Exim user is in any additional groups,
4515 they will be used during message reception.
4521 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob